diff --git "a/checkpoint-14000/trainer_state.json" "b/checkpoint-14000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-14000/trainer_state.json"
@@ -0,0 +1,98033 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.6264885274469938,
+  "eval_steps": 500,
+  "global_step": 14000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00011617775196049956,
+      "grad_norm": 1.1400724649429321,
+      "learning_rate": 0.0001,
+      "loss": 1.9564,
+      "step": 1
+    },
+    {
+      "epoch": 0.00023235550392099912,
+      "grad_norm": 0.5516552925109863,
+      "learning_rate": 0.0001,
+      "loss": 1.9434,
+      "step": 2
+    },
+    {
+      "epoch": 0.0003485332558814987,
+      "grad_norm": 2.07415771484375,
+      "learning_rate": 0.0001,
+      "loss": 1.9873,
+      "step": 3
+    },
+    {
+      "epoch": 0.00046471100784199824,
+      "grad_norm": 0.6526162028312683,
+      "learning_rate": 0.0001,
+      "loss": 1.9786,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005808887598024978,
+      "grad_norm": 0.5003806352615356,
+      "learning_rate": 0.0001,
+      "loss": 1.8828,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006970665117629974,
+      "grad_norm": 0.5392526388168335,
+      "learning_rate": 0.0001,
+      "loss": 1.8913,
+      "step": 6
+    },
+    {
+      "epoch": 0.000813244263723497,
+      "grad_norm": 0.4057665765285492,
+      "learning_rate": 0.0001,
+      "loss": 1.7552,
+      "step": 7
+    },
+    {
+      "epoch": 0.0009294220156839965,
+      "grad_norm": 0.42006543278694153,
+      "learning_rate": 0.0001,
+      "loss": 2.0082,
+      "step": 8
+    },
+    {
+      "epoch": 0.001045599767644496,
+      "grad_norm": 1.1451431512832642,
+      "learning_rate": 0.0001,
+      "loss": 1.8661,
+      "step": 9
+    },
+    {
+      "epoch": 0.0011617775196049957,
+      "grad_norm": 0.44178342819213867,
+      "learning_rate": 0.0001,
+      "loss": 1.914,
+      "step": 10
+    },
+    {
+      "epoch": 0.0012779552715654952,
+      "grad_norm": 0.3882751762866974,
+      "learning_rate": 0.0001,
+      "loss": 1.902,
+      "step": 11
+    },
+    {
+      "epoch": 0.0013941330235259948,
+      "grad_norm": 0.35292503237724304,
+      "learning_rate": 0.0001,
+      "loss": 1.89,
+      "step": 12
+    },
+    {
+      "epoch": 0.0015103107754864944,
+      "grad_norm": 0.4178386926651001,
+      "learning_rate": 0.0001,
+      "loss": 1.8888,
+      "step": 13
+    },
+    {
+      "epoch": 0.001626488527446994,
+      "grad_norm": 0.39220985770225525,
+      "learning_rate": 0.0001,
+      "loss": 1.7775,
+      "step": 14
+    },
+    {
+      "epoch": 0.0017426662794074934,
+      "grad_norm": 0.4134705364704132,
+      "learning_rate": 0.0001,
+      "loss": 1.8794,
+      "step": 15
+    },
+    {
+      "epoch": 0.001858844031367993,
+      "grad_norm": 0.3392408788204193,
+      "learning_rate": 0.0001,
+      "loss": 1.6998,
+      "step": 16
+    },
+    {
+      "epoch": 0.0019750217833284928,
+      "grad_norm": 0.3672613203525543,
+      "learning_rate": 0.0001,
+      "loss": 2.0741,
+      "step": 17
+    },
+    {
+      "epoch": 0.002091199535288992,
+      "grad_norm": 0.4194653630256653,
+      "learning_rate": 0.0001,
+      "loss": 1.9617,
+      "step": 18
+    },
+    {
+      "epoch": 0.002207377287249492,
+      "grad_norm": 0.38323554396629333,
+      "learning_rate": 0.0001,
+      "loss": 1.8911,
+      "step": 19
+    },
+    {
+      "epoch": 0.0023235550392099913,
+      "grad_norm": 0.35623419284820557,
+      "learning_rate": 0.0001,
+      "loss": 1.8062,
+      "step": 20
+    },
+    {
+      "epoch": 0.0024397327911704907,
+      "grad_norm": 0.3997187614440918,
+      "learning_rate": 0.0001,
+      "loss": 2.1459,
+      "step": 21
+    },
+    {
+      "epoch": 0.0025559105431309905,
+      "grad_norm": 0.31873300671577454,
+      "learning_rate": 0.0001,
+      "loss": 1.813,
+      "step": 22
+    },
+    {
+      "epoch": 0.00267208829509149,
+      "grad_norm": 0.36987119913101196,
+      "learning_rate": 0.0001,
+      "loss": 1.8538,
+      "step": 23
+    },
+    {
+      "epoch": 0.0027882660470519897,
+      "grad_norm": 0.38269954919815063,
+      "learning_rate": 0.0001,
+      "loss": 1.9971,
+      "step": 24
+    },
+    {
+      "epoch": 0.002904443799012489,
+      "grad_norm": 0.3547232449054718,
+      "learning_rate": 0.0001,
+      "loss": 1.9081,
+      "step": 25
+    },
+    {
+      "epoch": 0.003020621550972989,
+      "grad_norm": 0.33162739872932434,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 26
+    },
+    {
+      "epoch": 0.003136799302933488,
+      "grad_norm": 0.3379284143447876,
+      "learning_rate": 0.0001,
+      "loss": 1.856,
+      "step": 27
+    },
+    {
+      "epoch": 0.003252977054893988,
+      "grad_norm": 0.35247287154197693,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 28
+    },
+    {
+      "epoch": 0.0033691548068544874,
+      "grad_norm": 0.3893706500530243,
+      "learning_rate": 0.0001,
+      "loss": 2.0282,
+      "step": 29
+    },
+    {
+      "epoch": 0.0034853325588149867,
+      "grad_norm": 0.39306116104125977,
+      "learning_rate": 0.0001,
+      "loss": 1.7966,
+      "step": 30
+    },
+    {
+      "epoch": 0.0036015103107754865,
+      "grad_norm": 0.36439645290374756,
+      "learning_rate": 0.0001,
+      "loss": 1.9552,
+      "step": 31
+    },
+    {
+      "epoch": 0.003717688062735986,
+      "grad_norm": 0.3758845925331116,
+      "learning_rate": 0.0001,
+      "loss": 1.9639,
+      "step": 32
+    },
+    {
+      "epoch": 0.0038338658146964857,
+      "grad_norm": 0.3357931971549988,
+      "learning_rate": 0.0001,
+      "loss": 1.8929,
+      "step": 33
+    },
+    {
+      "epoch": 0.0039500435666569855,
+      "grad_norm": 0.3393707573413849,
+      "learning_rate": 0.0001,
+      "loss": 1.7972,
+      "step": 34
+    },
+    {
+      "epoch": 0.004066221318617485,
+      "grad_norm": 0.3457406163215637,
+      "learning_rate": 0.0001,
+      "loss": 1.7516,
+      "step": 35
+    },
+    {
+      "epoch": 0.004182399070577984,
+      "grad_norm": 0.33674880862236023,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 36
+    },
+    {
+      "epoch": 0.004298576822538484,
+      "grad_norm": 0.33614569902420044,
+      "learning_rate": 0.0001,
+      "loss": 1.8855,
+      "step": 37
+    },
+    {
+      "epoch": 0.004414754574498984,
+      "grad_norm": 0.34421899914741516,
+      "learning_rate": 0.0001,
+      "loss": 1.8091,
+      "step": 38
+    },
+    {
+      "epoch": 0.004530932326459483,
+      "grad_norm": 0.39817526936531067,
+      "learning_rate": 0.0001,
+      "loss": 1.941,
+      "step": 39
+    },
+    {
+      "epoch": 0.004647110078419983,
+      "grad_norm": 0.3206139802932739,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 40
+    },
+    {
+      "epoch": 0.004763287830380482,
+      "grad_norm": 0.37263521552085876,
+      "learning_rate": 0.0001,
+      "loss": 1.8275,
+      "step": 41
+    },
+    {
+      "epoch": 0.004879465582340981,
+      "grad_norm": 0.3837355077266693,
+      "learning_rate": 0.0001,
+      "loss": 1.9595,
+      "step": 42
+    },
+    {
+      "epoch": 0.004995643334301482,
+      "grad_norm": 0.33184581995010376,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 43
+    },
+    {
+      "epoch": 0.005111821086261981,
+      "grad_norm": 0.32549911737442017,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 44
+    },
+    {
+      "epoch": 0.00522799883822248,
+      "grad_norm": 0.35189497470855713,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 45
+    },
+    {
+      "epoch": 0.00534417659018298,
+      "grad_norm": 0.34523481130599976,
+      "learning_rate": 0.0001,
+      "loss": 1.9006,
+      "step": 46
+    },
+    {
+      "epoch": 0.00546035434214348,
+      "grad_norm": 0.3432101905345917,
+      "learning_rate": 0.0001,
+      "loss": 2.0098,
+      "step": 47
+    },
+    {
+      "epoch": 0.005576532094103979,
+      "grad_norm": 0.3387945294380188,
+      "learning_rate": 0.0001,
+      "loss": 1.8139,
+      "step": 48
+    },
+    {
+      "epoch": 0.005692709846064479,
+      "grad_norm": 0.33659058809280396,
+      "learning_rate": 0.0001,
+      "loss": 1.8409,
+      "step": 49
+    },
+    {
+      "epoch": 0.005808887598024978,
+      "grad_norm": 0.32735776901245117,
+      "learning_rate": 0.0001,
+      "loss": 1.7861,
+      "step": 50
+    },
+    {
+      "epoch": 0.005925065349985477,
+      "grad_norm": 0.35173168778419495,
+      "learning_rate": 0.0001,
+      "loss": 1.8177,
+      "step": 51
+    },
+    {
+      "epoch": 0.006041243101945978,
+      "grad_norm": 0.3859328031539917,
+      "learning_rate": 0.0001,
+      "loss": 2.0822,
+      "step": 52
+    },
+    {
+      "epoch": 0.006157420853906477,
+      "grad_norm": 0.3218703866004944,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 53
+    },
+    {
+      "epoch": 0.006273598605866976,
+      "grad_norm": 0.34024327993392944,
+      "learning_rate": 0.0001,
+      "loss": 1.8432,
+      "step": 54
+    },
+    {
+      "epoch": 0.006389776357827476,
+      "grad_norm": 0.3500777781009674,
+      "learning_rate": 0.0001,
+      "loss": 1.946,
+      "step": 55
+    },
+    {
+      "epoch": 0.006505954109787976,
+      "grad_norm": 0.35199397802352905,
+      "learning_rate": 0.0001,
+      "loss": 1.7764,
+      "step": 56
+    },
+    {
+      "epoch": 0.006622131861748475,
+      "grad_norm": 0.3386881947517395,
+      "learning_rate": 0.0001,
+      "loss": 1.8424,
+      "step": 57
+    },
+    {
+      "epoch": 0.006738309613708975,
+      "grad_norm": 0.333283931016922,
+      "learning_rate": 0.0001,
+      "loss": 1.8125,
+      "step": 58
+    },
+    {
+      "epoch": 0.006854487365669474,
+      "grad_norm": 0.32521334290504456,
+      "learning_rate": 0.0001,
+      "loss": 1.7354,
+      "step": 59
+    },
+    {
+      "epoch": 0.0069706651176299735,
+      "grad_norm": 0.36899396777153015,
+      "learning_rate": 0.0001,
+      "loss": 1.8834,
+      "step": 60
+    },
+    {
+      "epoch": 0.007086842869590474,
+      "grad_norm": 0.33825087547302246,
+      "learning_rate": 0.0001,
+      "loss": 1.905,
+      "step": 61
+    },
+    {
+      "epoch": 0.007203020621550973,
+      "grad_norm": 0.3499239981174469,
+      "learning_rate": 0.0001,
+      "loss": 1.8053,
+      "step": 62
+    },
+    {
+      "epoch": 0.0073191983735114725,
+      "grad_norm": 0.3182366192340851,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 63
+    },
+    {
+      "epoch": 0.007435376125471972,
+      "grad_norm": 0.34978607296943665,
+      "learning_rate": 0.0001,
+      "loss": 1.9504,
+      "step": 64
+    },
+    {
+      "epoch": 0.007551553877432472,
+      "grad_norm": 0.3370610177516937,
+      "learning_rate": 0.0001,
+      "loss": 1.8897,
+      "step": 65
+    },
+    {
+      "epoch": 0.0076677316293929714,
+      "grad_norm": 0.3260805606842041,
+      "learning_rate": 0.0001,
+      "loss": 1.8774,
+      "step": 66
+    },
+    {
+      "epoch": 0.007783909381353471,
+      "grad_norm": 0.32136961817741394,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 67
+    },
+    {
+      "epoch": 0.007900087133313971,
+      "grad_norm": 0.34404799342155457,
+      "learning_rate": 0.0001,
+      "loss": 1.9441,
+      "step": 68
+    },
+    {
+      "epoch": 0.00801626488527447,
+      "grad_norm": 0.3686494529247284,
+      "learning_rate": 0.0001,
+      "loss": 2.0653,
+      "step": 69
+    },
+    {
+      "epoch": 0.00813244263723497,
+      "grad_norm": 0.34148523211479187,
+      "learning_rate": 0.0001,
+      "loss": 1.9369,
+      "step": 70
+    },
+    {
+      "epoch": 0.00824862038919547,
+      "grad_norm": 0.3419267535209656,
+      "learning_rate": 0.0001,
+      "loss": 1.8754,
+      "step": 71
+    },
+    {
+      "epoch": 0.008364798141155969,
+      "grad_norm": 0.34724754095077515,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 72
+    },
+    {
+      "epoch": 0.008480975893116468,
+      "grad_norm": 0.3402698338031769,
+      "learning_rate": 0.0001,
+      "loss": 1.8966,
+      "step": 73
+    },
+    {
+      "epoch": 0.008597153645076967,
+      "grad_norm": 0.3450843393802643,
+      "learning_rate": 0.0001,
+      "loss": 1.9082,
+      "step": 74
+    },
+    {
+      "epoch": 0.008713331397037467,
+      "grad_norm": 0.3541632294654846,
+      "learning_rate": 0.0001,
+      "loss": 1.8932,
+      "step": 75
+    },
+    {
+      "epoch": 0.008829509148997968,
+      "grad_norm": 0.3470132648944855,
+      "learning_rate": 0.0001,
+      "loss": 1.7195,
+      "step": 76
+    },
+    {
+      "epoch": 0.008945686900958467,
+      "grad_norm": 0.33801767230033875,
+      "learning_rate": 0.0001,
+      "loss": 1.8521,
+      "step": 77
+    },
+    {
+      "epoch": 0.009061864652918966,
+      "grad_norm": 0.3319573402404785,
+      "learning_rate": 0.0001,
+      "loss": 1.7786,
+      "step": 78
+    },
+    {
+      "epoch": 0.009178042404879466,
+      "grad_norm": 0.35573163628578186,
+      "learning_rate": 0.0001,
+      "loss": 1.8646,
+      "step": 79
+    },
+    {
+      "epoch": 0.009294220156839965,
+      "grad_norm": 0.36378416419029236,
+      "learning_rate": 0.0001,
+      "loss": 2.0346,
+      "step": 80
+    },
+    {
+      "epoch": 0.009410397908800465,
+      "grad_norm": 0.3243565559387207,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 81
+    },
+    {
+      "epoch": 0.009526575660760964,
+      "grad_norm": 0.32637423276901245,
+      "learning_rate": 0.0001,
+      "loss": 1.8714,
+      "step": 82
+    },
+    {
+      "epoch": 0.009642753412721463,
+      "grad_norm": 0.3473580479621887,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 83
+    },
+    {
+      "epoch": 0.009758931164681963,
+      "grad_norm": 0.3237003982067108,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 84
+    },
+    {
+      "epoch": 0.009875108916642464,
+      "grad_norm": 0.3218970000743866,
+      "learning_rate": 0.0001,
+      "loss": 1.7922,
+      "step": 85
+    },
+    {
+      "epoch": 0.009991286668602963,
+      "grad_norm": 0.3358110189437866,
+      "learning_rate": 0.0001,
+      "loss": 1.7284,
+      "step": 86
+    },
+    {
+      "epoch": 0.010107464420563463,
+      "grad_norm": 0.34907999634742737,
+      "learning_rate": 0.0001,
+      "loss": 1.8581,
+      "step": 87
+    },
+    {
+      "epoch": 0.010223642172523962,
+      "grad_norm": 0.3303886950016022,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 88
+    },
+    {
+      "epoch": 0.010339819924484461,
+      "grad_norm": 0.33527591824531555,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 89
+    },
+    {
+      "epoch": 0.01045599767644496,
+      "grad_norm": 0.3530234098434448,
+      "learning_rate": 0.0001,
+      "loss": 1.8076,
+      "step": 90
+    },
+    {
+      "epoch": 0.01057217542840546,
+      "grad_norm": 0.35667964816093445,
+      "learning_rate": 0.0001,
+      "loss": 1.9292,
+      "step": 91
+    },
+    {
+      "epoch": 0.01068835318036596,
+      "grad_norm": 0.32774579524993896,
+      "learning_rate": 0.0001,
+      "loss": 1.902,
+      "step": 92
+    },
+    {
+      "epoch": 0.010804530932326459,
+      "grad_norm": 0.3343551456928253,
+      "learning_rate": 0.0001,
+      "loss": 1.8557,
+      "step": 93
+    },
+    {
+      "epoch": 0.01092070868428696,
+      "grad_norm": 0.3439468741416931,
+      "learning_rate": 0.0001,
+      "loss": 1.8196,
+      "step": 94
+    },
+    {
+      "epoch": 0.01103688643624746,
+      "grad_norm": 0.31505489349365234,
+      "learning_rate": 0.0001,
+      "loss": 1.7903,
+      "step": 95
+    },
+    {
+      "epoch": 0.011153064188207959,
+      "grad_norm": 0.3402290344238281,
+      "learning_rate": 0.0001,
+      "loss": 1.7981,
+      "step": 96
+    },
+    {
+      "epoch": 0.011269241940168458,
+      "grad_norm": 0.31825143098831177,
+      "learning_rate": 0.0001,
+      "loss": 1.7931,
+      "step": 97
+    },
+    {
+      "epoch": 0.011385419692128957,
+      "grad_norm": 0.32378312945365906,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 98
+    },
+    {
+      "epoch": 0.011501597444089457,
+      "grad_norm": 0.33641722798347473,
+      "learning_rate": 0.0001,
+      "loss": 1.8258,
+      "step": 99
+    },
+    {
+      "epoch": 0.011617775196049956,
+      "grad_norm": 0.3626033663749695,
+      "learning_rate": 0.0001,
+      "loss": 1.9063,
+      "step": 100
+    },
+    {
+      "epoch": 0.011733952948010455,
+      "grad_norm": 0.3551986813545227,
+      "learning_rate": 0.0001,
+      "loss": 1.876,
+      "step": 101
+    },
+    {
+      "epoch": 0.011850130699970955,
+      "grad_norm": 0.364032506942749,
+      "learning_rate": 0.0001,
+      "loss": 1.8289,
+      "step": 102
+    },
+    {
+      "epoch": 0.011966308451931456,
+      "grad_norm": 0.3228561282157898,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 103
+    },
+    {
+      "epoch": 0.012082486203891955,
+      "grad_norm": 0.34352198243141174,
+      "learning_rate": 0.0001,
+      "loss": 1.8273,
+      "step": 104
+    },
+    {
+      "epoch": 0.012198663955852455,
+      "grad_norm": 0.3475358784198761,
+      "learning_rate": 0.0001,
+      "loss": 1.6767,
+      "step": 105
+    },
+    {
+      "epoch": 0.012314841707812954,
+      "grad_norm": 0.3112235367298126,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 106
+    },
+    {
+      "epoch": 0.012431019459773453,
+      "grad_norm": 0.3369406759738922,
+      "learning_rate": 0.0001,
+      "loss": 1.9139,
+      "step": 107
+    },
+    {
+      "epoch": 0.012547197211733953,
+      "grad_norm": 0.3227371573448181,
+      "learning_rate": 0.0001,
+      "loss": 1.8202,
+      "step": 108
+    },
+    {
+      "epoch": 0.012663374963694452,
+      "grad_norm": 0.36206531524658203,
+      "learning_rate": 0.0001,
+      "loss": 1.9441,
+      "step": 109
+    },
+    {
+      "epoch": 0.012779552715654952,
+      "grad_norm": 0.3347959518432617,
+      "learning_rate": 0.0001,
+      "loss": 1.9206,
+      "step": 110
+    },
+    {
+      "epoch": 0.012895730467615451,
+      "grad_norm": 0.32953107357025146,
+      "learning_rate": 0.0001,
+      "loss": 1.9498,
+      "step": 111
+    },
+    {
+      "epoch": 0.013011908219575952,
+      "grad_norm": 0.30254584550857544,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 112
+    },
+    {
+      "epoch": 0.013128085971536451,
+      "grad_norm": 0.3403797149658203,
+      "learning_rate": 0.0001,
+      "loss": 1.8821,
+      "step": 113
+    },
+    {
+      "epoch": 0.01324426372349695,
+      "grad_norm": 0.31918781995773315,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 114
+    },
+    {
+      "epoch": 0.01336044147545745,
+      "grad_norm": 0.3319108486175537,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 115
+    },
+    {
+      "epoch": 0.01347661922741795,
+      "grad_norm": 0.34339022636413574,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 116
+    },
+    {
+      "epoch": 0.013592796979378449,
+      "grad_norm": 0.33843010663986206,
+      "learning_rate": 0.0001,
+      "loss": 1.8448,
+      "step": 117
+    },
+    {
+      "epoch": 0.013708974731338948,
+      "grad_norm": 0.32837411761283875,
+      "learning_rate": 0.0001,
+      "loss": 1.7597,
+      "step": 118
+    },
+    {
+      "epoch": 0.013825152483299448,
+      "grad_norm": 0.3620765507221222,
+      "learning_rate": 0.0001,
+      "loss": 1.8675,
+      "step": 119
+    },
+    {
+      "epoch": 0.013941330235259947,
+      "grad_norm": 0.33245301246643066,
+      "learning_rate": 0.0001,
+      "loss": 1.8119,
+      "step": 120
+    },
+    {
+      "epoch": 0.014057507987220448,
+      "grad_norm": 0.34569036960601807,
+      "learning_rate": 0.0001,
+      "loss": 1.9758,
+      "step": 121
+    },
+    {
+      "epoch": 0.014173685739180947,
+      "grad_norm": 0.374098539352417,
+      "learning_rate": 0.0001,
+      "loss": 1.78,
+      "step": 122
+    },
+    {
+      "epoch": 0.014289863491141447,
+      "grad_norm": 0.3418053686618805,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 123
+    },
+    {
+      "epoch": 0.014406041243101946,
+      "grad_norm": 0.3384765088558197,
+      "learning_rate": 0.0001,
+      "loss": 1.7649,
+      "step": 124
+    },
+    {
+      "epoch": 0.014522218995062446,
+      "grad_norm": 0.3590647280216217,
+      "learning_rate": 0.0001,
+      "loss": 1.9065,
+      "step": 125
+    },
+    {
+      "epoch": 0.014638396747022945,
+      "grad_norm": 0.3395555019378662,
+      "learning_rate": 0.0001,
+      "loss": 1.8293,
+      "step": 126
+    },
+    {
+      "epoch": 0.014754574498983444,
+      "grad_norm": 0.3610383868217468,
+      "learning_rate": 0.0001,
+      "loss": 1.8426,
+      "step": 127
+    },
+    {
+      "epoch": 0.014870752250943944,
+      "grad_norm": 0.33869925141334534,
+      "learning_rate": 0.0001,
+      "loss": 1.7425,
+      "step": 128
+    },
+    {
+      "epoch": 0.014986930002904443,
+      "grad_norm": 0.338553249835968,
+      "learning_rate": 0.0001,
+      "loss": 1.8638,
+      "step": 129
+    },
+    {
+      "epoch": 0.015103107754864944,
+      "grad_norm": 0.3297845125198364,
+      "learning_rate": 0.0001,
+      "loss": 1.8374,
+      "step": 130
+    },
+    {
+      "epoch": 0.015219285506825444,
+      "grad_norm": 0.3697233498096466,
+      "learning_rate": 0.0001,
+      "loss": 1.9652,
+      "step": 131
+    },
+    {
+      "epoch": 0.015335463258785943,
+      "grad_norm": 0.3488331139087677,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 132
+    },
+    {
+      "epoch": 0.015451641010746442,
+      "grad_norm": 0.3396114706993103,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 133
+    },
+    {
+      "epoch": 0.015567818762706942,
+      "grad_norm": 0.34215307235717773,
+      "learning_rate": 0.0001,
+      "loss": 1.9031,
+      "step": 134
+    },
+    {
+      "epoch": 0.015683996514667443,
+      "grad_norm": 0.32633256912231445,
+      "learning_rate": 0.0001,
+      "loss": 1.794,
+      "step": 135
+    },
+    {
+      "epoch": 0.015800174266627942,
+      "grad_norm": 0.3819684088230133,
+      "learning_rate": 0.0001,
+      "loss": 1.7509,
+      "step": 136
+    },
+    {
+      "epoch": 0.01591635201858844,
+      "grad_norm": 0.3342839479446411,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 137
+    },
+    {
+      "epoch": 0.01603252977054894,
+      "grad_norm": 0.32045823335647583,
+      "learning_rate": 0.0001,
+      "loss": 1.9446,
+      "step": 138
+    },
+    {
+      "epoch": 0.01614870752250944,
+      "grad_norm": 0.3185892701148987,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 139
+    },
+    {
+      "epoch": 0.01626488527446994,
+      "grad_norm": 0.3285033106803894,
+      "learning_rate": 0.0001,
+      "loss": 1.8363,
+      "step": 140
+    },
+    {
+      "epoch": 0.01638106302643044,
+      "grad_norm": 0.35822629928588867,
+      "learning_rate": 0.0001,
+      "loss": 1.7425,
+      "step": 141
+    },
+    {
+      "epoch": 0.01649724077839094,
+      "grad_norm": 0.3599529266357422,
+      "learning_rate": 0.0001,
+      "loss": 1.7378,
+      "step": 142
+    },
+    {
+      "epoch": 0.016613418530351438,
+      "grad_norm": 0.3105633556842804,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 143
+    },
+    {
+      "epoch": 0.016729596282311937,
+      "grad_norm": 0.3310804069042206,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 144
+    },
+    {
+      "epoch": 0.016845774034272436,
+      "grad_norm": 0.32904812693595886,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 145
+    },
+    {
+      "epoch": 0.016961951786232936,
+      "grad_norm": 0.3462965190410614,
+      "learning_rate": 0.0001,
+      "loss": 1.8588,
+      "step": 146
+    },
+    {
+      "epoch": 0.017078129538193435,
+      "grad_norm": 0.3536953926086426,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 147
+    },
+    {
+      "epoch": 0.017194307290153935,
+      "grad_norm": 0.3167782723903656,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 148
+    },
+    {
+      "epoch": 0.017310485042114434,
+      "grad_norm": 0.3438095450401306,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 149
+    },
+    {
+      "epoch": 0.017426662794074933,
+      "grad_norm": 0.32976609468460083,
+      "learning_rate": 0.0001,
+      "loss": 1.8282,
+      "step": 150
+    },
+    {
+      "epoch": 0.017542840546035433,
+      "grad_norm": 0.36160793900489807,
+      "learning_rate": 0.0001,
+      "loss": 1.982,
+      "step": 151
+    },
+    {
+      "epoch": 0.017659018297995936,
+      "grad_norm": 0.31704050302505493,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 152
+    },
+    {
+      "epoch": 0.017775196049956435,
+      "grad_norm": 0.3186262547969818,
+      "learning_rate": 0.0001,
+      "loss": 1.7416,
+      "step": 153
+    },
+    {
+      "epoch": 0.017891373801916934,
+      "grad_norm": 0.3476766049861908,
+      "learning_rate": 0.0001,
+      "loss": 1.822,
+      "step": 154
+    },
+    {
+      "epoch": 0.018007551553877434,
+      "grad_norm": 0.35061368346214294,
+      "learning_rate": 0.0001,
+      "loss": 1.7696,
+      "step": 155
+    },
+    {
+      "epoch": 0.018123729305837933,
+      "grad_norm": 0.33829307556152344,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 156
+    },
+    {
+      "epoch": 0.018239907057798432,
+      "grad_norm": 0.3412603735923767,
+      "learning_rate": 0.0001,
+      "loss": 1.8324,
+      "step": 157
+    },
+    {
+      "epoch": 0.01835608480975893,
+      "grad_norm": 0.3786666989326477,
+      "learning_rate": 0.0001,
+      "loss": 1.9237,
+      "step": 158
+    },
+    {
+      "epoch": 0.01847226256171943,
+      "grad_norm": 0.31077513098716736,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 159
+    },
+    {
+      "epoch": 0.01858844031367993,
+      "grad_norm": 0.32142356038093567,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 160
+    },
+    {
+      "epoch": 0.01870461806564043,
+      "grad_norm": 0.30329957604408264,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 161
+    },
+    {
+      "epoch": 0.01882079581760093,
+      "grad_norm": 0.3338776230812073,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 162
+    },
+    {
+      "epoch": 0.01893697356956143,
+      "grad_norm": 0.32876482605934143,
+      "learning_rate": 0.0001,
+      "loss": 1.8144,
+      "step": 163
+    },
+    {
+      "epoch": 0.019053151321521928,
+      "grad_norm": 0.32878580689430237,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 164
+    },
+    {
+      "epoch": 0.019169329073482427,
+      "grad_norm": 0.33372652530670166,
+      "learning_rate": 0.0001,
+      "loss": 1.7709,
+      "step": 165
+    },
+    {
+      "epoch": 0.019285506825442927,
+      "grad_norm": 0.3314981460571289,
+      "learning_rate": 0.0001,
+      "loss": 1.828,
+      "step": 166
+    },
+    {
+      "epoch": 0.019401684577403426,
+      "grad_norm": 0.3250749707221985,
+      "learning_rate": 0.0001,
+      "loss": 1.7861,
+      "step": 167
+    },
+    {
+      "epoch": 0.019517862329363925,
+      "grad_norm": 0.31878861784935,
+      "learning_rate": 0.0001,
+      "loss": 1.9088,
+      "step": 168
+    },
+    {
+      "epoch": 0.019634040081324425,
+      "grad_norm": 0.3233966827392578,
+      "learning_rate": 0.0001,
+      "loss": 1.759,
+      "step": 169
+    },
+    {
+      "epoch": 0.019750217833284928,
+      "grad_norm": 0.3260466754436493,
+      "learning_rate": 0.0001,
+      "loss": 1.9208,
+      "step": 170
+    },
+    {
+      "epoch": 0.019866395585245427,
+      "grad_norm": 0.33691638708114624,
+      "learning_rate": 0.0001,
+      "loss": 1.8521,
+      "step": 171
+    },
+    {
+      "epoch": 0.019982573337205926,
+      "grad_norm": 0.33654195070266724,
+      "learning_rate": 0.0001,
+      "loss": 1.7191,
+      "step": 172
+    },
+    {
+      "epoch": 0.020098751089166426,
+      "grad_norm": 0.32309457659721375,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 173
+    },
+    {
+      "epoch": 0.020214928841126925,
+      "grad_norm": 0.33408665657043457,
+      "learning_rate": 0.0001,
+      "loss": 1.8275,
+      "step": 174
+    },
+    {
+      "epoch": 0.020331106593087424,
+      "grad_norm": 0.32850080728530884,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 175
+    },
+    {
+      "epoch": 0.020447284345047924,
+      "grad_norm": 0.35644981265068054,
+      "learning_rate": 0.0001,
+      "loss": 1.9859,
+      "step": 176
+    },
+    {
+      "epoch": 0.020563462097008423,
+      "grad_norm": 0.36484628915786743,
+      "learning_rate": 0.0001,
+      "loss": 1.8343,
+      "step": 177
+    },
+    {
+      "epoch": 0.020679639848968923,
+      "grad_norm": 0.3122706413269043,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 178
+    },
+    {
+      "epoch": 0.020795817600929422,
+      "grad_norm": 0.35981225967407227,
+      "learning_rate": 0.0001,
+      "loss": 1.9379,
+      "step": 179
+    },
+    {
+      "epoch": 0.02091199535288992,
+      "grad_norm": 0.3504844009876251,
+      "learning_rate": 0.0001,
+      "loss": 1.7112,
+      "step": 180
+    },
+    {
+      "epoch": 0.02102817310485042,
+      "grad_norm": 0.3608277142047882,
+      "learning_rate": 0.0001,
+      "loss": 1.705,
+      "step": 181
+    },
+    {
+      "epoch": 0.02114435085681092,
+      "grad_norm": 0.3393659293651581,
+      "learning_rate": 0.0001,
+      "loss": 1.87,
+      "step": 182
+    },
+    {
+      "epoch": 0.02126052860877142,
+      "grad_norm": 0.34214162826538086,
+      "learning_rate": 0.0001,
+      "loss": 1.8054,
+      "step": 183
+    },
+    {
+      "epoch": 0.02137670636073192,
+      "grad_norm": 0.33791059255599976,
+      "learning_rate": 0.0001,
+      "loss": 1.9013,
+      "step": 184
+    },
+    {
+      "epoch": 0.021492884112692418,
+      "grad_norm": 0.3184012770652771,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 185
+    },
+    {
+      "epoch": 0.021609061864652918,
+      "grad_norm": 0.3612908124923706,
+      "learning_rate": 0.0001,
+      "loss": 1.8917,
+      "step": 186
+    },
+    {
+      "epoch": 0.021725239616613417,
+      "grad_norm": 0.3533594608306885,
+      "learning_rate": 0.0001,
+      "loss": 1.8029,
+      "step": 187
+    },
+    {
+      "epoch": 0.02184141736857392,
+      "grad_norm": 0.33328378200531006,
+      "learning_rate": 0.0001,
+      "loss": 1.7835,
+      "step": 188
+    },
+    {
+      "epoch": 0.02195759512053442,
+      "grad_norm": 0.3383578360080719,
+      "learning_rate": 0.0001,
+      "loss": 1.7839,
+      "step": 189
+    },
+    {
+      "epoch": 0.02207377287249492,
+      "grad_norm": 0.3100789785385132,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 190
+    },
+    {
+      "epoch": 0.022189950624455418,
+      "grad_norm": 0.3665405511856079,
+      "learning_rate": 0.0001,
+      "loss": 1.8639,
+      "step": 191
+    },
+    {
+      "epoch": 0.022306128376415917,
+      "grad_norm": 0.36163437366485596,
+      "learning_rate": 0.0001,
+      "loss": 1.8558,
+      "step": 192
+    },
+    {
+      "epoch": 0.022422306128376417,
+      "grad_norm": 0.367439866065979,
+      "learning_rate": 0.0001,
+      "loss": 1.9491,
+      "step": 193
+    },
+    {
+      "epoch": 0.022538483880336916,
+      "grad_norm": 0.3353922963142395,
+      "learning_rate": 0.0001,
+      "loss": 1.8423,
+      "step": 194
+    },
+    {
+      "epoch": 0.022654661632297415,
+      "grad_norm": 0.33214300870895386,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 195
+    },
+    {
+      "epoch": 0.022770839384257915,
+      "grad_norm": 0.33071082830429077,
+      "learning_rate": 0.0001,
+      "loss": 1.8457,
+      "step": 196
+    },
+    {
+      "epoch": 0.022887017136218414,
+      "grad_norm": 0.3185707926750183,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 197
+    },
+    {
+      "epoch": 0.023003194888178913,
+      "grad_norm": 0.32431602478027344,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 198
+    },
+    {
+      "epoch": 0.023119372640139413,
+      "grad_norm": 0.3300994634628296,
+      "learning_rate": 0.0001,
+      "loss": 1.7727,
+      "step": 199
+    },
+    {
+      "epoch": 0.023235550392099912,
+      "grad_norm": 0.3367205858230591,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 200
+    },
+    {
+      "epoch": 0.02335172814406041,
+      "grad_norm": 0.3295809328556061,
+      "learning_rate": 0.0001,
+      "loss": 1.703,
+      "step": 201
+    },
+    {
+      "epoch": 0.02346790589602091,
+      "grad_norm": 0.3326011598110199,
+      "learning_rate": 0.0001,
+      "loss": 1.7791,
+      "step": 202
+    },
+    {
+      "epoch": 0.02358408364798141,
+      "grad_norm": 0.3159317374229431,
+      "learning_rate": 0.0001,
+      "loss": 1.7935,
+      "step": 203
+    },
+    {
+      "epoch": 0.02370026139994191,
+      "grad_norm": 0.35444700717926025,
+      "learning_rate": 0.0001,
+      "loss": 1.8029,
+      "step": 204
+    },
+    {
+      "epoch": 0.02381643915190241,
+      "grad_norm": 0.35230258107185364,
+      "learning_rate": 0.0001,
+      "loss": 1.7454,
+      "step": 205
+    },
+    {
+      "epoch": 0.023932616903862912,
+      "grad_norm": 0.3529949486255646,
+      "learning_rate": 0.0001,
+      "loss": 1.8424,
+      "step": 206
+    },
+    {
+      "epoch": 0.02404879465582341,
+      "grad_norm": 0.3865301012992859,
+      "learning_rate": 0.0001,
+      "loss": 1.8076,
+      "step": 207
+    },
+    {
+      "epoch": 0.02416497240778391,
+      "grad_norm": 0.32935428619384766,
+      "learning_rate": 0.0001,
+      "loss": 1.7917,
+      "step": 208
+    },
+    {
+      "epoch": 0.02428115015974441,
+      "grad_norm": 0.339444100856781,
+      "learning_rate": 0.0001,
+      "loss": 1.8285,
+      "step": 209
+    },
+    {
+      "epoch": 0.02439732791170491,
+      "grad_norm": 0.3153437077999115,
+      "learning_rate": 0.0001,
+      "loss": 1.5033,
+      "step": 210
+    },
+    {
+      "epoch": 0.02451350566366541,
+      "grad_norm": 0.3516276478767395,
+      "learning_rate": 0.0001,
+      "loss": 1.9109,
+      "step": 211
+    },
+    {
+      "epoch": 0.024629683415625908,
+      "grad_norm": 0.3484845757484436,
+      "learning_rate": 0.0001,
+      "loss": 1.6998,
+      "step": 212
+    },
+    {
+      "epoch": 0.024745861167586407,
+      "grad_norm": 0.3327982723712921,
+      "learning_rate": 0.0001,
+      "loss": 1.792,
+      "step": 213
+    },
+    {
+      "epoch": 0.024862038919546907,
+      "grad_norm": 0.37349483370780945,
+      "learning_rate": 0.0001,
+      "loss": 1.8074,
+      "step": 214
+    },
+    {
+      "epoch": 0.024978216671507406,
+      "grad_norm": 0.3740026354789734,
+      "learning_rate": 0.0001,
+      "loss": 1.7827,
+      "step": 215
+    },
+    {
+      "epoch": 0.025094394423467906,
+      "grad_norm": 0.34118950366973877,
+      "learning_rate": 0.0001,
+      "loss": 1.7631,
+      "step": 216
+    },
+    {
+      "epoch": 0.025210572175428405,
+      "grad_norm": 0.3344680368900299,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 217
+    },
+    {
+      "epoch": 0.025326749927388904,
+      "grad_norm": 0.3705542981624603,
+      "learning_rate": 0.0001,
+      "loss": 1.8239,
+      "step": 218
+    },
+    {
+      "epoch": 0.025442927679349404,
+      "grad_norm": 0.3481920063495636,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 219
+    },
+    {
+      "epoch": 0.025559105431309903,
+      "grad_norm": 0.3217613399028778,
+      "learning_rate": 0.0001,
+      "loss": 1.7362,
+      "step": 220
+    },
+    {
+      "epoch": 0.025675283183270402,
+      "grad_norm": 0.3721480667591095,
+      "learning_rate": 0.0001,
+      "loss": 1.8672,
+      "step": 221
+    },
+    {
+      "epoch": 0.025791460935230902,
+      "grad_norm": 0.32509273290634155,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 222
+    },
+    {
+      "epoch": 0.0259076386871914,
+      "grad_norm": 0.33785101771354675,
+      "learning_rate": 0.0001,
+      "loss": 1.7899,
+      "step": 223
+    },
+    {
+      "epoch": 0.026023816439151904,
+      "grad_norm": 0.3636549115180969,
+      "learning_rate": 0.0001,
+      "loss": 1.8676,
+      "step": 224
+    },
+    {
+      "epoch": 0.026139994191112403,
+      "grad_norm": 0.31692421436309814,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 225
+    },
+    {
+      "epoch": 0.026256171943072903,
+      "grad_norm": 0.3383740186691284,
+      "learning_rate": 0.0001,
+      "loss": 1.7732,
+      "step": 226
+    },
+    {
+      "epoch": 0.026372349695033402,
+      "grad_norm": 0.355059951543808,
+      "learning_rate": 0.0001,
+      "loss": 1.677,
+      "step": 227
+    },
+    {
+      "epoch": 0.0264885274469939,
+      "grad_norm": 0.3562302589416504,
+      "learning_rate": 0.0001,
+      "loss": 1.9306,
+      "step": 228
+    },
+    {
+      "epoch": 0.0266047051989544,
+      "grad_norm": 0.3225594162940979,
+      "learning_rate": 0.0001,
+      "loss": 1.7609,
+      "step": 229
+    },
+    {
+      "epoch": 0.0267208829509149,
+      "grad_norm": 0.33251953125,
+      "learning_rate": 0.0001,
+      "loss": 1.8027,
+      "step": 230
+    },
+    {
+      "epoch": 0.0268370607028754,
+      "grad_norm": 0.3734368085861206,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 231
+    },
+    {
+      "epoch": 0.0269532384548359,
+      "grad_norm": 0.33438679575920105,
+      "learning_rate": 0.0001,
+      "loss": 1.7758,
+      "step": 232
+    },
+    {
+      "epoch": 0.0270694162067964,
+      "grad_norm": 0.37446722388267517,
+      "learning_rate": 0.0001,
+      "loss": 1.929,
+      "step": 233
+    },
+    {
+      "epoch": 0.027185593958756898,
+      "grad_norm": 0.3230035603046417,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 234
+    },
+    {
+      "epoch": 0.027301771710717397,
+      "grad_norm": 0.36308127641677856,
+      "learning_rate": 0.0001,
+      "loss": 1.8119,
+      "step": 235
+    },
+    {
+      "epoch": 0.027417949462677896,
+      "grad_norm": 0.31774529814720154,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 236
+    },
+    {
+      "epoch": 0.027534127214638396,
+      "grad_norm": 0.3248523473739624,
+      "learning_rate": 0.0001,
+      "loss": 1.8632,
+      "step": 237
+    },
+    {
+      "epoch": 0.027650304966598895,
+      "grad_norm": 0.3736507296562195,
+      "learning_rate": 0.0001,
+      "loss": 1.9092,
+      "step": 238
+    },
+    {
+      "epoch": 0.027766482718559395,
+      "grad_norm": 0.3262912631034851,
+      "learning_rate": 0.0001,
+      "loss": 1.697,
+      "step": 239
+    },
+    {
+      "epoch": 0.027882660470519894,
+      "grad_norm": 0.3249903619289398,
+      "learning_rate": 0.0001,
+      "loss": 1.7496,
+      "step": 240
+    },
+    {
+      "epoch": 0.027998838222480393,
+      "grad_norm": 0.3449850380420685,
+      "learning_rate": 0.0001,
+      "loss": 1.825,
+      "step": 241
+    },
+    {
+      "epoch": 0.028115015974440896,
+      "grad_norm": 0.41747432947158813,
+      "learning_rate": 0.0001,
+      "loss": 2.1273,
+      "step": 242
+    },
+    {
+      "epoch": 0.028231193726401396,
+      "grad_norm": 0.35569465160369873,
+      "learning_rate": 0.0001,
+      "loss": 1.8643,
+      "step": 243
+    },
+    {
+      "epoch": 0.028347371478361895,
+      "grad_norm": 0.3352862596511841,
+      "learning_rate": 0.0001,
+      "loss": 1.8427,
+      "step": 244
+    },
+    {
+      "epoch": 0.028463549230322394,
+      "grad_norm": 0.38555145263671875,
+      "learning_rate": 0.0001,
+      "loss": 1.9834,
+      "step": 245
+    },
+    {
+      "epoch": 0.028579726982282894,
+      "grad_norm": 0.3578146696090698,
+      "learning_rate": 0.0001,
+      "loss": 1.7827,
+      "step": 246
+    },
+    {
+      "epoch": 0.028695904734243393,
+      "grad_norm": 0.3353215754032135,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 247
+    },
+    {
+      "epoch": 0.028812082486203892,
+      "grad_norm": 0.3413390815258026,
+      "learning_rate": 0.0001,
+      "loss": 1.7664,
+      "step": 248
+    },
+    {
+      "epoch": 0.028928260238164392,
+      "grad_norm": 0.3437090218067169,
+      "learning_rate": 0.0001,
+      "loss": 1.8693,
+      "step": 249
+    },
+    {
+      "epoch": 0.02904443799012489,
+      "grad_norm": 0.3443460762500763,
+      "learning_rate": 0.0001,
+      "loss": 1.8738,
+      "step": 250
+    },
+    {
+      "epoch": 0.02916061574208539,
+      "grad_norm": 0.3518659770488739,
+      "learning_rate": 0.0001,
+      "loss": 1.8869,
+      "step": 251
+    },
+    {
+      "epoch": 0.02927679349404589,
+      "grad_norm": 0.3433094322681427,
+      "learning_rate": 0.0001,
+      "loss": 1.8771,
+      "step": 252
+    },
+    {
+      "epoch": 0.02939297124600639,
+      "grad_norm": 0.3407760262489319,
+      "learning_rate": 0.0001,
+      "loss": 1.914,
+      "step": 253
+    },
+    {
+      "epoch": 0.02950914899796689,
+      "grad_norm": 0.3268572688102722,
+      "learning_rate": 0.0001,
+      "loss": 1.8681,
+      "step": 254
+    },
+    {
+      "epoch": 0.029625326749927388,
+      "grad_norm": 0.3242436349391937,
+      "learning_rate": 0.0001,
+      "loss": 1.8188,
+      "step": 255
+    },
+    {
+      "epoch": 0.029741504501887887,
+      "grad_norm": 0.33246999979019165,
+      "learning_rate": 0.0001,
+      "loss": 1.7859,
+      "step": 256
+    },
+    {
+      "epoch": 0.029857682253848387,
+      "grad_norm": 0.3286731541156769,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 257
+    },
+    {
+      "epoch": 0.029973860005808886,
+      "grad_norm": 0.3115208148956299,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 258
+    },
+    {
+      "epoch": 0.030090037757769385,
+      "grad_norm": 0.3387967646121979,
+      "learning_rate": 0.0001,
+      "loss": 1.8529,
+      "step": 259
+    },
+    {
+      "epoch": 0.03020621550972989,
+      "grad_norm": 0.35752734541893005,
+      "learning_rate": 0.0001,
+      "loss": 1.8307,
+      "step": 260
+    },
+    {
+      "epoch": 0.030322393261690388,
+      "grad_norm": 0.3699615001678467,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 261
+    },
+    {
+      "epoch": 0.030438571013650887,
+      "grad_norm": 0.328961044549942,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 262
+    },
+    {
+      "epoch": 0.030554748765611386,
+      "grad_norm": 0.36490383744239807,
+      "learning_rate": 0.0001,
+      "loss": 1.8146,
+      "step": 263
+    },
+    {
+      "epoch": 0.030670926517571886,
+      "grad_norm": 0.32661300897598267,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 264
+    },
+    {
+      "epoch": 0.030787104269532385,
+      "grad_norm": 0.34228524565696716,
+      "learning_rate": 0.0001,
+      "loss": 1.8087,
+      "step": 265
+    },
+    {
+      "epoch": 0.030903282021492885,
+      "grad_norm": 0.35033437609672546,
+      "learning_rate": 0.0001,
+      "loss": 1.8297,
+      "step": 266
+    },
+    {
+      "epoch": 0.031019459773453384,
+      "grad_norm": 0.3491237759590149,
+      "learning_rate": 0.0001,
+      "loss": 1.945,
+      "step": 267
+    },
+    {
+      "epoch": 0.031135637525413883,
+      "grad_norm": 0.34657391905784607,
+      "learning_rate": 0.0001,
+      "loss": 1.823,
+      "step": 268
+    },
+    {
+      "epoch": 0.03125181527737438,
+      "grad_norm": 0.36879056692123413,
+      "learning_rate": 0.0001,
+      "loss": 1.859,
+      "step": 269
+    },
+    {
+      "epoch": 0.031367993029334885,
+      "grad_norm": 0.36880430579185486,
+      "learning_rate": 0.0001,
+      "loss": 1.7599,
+      "step": 270
+    },
+    {
+      "epoch": 0.03148417078129538,
+      "grad_norm": 0.3530198931694031,
+      "learning_rate": 0.0001,
+      "loss": 1.9935,
+      "step": 271
+    },
+    {
+      "epoch": 0.031600348533255884,
+      "grad_norm": 0.35175904631614685,
+      "learning_rate": 0.0001,
+      "loss": 1.8837,
+      "step": 272
+    },
+    {
+      "epoch": 0.03171652628521638,
+      "grad_norm": 0.3176674246788025,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 273
+    },
+    {
+      "epoch": 0.03183270403717688,
+      "grad_norm": 0.33333176374435425,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 274
+    },
+    {
+      "epoch": 0.03194888178913738,
+      "grad_norm": 0.31809431314468384,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 275
+    },
+    {
+      "epoch": 0.03206505954109788,
+      "grad_norm": 0.3419245183467865,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 276
+    },
+    {
+      "epoch": 0.03218123729305838,
+      "grad_norm": 0.3137516379356384,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 277
+    },
+    {
+      "epoch": 0.03229741504501888,
+      "grad_norm": 0.3313903510570526,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 278
+    },
+    {
+      "epoch": 0.032413592796979376,
+      "grad_norm": 0.3197391629219055,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 279
+    },
+    {
+      "epoch": 0.03252977054893988,
+      "grad_norm": 0.3287334442138672,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 280
+    },
+    {
+      "epoch": 0.032645948300900375,
+      "grad_norm": 0.3322586715221405,
+      "learning_rate": 0.0001,
+      "loss": 1.8868,
+      "step": 281
+    },
+    {
+      "epoch": 0.03276212605286088,
+      "grad_norm": 0.35943326354026794,
+      "learning_rate": 0.0001,
+      "loss": 1.8031,
+      "step": 282
+    },
+    {
+      "epoch": 0.032878303804821374,
+      "grad_norm": 0.3296755254268646,
+      "learning_rate": 0.0001,
+      "loss": 1.7838,
+      "step": 283
+    },
+    {
+      "epoch": 0.03299448155678188,
+      "grad_norm": 0.3416058123111725,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 284
+    },
+    {
+      "epoch": 0.03311065930874237,
+      "grad_norm": 0.3588326573371887,
+      "learning_rate": 0.0001,
+      "loss": 1.8722,
+      "step": 285
+    },
+    {
+      "epoch": 0.033226837060702875,
+      "grad_norm": 0.3292895555496216,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 286
+    },
+    {
+      "epoch": 0.03334301481266338,
+      "grad_norm": 0.3382332921028137,
+      "learning_rate": 0.0001,
+      "loss": 1.8179,
+      "step": 287
+    },
+    {
+      "epoch": 0.033459192564623874,
+      "grad_norm": 0.33576157689094543,
+      "learning_rate": 0.0001,
+      "loss": 1.8442,
+      "step": 288
+    },
+    {
+      "epoch": 0.03357537031658438,
+      "grad_norm": 0.3294540047645569,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 289
+    },
+    {
+      "epoch": 0.03369154806854487,
+      "grad_norm": 0.3160039782524109,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 290
+    },
+    {
+      "epoch": 0.033807725820505376,
+      "grad_norm": 0.3588349223136902,
+      "learning_rate": 0.0001,
+      "loss": 1.8304,
+      "step": 291
+    },
+    {
+      "epoch": 0.03392390357246587,
+      "grad_norm": 0.30049145221710205,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 292
+    },
+    {
+      "epoch": 0.034040081324426374,
+      "grad_norm": 0.3043253719806671,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 293
+    },
+    {
+      "epoch": 0.03415625907638687,
+      "grad_norm": 0.377269983291626,
+      "learning_rate": 0.0001,
+      "loss": 1.8306,
+      "step": 294
+    },
+    {
+      "epoch": 0.03427243682834737,
+      "grad_norm": 0.378327876329422,
+      "learning_rate": 0.0001,
+      "loss": 1.9481,
+      "step": 295
+    },
+    {
+      "epoch": 0.03438861458030787,
+      "grad_norm": 0.3419339060783386,
+      "learning_rate": 0.0001,
+      "loss": 1.8375,
+      "step": 296
+    },
+    {
+      "epoch": 0.03450479233226837,
+      "grad_norm": 0.3224831521511078,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 297
+    },
+    {
+      "epoch": 0.03462097008422887,
+      "grad_norm": 0.35075846314430237,
+      "learning_rate": 0.0001,
+      "loss": 1.7834,
+      "step": 298
+    },
+    {
+      "epoch": 0.03473714783618937,
+      "grad_norm": 0.345042884349823,
+      "learning_rate": 0.0001,
+      "loss": 1.8939,
+      "step": 299
+    },
+    {
+      "epoch": 0.03485332558814987,
+      "grad_norm": 0.3130955100059509,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 300
+    },
+    {
+      "epoch": 0.03496950334011037,
+      "grad_norm": 0.35810473561286926,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 301
+    },
+    {
+      "epoch": 0.035085681092070865,
+      "grad_norm": 0.33703386783599854,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 302
+    },
+    {
+      "epoch": 0.03520185884403137,
+      "grad_norm": 0.3259965777397156,
+      "learning_rate": 0.0001,
+      "loss": 1.7835,
+      "step": 303
+    },
+    {
+      "epoch": 0.03531803659599187,
+      "grad_norm": 0.3411068618297577,
+      "learning_rate": 0.0001,
+      "loss": 1.7623,
+      "step": 304
+    },
+    {
+      "epoch": 0.03543421434795237,
+      "grad_norm": 0.35380053520202637,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 305
+    },
+    {
+      "epoch": 0.03555039209991287,
+      "grad_norm": 0.32706257700920105,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 306
+    },
+    {
+      "epoch": 0.035666569851873366,
+      "grad_norm": 0.3144983649253845,
+      "learning_rate": 0.0001,
+      "loss": 1.6658,
+      "step": 307
+    },
+    {
+      "epoch": 0.03578274760383387,
+      "grad_norm": 0.3492032289505005,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 308
+    },
+    {
+      "epoch": 0.035898925355794364,
+      "grad_norm": 0.34730949997901917,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 309
+    },
+    {
+      "epoch": 0.03601510310775487,
+      "grad_norm": 0.31954050064086914,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 310
+    },
+    {
+      "epoch": 0.03613128085971536,
+      "grad_norm": 0.3348424732685089,
+      "learning_rate": 0.0001,
+      "loss": 1.7981,
+      "step": 311
+    },
+    {
+      "epoch": 0.036247458611675866,
+      "grad_norm": 0.3164363503456116,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 312
+    },
+    {
+      "epoch": 0.03636363636363636,
+      "grad_norm": 0.35313692688941956,
+      "learning_rate": 0.0001,
+      "loss": 1.9214,
+      "step": 313
+    },
+    {
+      "epoch": 0.036479814115596865,
+      "grad_norm": 0.3127838671207428,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 314
+    },
+    {
+      "epoch": 0.03659599186755736,
+      "grad_norm": 0.33738380670547485,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 315
+    },
+    {
+      "epoch": 0.03671216961951786,
+      "grad_norm": 0.34780019521713257,
+      "learning_rate": 0.0001,
+      "loss": 1.863,
+      "step": 316
+    },
+    {
+      "epoch": 0.03682834737147836,
+      "grad_norm": 0.3482603430747986,
+      "learning_rate": 0.0001,
+      "loss": 1.9564,
+      "step": 317
+    },
+    {
+      "epoch": 0.03694452512343886,
+      "grad_norm": 0.3411928415298462,
+      "learning_rate": 0.0001,
+      "loss": 1.8199,
+      "step": 318
+    },
+    {
+      "epoch": 0.03706070287539936,
+      "grad_norm": 0.33409395813941956,
+      "learning_rate": 0.0001,
+      "loss": 1.8027,
+      "step": 319
+    },
+    {
+      "epoch": 0.03717688062735986,
+      "grad_norm": 0.3127480447292328,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 320
+    },
+    {
+      "epoch": 0.03729305837932036,
+      "grad_norm": 0.3322620689868927,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 321
+    },
+    {
+      "epoch": 0.03740923613128086,
+      "grad_norm": 0.3270343542098999,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 322
+    },
+    {
+      "epoch": 0.03752541388324136,
+      "grad_norm": 0.3130471110343933,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 323
+    },
+    {
+      "epoch": 0.03764159163520186,
+      "grad_norm": 0.33622002601623535,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 324
+    },
+    {
+      "epoch": 0.03775776938716236,
+      "grad_norm": 0.3530596196651459,
+      "learning_rate": 0.0001,
+      "loss": 1.9313,
+      "step": 325
+    },
+    {
+      "epoch": 0.03787394713912286,
+      "grad_norm": 0.38460564613342285,
+      "learning_rate": 0.0001,
+      "loss": 2.0007,
+      "step": 326
+    },
+    {
+      "epoch": 0.03799012489108336,
+      "grad_norm": 0.3392890393733978,
+      "learning_rate": 0.0001,
+      "loss": 1.8334,
+      "step": 327
+    },
+    {
+      "epoch": 0.038106302643043856,
+      "grad_norm": 0.33205297589302063,
+      "learning_rate": 0.0001,
+      "loss": 1.8434,
+      "step": 328
+    },
+    {
+      "epoch": 0.03822248039500436,
+      "grad_norm": 0.34488892555236816,
+      "learning_rate": 0.0001,
+      "loss": 1.7568,
+      "step": 329
+    },
+    {
+      "epoch": 0.038338658146964855,
+      "grad_norm": 0.357216477394104,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 330
+    },
+    {
+      "epoch": 0.03845483589892536,
+      "grad_norm": 0.31771931052207947,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 331
+    },
+    {
+      "epoch": 0.03857101365088585,
+      "grad_norm": 0.36487412452697754,
+      "learning_rate": 0.0001,
+      "loss": 1.8129,
+      "step": 332
+    },
+    {
+      "epoch": 0.038687191402846356,
+      "grad_norm": 0.3407606780529022,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 333
+    },
+    {
+      "epoch": 0.03880336915480685,
+      "grad_norm": 0.31532132625579834,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 334
+    },
+    {
+      "epoch": 0.038919546906767355,
+      "grad_norm": 0.3287792503833771,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 335
+    },
+    {
+      "epoch": 0.03903572465872785,
+      "grad_norm": 0.33964601159095764,
+      "learning_rate": 0.0001,
+      "loss": 1.8163,
+      "step": 336
+    },
+    {
+      "epoch": 0.039151902410688354,
+      "grad_norm": 0.34165436029434204,
+      "learning_rate": 0.0001,
+      "loss": 1.8581,
+      "step": 337
+    },
+    {
+      "epoch": 0.03926808016264885,
+      "grad_norm": 0.348545640707016,
+      "learning_rate": 0.0001,
+      "loss": 1.7487,
+      "step": 338
+    },
+    {
+      "epoch": 0.03938425791460935,
+      "grad_norm": 0.322898268699646,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 339
+    },
+    {
+      "epoch": 0.039500435666569855,
+      "grad_norm": 0.3299243748188019,
+      "learning_rate": 0.0001,
+      "loss": 1.7179,
+      "step": 340
+    },
+    {
+      "epoch": 0.03961661341853035,
+      "grad_norm": 0.34560680389404297,
+      "learning_rate": 0.0001,
+      "loss": 1.8015,
+      "step": 341
+    },
+    {
+      "epoch": 0.039732791170490854,
+      "grad_norm": 0.34533241391181946,
+      "learning_rate": 0.0001,
+      "loss": 1.8813,
+      "step": 342
+    },
+    {
+      "epoch": 0.03984896892245135,
+      "grad_norm": 0.3315548598766327,
+      "learning_rate": 0.0001,
+      "loss": 1.849,
+      "step": 343
+    },
+    {
+      "epoch": 0.03996514667441185,
+      "grad_norm": 0.31113383173942566,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 344
+    },
+    {
+      "epoch": 0.04008132442637235,
+      "grad_norm": 0.34004050493240356,
+      "learning_rate": 0.0001,
+      "loss": 1.7723,
+      "step": 345
+    },
+    {
+      "epoch": 0.04019750217833285,
+      "grad_norm": 0.3573724925518036,
+      "learning_rate": 0.0001,
+      "loss": 1.8572,
+      "step": 346
+    },
+    {
+      "epoch": 0.04031367993029335,
+      "grad_norm": 0.3385657072067261,
+      "learning_rate": 0.0001,
+      "loss": 1.7458,
+      "step": 347
+    },
+    {
+      "epoch": 0.04042985768225385,
+      "grad_norm": 0.34114423394203186,
+      "learning_rate": 0.0001,
+      "loss": 1.7496,
+      "step": 348
+    },
+    {
+      "epoch": 0.040546035434214346,
+      "grad_norm": 0.3590448796749115,
+      "learning_rate": 0.0001,
+      "loss": 1.8671,
+      "step": 349
+    },
+    {
+      "epoch": 0.04066221318617485,
+      "grad_norm": 0.34044602513313293,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 350
+    },
+    {
+      "epoch": 0.040778390938135345,
+      "grad_norm": 0.34901726245880127,
+      "learning_rate": 0.0001,
+      "loss": 1.7836,
+      "step": 351
+    },
+    {
+      "epoch": 0.04089456869009585,
+      "grad_norm": 0.33901697397232056,
+      "learning_rate": 0.0001,
+      "loss": 1.9265,
+      "step": 352
+    },
+    {
+      "epoch": 0.041010746442056344,
+      "grad_norm": 0.33621305227279663,
+      "learning_rate": 0.0001,
+      "loss": 1.7498,
+      "step": 353
+    },
+    {
+      "epoch": 0.041126924194016846,
+      "grad_norm": 0.3610171973705292,
+      "learning_rate": 0.0001,
+      "loss": 1.9162,
+      "step": 354
+    },
+    {
+      "epoch": 0.04124310194597734,
+      "grad_norm": 0.33839908242225647,
+      "learning_rate": 0.0001,
+      "loss": 1.8152,
+      "step": 355
+    },
+    {
+      "epoch": 0.041359279697937845,
+      "grad_norm": 0.32283854484558105,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 356
+    },
+    {
+      "epoch": 0.04147545744989834,
+      "grad_norm": 0.3764745891094208,
+      "learning_rate": 0.0001,
+      "loss": 1.9003,
+      "step": 357
+    },
+    {
+      "epoch": 0.041591635201858844,
+      "grad_norm": 0.34221282601356506,
+      "learning_rate": 0.0001,
+      "loss": 1.8303,
+      "step": 358
+    },
+    {
+      "epoch": 0.04170781295381935,
+      "grad_norm": 0.36195555329322815,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 359
+    },
+    {
+      "epoch": 0.04182399070577984,
+      "grad_norm": 0.33597198128700256,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 360
+    },
+    {
+      "epoch": 0.041940168457740346,
+      "grad_norm": 0.3600618243217468,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 361
+    },
+    {
+      "epoch": 0.04205634620970084,
+      "grad_norm": 0.3321481943130493,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 362
+    },
+    {
+      "epoch": 0.042172523961661344,
+      "grad_norm": 0.34914496541023254,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 363
+    },
+    {
+      "epoch": 0.04228870171362184,
+      "grad_norm": 0.34095922112464905,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 364
+    },
+    {
+      "epoch": 0.04240487946558234,
+      "grad_norm": 0.35311540961265564,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 365
+    },
+    {
+      "epoch": 0.04252105721754284,
+      "grad_norm": 0.383943647146225,
+      "learning_rate": 0.0001,
+      "loss": 1.8434,
+      "step": 366
+    },
+    {
+      "epoch": 0.04263723496950334,
+      "grad_norm": 0.33830368518829346,
+      "learning_rate": 0.0001,
+      "loss": 1.8743,
+      "step": 367
+    },
+    {
+      "epoch": 0.04275341272146384,
+      "grad_norm": 0.3408838212490082,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 368
+    },
+    {
+      "epoch": 0.04286959047342434,
+      "grad_norm": 0.37213000655174255,
+      "learning_rate": 0.0001,
+      "loss": 1.9295,
+      "step": 369
+    },
+    {
+      "epoch": 0.042985768225384836,
+      "grad_norm": 0.33618566393852234,
+      "learning_rate": 0.0001,
+      "loss": 1.9901,
+      "step": 370
+    },
+    {
+      "epoch": 0.04310194597734534,
+      "grad_norm": 0.3326283097267151,
+      "learning_rate": 0.0001,
+      "loss": 1.7401,
+      "step": 371
+    },
+    {
+      "epoch": 0.043218123729305835,
+      "grad_norm": 0.3324650526046753,
+      "learning_rate": 0.0001,
+      "loss": 1.7636,
+      "step": 372
+    },
+    {
+      "epoch": 0.04333430148126634,
+      "grad_norm": 0.36444899439811707,
+      "learning_rate": 0.0001,
+      "loss": 1.7127,
+      "step": 373
+    },
+    {
+      "epoch": 0.043450479233226834,
+      "grad_norm": 0.32787513732910156,
+      "learning_rate": 0.0001,
+      "loss": 1.7763,
+      "step": 374
+    },
+    {
+      "epoch": 0.04356665698518734,
+      "grad_norm": 0.3590793311595917,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 375
+    },
+    {
+      "epoch": 0.04368283473714784,
+      "grad_norm": 0.3318205773830414,
+      "learning_rate": 0.0001,
+      "loss": 1.8108,
+      "step": 376
+    },
+    {
+      "epoch": 0.043799012489108335,
+      "grad_norm": 0.34203052520751953,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 377
+    },
+    {
+      "epoch": 0.04391519024106884,
+      "grad_norm": 0.3480004072189331,
+      "learning_rate": 0.0001,
+      "loss": 1.8196,
+      "step": 378
+    },
+    {
+      "epoch": 0.044031367993029334,
+      "grad_norm": 0.3592599928379059,
+      "learning_rate": 0.0001,
+      "loss": 1.768,
+      "step": 379
+    },
+    {
+      "epoch": 0.04414754574498984,
+      "grad_norm": 0.3415685296058655,
+      "learning_rate": 0.0001,
+      "loss": 1.8739,
+      "step": 380
+    },
+    {
+      "epoch": 0.04426372349695033,
+      "grad_norm": 0.33502671122550964,
+      "learning_rate": 0.0001,
+      "loss": 1.7937,
+      "step": 381
+    },
+    {
+      "epoch": 0.044379901248910836,
+      "grad_norm": 0.3669358193874359,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 382
+    },
+    {
+      "epoch": 0.04449607900087133,
+      "grad_norm": 0.35224342346191406,
+      "learning_rate": 0.0001,
+      "loss": 1.787,
+      "step": 383
+    },
+    {
+      "epoch": 0.044612256752831835,
+      "grad_norm": 0.3165675699710846,
+      "learning_rate": 0.0001,
+      "loss": 1.7865,
+      "step": 384
+    },
+    {
+      "epoch": 0.04472843450479233,
+      "grad_norm": 0.34537431597709656,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 385
+    },
+    {
+      "epoch": 0.04484461225675283,
+      "grad_norm": 0.3536418080329895,
+      "learning_rate": 0.0001,
+      "loss": 1.8404,
+      "step": 386
+    },
+    {
+      "epoch": 0.04496079000871333,
+      "grad_norm": 0.33434730768203735,
+      "learning_rate": 0.0001,
+      "loss": 1.8322,
+      "step": 387
+    },
+    {
+      "epoch": 0.04507696776067383,
+      "grad_norm": 0.32901090383529663,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 388
+    },
+    {
+      "epoch": 0.04519314551263433,
+      "grad_norm": 0.3278455436229706,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 389
+    },
+    {
+      "epoch": 0.04530932326459483,
+      "grad_norm": 0.34721484780311584,
+      "learning_rate": 0.0001,
+      "loss": 1.8898,
+      "step": 390
+    },
+    {
+      "epoch": 0.04542550101655533,
+      "grad_norm": 0.3322944939136505,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 391
+    },
+    {
+      "epoch": 0.04554167876851583,
+      "grad_norm": 0.34092095494270325,
+      "learning_rate": 0.0001,
+      "loss": 1.8761,
+      "step": 392
+    },
+    {
+      "epoch": 0.045657856520476325,
+      "grad_norm": 0.3071598708629608,
+      "learning_rate": 0.0001,
+      "loss": 1.7564,
+      "step": 393
+    },
+    {
+      "epoch": 0.04577403427243683,
+      "grad_norm": 0.3361007273197174,
+      "learning_rate": 0.0001,
+      "loss": 1.7425,
+      "step": 394
+    },
+    {
+      "epoch": 0.04589021202439733,
+      "grad_norm": 0.38204225897789,
+      "learning_rate": 0.0001,
+      "loss": 1.9099,
+      "step": 395
+    },
+    {
+      "epoch": 0.04600638977635783,
+      "grad_norm": 0.32257604598999023,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 396
+    },
+    {
+      "epoch": 0.04612256752831833,
+      "grad_norm": 0.3314444124698639,
+      "learning_rate": 0.0001,
+      "loss": 1.7518,
+      "step": 397
+    },
+    {
+      "epoch": 0.046238745280278826,
+      "grad_norm": 0.34747904539108276,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 398
+    },
+    {
+      "epoch": 0.04635492303223933,
+      "grad_norm": 0.3240096867084503,
+      "learning_rate": 0.0001,
+      "loss": 1.7374,
+      "step": 399
+    },
+    {
+      "epoch": 0.046471100784199824,
+      "grad_norm": 0.3071964979171753,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 400
+    },
+    {
+      "epoch": 0.04658727853616033,
+      "grad_norm": 0.3766336739063263,
+      "learning_rate": 0.0001,
+      "loss": 1.9295,
+      "step": 401
+    },
+    {
+      "epoch": 0.04670345628812082,
+      "grad_norm": 0.35144826769828796,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 402
+    },
+    {
+      "epoch": 0.046819634040081326,
+      "grad_norm": 0.31121134757995605,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 403
+    },
+    {
+      "epoch": 0.04693581179204182,
+      "grad_norm": 0.3237099051475525,
+      "learning_rate": 0.0001,
+      "loss": 1.7629,
+      "step": 404
+    },
+    {
+      "epoch": 0.047051989544002325,
+      "grad_norm": 0.3905102610588074,
+      "learning_rate": 0.0001,
+      "loss": 1.9238,
+      "step": 405
+    },
+    {
+      "epoch": 0.04716816729596282,
+      "grad_norm": 0.31731948256492615,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 406
+    },
+    {
+      "epoch": 0.047284345047923323,
+      "grad_norm": 0.33582621812820435,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 407
+    },
+    {
+      "epoch": 0.04740052279988382,
+      "grad_norm": 0.3342641592025757,
+      "learning_rate": 0.0001,
+      "loss": 1.7827,
+      "step": 408
+    },
+    {
+      "epoch": 0.04751670055184432,
+      "grad_norm": 0.3307628035545349,
+      "learning_rate": 0.0001,
+      "loss": 1.8255,
+      "step": 409
+    },
+    {
+      "epoch": 0.04763287830380482,
+      "grad_norm": 0.3463532030582428,
+      "learning_rate": 0.0001,
+      "loss": 1.7545,
+      "step": 410
+    },
+    {
+      "epoch": 0.04774905605576532,
+      "grad_norm": 0.35025402903556824,
+      "learning_rate": 0.0001,
+      "loss": 1.7729,
+      "step": 411
+    },
+    {
+      "epoch": 0.047865233807725824,
+      "grad_norm": 0.380147784948349,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 412
+    },
+    {
+      "epoch": 0.04798141155968632,
+      "grad_norm": 0.33736342191696167,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 413
+    },
+    {
+      "epoch": 0.04809758931164682,
+      "grad_norm": 0.33364975452423096,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 414
+    },
+    {
+      "epoch": 0.04821376706360732,
+      "grad_norm": 0.3768046498298645,
+      "learning_rate": 0.0001,
+      "loss": 1.8859,
+      "step": 415
+    },
+    {
+      "epoch": 0.04832994481556782,
+      "grad_norm": 0.3302350342273712,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 416
+    },
+    {
+      "epoch": 0.04844612256752832,
+      "grad_norm": 0.3570808172225952,
+      "learning_rate": 0.0001,
+      "loss": 1.8595,
+      "step": 417
+    },
+    {
+      "epoch": 0.04856230031948882,
+      "grad_norm": 0.3274199962615967,
+      "learning_rate": 0.0001,
+      "loss": 1.8074,
+      "step": 418
+    },
+    {
+      "epoch": 0.048678478071449316,
+      "grad_norm": 0.36040613055229187,
+      "learning_rate": 0.0001,
+      "loss": 1.8473,
+      "step": 419
+    },
+    {
+      "epoch": 0.04879465582340982,
+      "grad_norm": 0.3288051187992096,
+      "learning_rate": 0.0001,
+      "loss": 1.7849,
+      "step": 420
+    },
+    {
+      "epoch": 0.048910833575370315,
+      "grad_norm": 0.32802271842956543,
+      "learning_rate": 0.0001,
+      "loss": 1.8159,
+      "step": 421
+    },
+    {
+      "epoch": 0.04902701132733082,
+      "grad_norm": 0.3353809118270874,
+      "learning_rate": 0.0001,
+      "loss": 1.7419,
+      "step": 422
+    },
+    {
+      "epoch": 0.04914318907929131,
+      "grad_norm": 0.325050413608551,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 423
+    },
+    {
+      "epoch": 0.049259366831251816,
+      "grad_norm": 0.34243983030319214,
+      "learning_rate": 0.0001,
+      "loss": 1.7955,
+      "step": 424
+    },
+    {
+      "epoch": 0.04937554458321231,
+      "grad_norm": 0.36178770661354065,
+      "learning_rate": 0.0001,
+      "loss": 1.8266,
+      "step": 425
+    },
+    {
+      "epoch": 0.049491722335172815,
+      "grad_norm": 0.35593631863594055,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 426
+    },
+    {
+      "epoch": 0.04960790008713331,
+      "grad_norm": 0.3078707158565521,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 427
+    },
+    {
+      "epoch": 0.049724077839093814,
+      "grad_norm": 0.3557792007923126,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 428
+    },
+    {
+      "epoch": 0.04984025559105431,
+      "grad_norm": 0.3614198863506317,
+      "learning_rate": 0.0001,
+      "loss": 1.8574,
+      "step": 429
+    },
+    {
+      "epoch": 0.04995643334301481,
+      "grad_norm": 0.3604932427406311,
+      "learning_rate": 0.0001,
+      "loss": 1.8913,
+      "step": 430
+    },
+    {
+      "epoch": 0.050072611094975315,
+      "grad_norm": 0.34103018045425415,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 431
+    },
+    {
+      "epoch": 0.05018878884693581,
+      "grad_norm": 0.34125006198883057,
+      "learning_rate": 0.0001,
+      "loss": 1.8007,
+      "step": 432
+    },
+    {
+      "epoch": 0.050304966598896314,
+      "grad_norm": 0.3314953148365021,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 433
+    },
+    {
+      "epoch": 0.05042114435085681,
+      "grad_norm": 0.3640129566192627,
+      "learning_rate": 0.0001,
+      "loss": 1.816,
+      "step": 434
+    },
+    {
+      "epoch": 0.05053732210281731,
+      "grad_norm": 0.3177947402000427,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 435
+    },
+    {
+      "epoch": 0.05065349985477781,
+      "grad_norm": 0.3107048273086548,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 436
+    },
+    {
+      "epoch": 0.05076967760673831,
+      "grad_norm": 0.3266063332557678,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 437
+    },
+    {
+      "epoch": 0.05088585535869881,
+      "grad_norm": 0.32090339064598083,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 438
+    },
+    {
+      "epoch": 0.05100203311065931,
+      "grad_norm": 0.3660094141960144,
+      "learning_rate": 0.0001,
+      "loss": 1.9414,
+      "step": 439
+    },
+    {
+      "epoch": 0.051118210862619806,
+      "grad_norm": 0.3175399899482727,
+      "learning_rate": 0.0001,
+      "loss": 1.6903,
+      "step": 440
+    },
+    {
+      "epoch": 0.05123438861458031,
+      "grad_norm": 0.3207061290740967,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 441
+    },
+    {
+      "epoch": 0.051350566366540805,
+      "grad_norm": 0.34974828362464905,
+      "learning_rate": 0.0001,
+      "loss": 1.8292,
+      "step": 442
+    },
+    {
+      "epoch": 0.05146674411850131,
+      "grad_norm": 0.40996474027633667,
+      "learning_rate": 0.0001,
+      "loss": 1.8227,
+      "step": 443
+    },
+    {
+      "epoch": 0.051582921870461804,
+      "grad_norm": 0.33745077252388,
+      "learning_rate": 0.0001,
+      "loss": 1.7853,
+      "step": 444
+    },
+    {
+      "epoch": 0.051699099622422306,
+      "grad_norm": 0.3403610289096832,
+      "learning_rate": 0.0001,
+      "loss": 1.8465,
+      "step": 445
+    },
+    {
+      "epoch": 0.0518152773743828,
+      "grad_norm": 0.30970969796180725,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 446
+    },
+    {
+      "epoch": 0.051931455126343305,
+      "grad_norm": 0.3483198583126068,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 447
+    },
+    {
+      "epoch": 0.05204763287830381,
+      "grad_norm": 0.3291124999523163,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 448
+    },
+    {
+      "epoch": 0.052163810630264304,
+      "grad_norm": 0.320839524269104,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 449
+    },
+    {
+      "epoch": 0.05227998838222481,
+      "grad_norm": 0.3379274308681488,
+      "learning_rate": 0.0001,
+      "loss": 1.7447,
+      "step": 450
+    },
+    {
+      "epoch": 0.0523961661341853,
+      "grad_norm": 0.31182000041007996,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 451
+    },
+    {
+      "epoch": 0.052512343886145806,
+      "grad_norm": 0.3308543562889099,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 452
+    },
+    {
+      "epoch": 0.0526285216381063,
+      "grad_norm": 0.34902223944664,
+      "learning_rate": 0.0001,
+      "loss": 1.92,
+      "step": 453
+    },
+    {
+      "epoch": 0.052744699390066804,
+      "grad_norm": 0.3305058181285858,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 454
+    },
+    {
+      "epoch": 0.0528608771420273,
+      "grad_norm": 0.3399059772491455,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 455
+    },
+    {
+      "epoch": 0.0529770548939878,
+      "grad_norm": 0.36334654688835144,
+      "learning_rate": 0.0001,
+      "loss": 1.8425,
+      "step": 456
+    },
+    {
+      "epoch": 0.0530932326459483,
+      "grad_norm": 0.34613272547721863,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 457
+    },
+    {
+      "epoch": 0.0532094103979088,
+      "grad_norm": 0.3351539373397827,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 458
+    },
+    {
+      "epoch": 0.0533255881498693,
+      "grad_norm": 0.34070366621017456,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 459
+    },
+    {
+      "epoch": 0.0534417659018298,
+      "grad_norm": 0.3561154305934906,
+      "learning_rate": 0.0001,
+      "loss": 1.8496,
+      "step": 460
+    },
+    {
+      "epoch": 0.053557943653790296,
+      "grad_norm": 0.32410141825675964,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 461
+    },
+    {
+      "epoch": 0.0536741214057508,
+      "grad_norm": 0.3558378219604492,
+      "learning_rate": 0.0001,
+      "loss": 1.7761,
+      "step": 462
+    },
+    {
+      "epoch": 0.053790299157711295,
+      "grad_norm": 0.3171548843383789,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 463
+    },
+    {
+      "epoch": 0.0539064769096718,
+      "grad_norm": 0.34194016456604004,
+      "learning_rate": 0.0001,
+      "loss": 1.8411,
+      "step": 464
+    },
+    {
+      "epoch": 0.0540226546616323,
+      "grad_norm": 0.3174999952316284,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 465
+    },
+    {
+      "epoch": 0.0541388324135928,
+      "grad_norm": 0.33531078696250916,
+      "learning_rate": 0.0001,
+      "loss": 1.8138,
+      "step": 466
+    },
+    {
+      "epoch": 0.0542550101655533,
+      "grad_norm": 0.33532780408859253,
+      "learning_rate": 0.0001,
+      "loss": 1.7827,
+      "step": 467
+    },
+    {
+      "epoch": 0.054371187917513795,
+      "grad_norm": 0.40104812383651733,
+      "learning_rate": 0.0001,
+      "loss": 2.0425,
+      "step": 468
+    },
+    {
+      "epoch": 0.0544873656694743,
+      "grad_norm": 0.33999940752983093,
+      "learning_rate": 0.0001,
+      "loss": 1.7676,
+      "step": 469
+    },
+    {
+      "epoch": 0.054603543421434794,
+      "grad_norm": 0.3406895399093628,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 470
+    },
+    {
+      "epoch": 0.0547197211733953,
+      "grad_norm": 0.32375043630599976,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 471
+    },
+    {
+      "epoch": 0.05483589892535579,
+      "grad_norm": 0.3496737480163574,
+      "learning_rate": 0.0001,
+      "loss": 1.8868,
+      "step": 472
+    },
+    {
+      "epoch": 0.054952076677316296,
+      "grad_norm": 0.32865509390830994,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 473
+    },
+    {
+      "epoch": 0.05506825442927679,
+      "grad_norm": 0.3332083225250244,
+      "learning_rate": 0.0001,
+      "loss": 1.7202,
+      "step": 474
+    },
+    {
+      "epoch": 0.055184432181237295,
+      "grad_norm": 0.3402409851551056,
+      "learning_rate": 0.0001,
+      "loss": 1.8425,
+      "step": 475
+    },
+    {
+      "epoch": 0.05530060993319779,
+      "grad_norm": 0.37318024039268494,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 476
+    },
+    {
+      "epoch": 0.05541678768515829,
+      "grad_norm": 0.3607703745365143,
+      "learning_rate": 0.0001,
+      "loss": 1.8452,
+      "step": 477
+    },
+    {
+      "epoch": 0.05553296543711879,
+      "grad_norm": 0.3592975437641144,
+      "learning_rate": 0.0001,
+      "loss": 1.8463,
+      "step": 478
+    },
+    {
+      "epoch": 0.05564914318907929,
+      "grad_norm": 0.3610447645187378,
+      "learning_rate": 0.0001,
+      "loss": 1.8906,
+      "step": 479
+    },
+    {
+      "epoch": 0.05576532094103979,
+      "grad_norm": 0.33733242750167847,
+      "learning_rate": 0.0001,
+      "loss": 1.8464,
+      "step": 480
+    },
+    {
+      "epoch": 0.05588149869300029,
+      "grad_norm": 0.35732340812683105,
+      "learning_rate": 0.0001,
+      "loss": 1.8149,
+      "step": 481
+    },
+    {
+      "epoch": 0.05599767644496079,
+      "grad_norm": 0.3425617814064026,
+      "learning_rate": 0.0001,
+      "loss": 1.8263,
+      "step": 482
+    },
+    {
+      "epoch": 0.05611385419692129,
+      "grad_norm": 0.33323052525520325,
+      "learning_rate": 0.0001,
+      "loss": 1.7538,
+      "step": 483
+    },
+    {
+      "epoch": 0.05623003194888179,
+      "grad_norm": 0.35951876640319824,
+      "learning_rate": 0.0001,
+      "loss": 1.8618,
+      "step": 484
+    },
+    {
+      "epoch": 0.05634620970084229,
+      "grad_norm": 0.3292408883571625,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 485
+    },
+    {
+      "epoch": 0.05646238745280279,
+      "grad_norm": 0.37031203508377075,
+      "learning_rate": 0.0001,
+      "loss": 1.7308,
+      "step": 486
+    },
+    {
+      "epoch": 0.05657856520476329,
+      "grad_norm": 0.32847365736961365,
+      "learning_rate": 0.0001,
+      "loss": 1.7937,
+      "step": 487
+    },
+    {
+      "epoch": 0.05669474295672379,
+      "grad_norm": 0.3220025599002838,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 488
+    },
+    {
+      "epoch": 0.056810920708684286,
+      "grad_norm": 0.36746835708618164,
+      "learning_rate": 0.0001,
+      "loss": 1.6696,
+      "step": 489
+    },
+    {
+      "epoch": 0.05692709846064479,
+      "grad_norm": 0.3531992733478546,
+      "learning_rate": 0.0001,
+      "loss": 1.7423,
+      "step": 490
+    },
+    {
+      "epoch": 0.057043276212605284,
+      "grad_norm": 0.3173719346523285,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 491
+    },
+    {
+      "epoch": 0.05715945396456579,
+      "grad_norm": 0.35342299938201904,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 492
+    },
+    {
+      "epoch": 0.05727563171652628,
+      "grad_norm": 0.3307000696659088,
+      "learning_rate": 0.0001,
+      "loss": 1.7597,
+      "step": 493
+    },
+    {
+      "epoch": 0.057391809468486786,
+      "grad_norm": 0.3486880362033844,
+      "learning_rate": 0.0001,
+      "loss": 1.795,
+      "step": 494
+    },
+    {
+      "epoch": 0.05750798722044728,
+      "grad_norm": 0.37966904044151306,
+      "learning_rate": 0.0001,
+      "loss": 1.8754,
+      "step": 495
+    },
+    {
+      "epoch": 0.057624164972407785,
+      "grad_norm": 0.3487970530986786,
+      "learning_rate": 0.0001,
+      "loss": 1.8312,
+      "step": 496
+    },
+    {
+      "epoch": 0.05774034272436828,
+      "grad_norm": 0.30827823281288147,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 497
+    },
+    {
+      "epoch": 0.057856520476328784,
+      "grad_norm": 0.33888596296310425,
+      "learning_rate": 0.0001,
+      "loss": 1.7648,
+      "step": 498
+    },
+    {
+      "epoch": 0.05797269822828928,
+      "grad_norm": 0.35201674699783325,
+      "learning_rate": 0.0001,
+      "loss": 1.7952,
+      "step": 499
+    },
+    {
+      "epoch": 0.05808887598024978,
+      "grad_norm": 0.35336655378341675,
+      "learning_rate": 0.0001,
+      "loss": 1.808,
+      "step": 500
+    },
+    {
+      "epoch": 0.058205053732210285,
+      "grad_norm": 0.36219102144241333,
+      "learning_rate": 0.0001,
+      "loss": 1.7659,
+      "step": 501
+    },
+    {
+      "epoch": 0.05832123148417078,
+      "grad_norm": 0.3310224711894989,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 502
+    },
+    {
+      "epoch": 0.058437409236131284,
+      "grad_norm": 0.372923344373703,
+      "learning_rate": 0.0001,
+      "loss": 1.7554,
+      "step": 503
+    },
+    {
+      "epoch": 0.05855358698809178,
+      "grad_norm": 0.358759343624115,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 504
+    },
+    {
+      "epoch": 0.05866976474005228,
+      "grad_norm": 0.41483983397483826,
+      "learning_rate": 0.0001,
+      "loss": 1.7648,
+      "step": 505
+    },
+    {
+      "epoch": 0.05878594249201278,
+      "grad_norm": 0.3552764058113098,
+      "learning_rate": 0.0001,
+      "loss": 1.8973,
+      "step": 506
+    },
+    {
+      "epoch": 0.05890212024397328,
+      "grad_norm": 0.3468693494796753,
+      "learning_rate": 0.0001,
+      "loss": 1.7336,
+      "step": 507
+    },
+    {
+      "epoch": 0.05901829799593378,
+      "grad_norm": 0.359375536441803,
+      "learning_rate": 0.0001,
+      "loss": 1.8677,
+      "step": 508
+    },
+    {
+      "epoch": 0.05913447574789428,
+      "grad_norm": 0.32384392619132996,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 509
+    },
+    {
+      "epoch": 0.059250653499854776,
+      "grad_norm": 0.3563079237937927,
+      "learning_rate": 0.0001,
+      "loss": 1.8236,
+      "step": 510
+    },
+    {
+      "epoch": 0.05936683125181528,
+      "grad_norm": 0.32716381549835205,
+      "learning_rate": 0.0001,
+      "loss": 1.8558,
+      "step": 511
+    },
+    {
+      "epoch": 0.059483009003775775,
+      "grad_norm": 0.3453499674797058,
+      "learning_rate": 0.0001,
+      "loss": 1.8389,
+      "step": 512
+    },
+    {
+      "epoch": 0.05959918675573628,
+      "grad_norm": 0.32241371273994446,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 513
+    },
+    {
+      "epoch": 0.05971536450769677,
+      "grad_norm": 0.34780648350715637,
+      "learning_rate": 0.0001,
+      "loss": 1.7995,
+      "step": 514
+    },
+    {
+      "epoch": 0.059831542259657276,
+      "grad_norm": 0.3340502977371216,
+      "learning_rate": 0.0001,
+      "loss": 1.8202,
+      "step": 515
+    },
+    {
+      "epoch": 0.05994772001161777,
+      "grad_norm": 0.331016480922699,
+      "learning_rate": 0.0001,
+      "loss": 1.8144,
+      "step": 516
+    },
+    {
+      "epoch": 0.060063897763578275,
+      "grad_norm": 0.34053540229797363,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 517
+    },
+    {
+      "epoch": 0.06018007551553877,
+      "grad_norm": 0.31866925954818726,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 518
+    },
+    {
+      "epoch": 0.060296253267499274,
+      "grad_norm": 0.36000797152519226,
+      "learning_rate": 0.0001,
+      "loss": 1.92,
+      "step": 519
+    },
+    {
+      "epoch": 0.06041243101945978,
+      "grad_norm": 0.37475964426994324,
+      "learning_rate": 0.0001,
+      "loss": 1.8022,
+      "step": 520
+    },
+    {
+      "epoch": 0.06052860877142027,
+      "grad_norm": 0.3554963767528534,
+      "learning_rate": 0.0001,
+      "loss": 1.8536,
+      "step": 521
+    },
+    {
+      "epoch": 0.060644786523380775,
+      "grad_norm": 0.34274694323539734,
+      "learning_rate": 0.0001,
+      "loss": 1.7868,
+      "step": 522
+    },
+    {
+      "epoch": 0.06076096427534127,
+      "grad_norm": 0.31418636441230774,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 523
+    },
+    {
+      "epoch": 0.060877142027301774,
+      "grad_norm": 0.3319328725337982,
+      "learning_rate": 0.0001,
+      "loss": 1.7644,
+      "step": 524
+    },
+    {
+      "epoch": 0.06099331977926227,
+      "grad_norm": 0.34280523657798767,
+      "learning_rate": 0.0001,
+      "loss": 1.759,
+      "step": 525
+    },
+    {
+      "epoch": 0.06110949753122277,
+      "grad_norm": 0.3405478298664093,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 526
+    },
+    {
+      "epoch": 0.06122567528318327,
+      "grad_norm": 0.34357553720474243,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 527
+    },
+    {
+      "epoch": 0.06134185303514377,
+      "grad_norm": 0.3327886164188385,
+      "learning_rate": 0.0001,
+      "loss": 1.825,
+      "step": 528
+    },
+    {
+      "epoch": 0.06145803078710427,
+      "grad_norm": 0.31985750794410706,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 529
+    },
+    {
+      "epoch": 0.06157420853906477,
+      "grad_norm": 0.32863008975982666,
+      "learning_rate": 0.0001,
+      "loss": 1.7012,
+      "step": 530
+    },
+    {
+      "epoch": 0.061690386291025266,
+      "grad_norm": 0.33859503269195557,
+      "learning_rate": 0.0001,
+      "loss": 1.7983,
+      "step": 531
+    },
+    {
+      "epoch": 0.06180656404298577,
+      "grad_norm": 0.354203999042511,
+      "learning_rate": 0.0001,
+      "loss": 1.801,
+      "step": 532
+    },
+    {
+      "epoch": 0.061922741794946265,
+      "grad_norm": 0.3500482439994812,
+      "learning_rate": 0.0001,
+      "loss": 1.7481,
+      "step": 533
+    },
+    {
+      "epoch": 0.06203891954690677,
+      "grad_norm": 0.3380871117115021,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 534
+    },
+    {
+      "epoch": 0.062155097298867264,
+      "grad_norm": 0.3638668656349182,
+      "learning_rate": 0.0001,
+      "loss": 1.8013,
+      "step": 535
+    },
+    {
+      "epoch": 0.06227127505082777,
+      "grad_norm": 0.3231678903102875,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 536
+    },
+    {
+      "epoch": 0.06238745280278827,
+      "grad_norm": 0.3448564112186432,
+      "learning_rate": 0.0001,
+      "loss": 1.7836,
+      "step": 537
+    },
+    {
+      "epoch": 0.06250363055474877,
+      "grad_norm": 0.34526991844177246,
+      "learning_rate": 0.0001,
+      "loss": 1.7609,
+      "step": 538
+    },
+    {
+      "epoch": 0.06261980830670927,
+      "grad_norm": 0.351270467042923,
+      "learning_rate": 0.0001,
+      "loss": 1.8093,
+      "step": 539
+    },
+    {
+      "epoch": 0.06273598605866977,
+      "grad_norm": 0.33383870124816895,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 540
+    },
+    {
+      "epoch": 0.06285216381063026,
+      "grad_norm": 0.31945285201072693,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 541
+    },
+    {
+      "epoch": 0.06296834156259076,
+      "grad_norm": 0.32655853033065796,
+      "learning_rate": 0.0001,
+      "loss": 1.7056,
+      "step": 542
+    },
+    {
+      "epoch": 0.06308451931455127,
+      "grad_norm": 0.3944481313228607,
+      "learning_rate": 0.0001,
+      "loss": 1.8958,
+      "step": 543
+    },
+    {
+      "epoch": 0.06320069706651177,
+      "grad_norm": 0.38392776250839233,
+      "learning_rate": 0.0001,
+      "loss": 1.8377,
+      "step": 544
+    },
+    {
+      "epoch": 0.06331687481847226,
+      "grad_norm": 0.3638809621334076,
+      "learning_rate": 0.0001,
+      "loss": 1.8184,
+      "step": 545
+    },
+    {
+      "epoch": 0.06343305257043276,
+      "grad_norm": 0.3336743116378784,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 546
+    },
+    {
+      "epoch": 0.06354923032239326,
+      "grad_norm": 0.32870176434516907,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 547
+    },
+    {
+      "epoch": 0.06366540807435377,
+      "grad_norm": 0.3251485526561737,
+      "learning_rate": 0.0001,
+      "loss": 1.797,
+      "step": 548
+    },
+    {
+      "epoch": 0.06378158582631425,
+      "grad_norm": 0.36087340116500854,
+      "learning_rate": 0.0001,
+      "loss": 1.8023,
+      "step": 549
+    },
+    {
+      "epoch": 0.06389776357827476,
+      "grad_norm": 0.35436710715293884,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 550
+    },
+    {
+      "epoch": 0.06401394133023526,
+      "grad_norm": 0.3625239133834839,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 551
+    },
+    {
+      "epoch": 0.06413011908219576,
+      "grad_norm": 0.33559513092041016,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 552
+    },
+    {
+      "epoch": 0.06424629683415625,
+      "grad_norm": 0.3335849642753601,
+      "learning_rate": 0.0001,
+      "loss": 1.832,
+      "step": 553
+    },
+    {
+      "epoch": 0.06436247458611676,
+      "grad_norm": 0.3296360969543457,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 554
+    },
+    {
+      "epoch": 0.06447865233807726,
+      "grad_norm": 0.3487899601459503,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 555
+    },
+    {
+      "epoch": 0.06459483009003776,
+      "grad_norm": 0.3731647729873657,
+      "learning_rate": 0.0001,
+      "loss": 2.0035,
+      "step": 556
+    },
+    {
+      "epoch": 0.06471100784199826,
+      "grad_norm": 0.33725181221961975,
+      "learning_rate": 0.0001,
+      "loss": 1.8034,
+      "step": 557
+    },
+    {
+      "epoch": 0.06482718559395875,
+      "grad_norm": 0.3277061879634857,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 558
+    },
+    {
+      "epoch": 0.06494336334591926,
+      "grad_norm": 0.3473774790763855,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 559
+    },
+    {
+      "epoch": 0.06505954109787976,
+      "grad_norm": 0.35049760341644287,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 560
+    },
+    {
+      "epoch": 0.06517571884984026,
+      "grad_norm": 0.3388431668281555,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 561
+    },
+    {
+      "epoch": 0.06529189660180075,
+      "grad_norm": 0.35445770621299744,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 562
+    },
+    {
+      "epoch": 0.06540807435376125,
+      "grad_norm": 0.3392694592475891,
+      "learning_rate": 0.0001,
+      "loss": 1.8481,
+      "step": 563
+    },
+    {
+      "epoch": 0.06552425210572176,
+      "grad_norm": 0.3575725257396698,
+      "learning_rate": 0.0001,
+      "loss": 1.7864,
+      "step": 564
+    },
+    {
+      "epoch": 0.06564042985768226,
+      "grad_norm": 0.3979930877685547,
+      "learning_rate": 0.0001,
+      "loss": 1.8221,
+      "step": 565
+    },
+    {
+      "epoch": 0.06575660760964275,
+      "grad_norm": 0.3467651605606079,
+      "learning_rate": 0.0001,
+      "loss": 1.8548,
+      "step": 566
+    },
+    {
+      "epoch": 0.06587278536160325,
+      "grad_norm": 0.3466176390647888,
+      "learning_rate": 0.0001,
+      "loss": 1.9021,
+      "step": 567
+    },
+    {
+      "epoch": 0.06598896311356375,
+      "grad_norm": 0.33713239431381226,
+      "learning_rate": 0.0001,
+      "loss": 1.835,
+      "step": 568
+    },
+    {
+      "epoch": 0.06610514086552426,
+      "grad_norm": 0.3413597047328949,
+      "learning_rate": 0.0001,
+      "loss": 1.8019,
+      "step": 569
+    },
+    {
+      "epoch": 0.06622131861748475,
+      "grad_norm": 0.40804192423820496,
+      "learning_rate": 0.0001,
+      "loss": 1.9077,
+      "step": 570
+    },
+    {
+      "epoch": 0.06633749636944525,
+      "grad_norm": 0.36405694484710693,
+      "learning_rate": 0.0001,
+      "loss": 1.7789,
+      "step": 571
+    },
+    {
+      "epoch": 0.06645367412140575,
+      "grad_norm": 0.3612009882926941,
+      "learning_rate": 0.0001,
+      "loss": 1.758,
+      "step": 572
+    },
+    {
+      "epoch": 0.06656985187336625,
+      "grad_norm": 0.3526972830295563,
+      "learning_rate": 0.0001,
+      "loss": 1.9189,
+      "step": 573
+    },
+    {
+      "epoch": 0.06668602962532676,
+      "grad_norm": 0.34476232528686523,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 574
+    },
+    {
+      "epoch": 0.06680220737728725,
+      "grad_norm": 0.35720011591911316,
+      "learning_rate": 0.0001,
+      "loss": 1.9326,
+      "step": 575
+    },
+    {
+      "epoch": 0.06691838512924775,
+      "grad_norm": 0.35136792063713074,
+      "learning_rate": 0.0001,
+      "loss": 1.8546,
+      "step": 576
+    },
+    {
+      "epoch": 0.06703456288120825,
+      "grad_norm": 0.32413557171821594,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 577
+    },
+    {
+      "epoch": 0.06715074063316875,
+      "grad_norm": 0.321250319480896,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 578
+    },
+    {
+      "epoch": 0.06726691838512924,
+      "grad_norm": 0.34484270215034485,
+      "learning_rate": 0.0001,
+      "loss": 1.7195,
+      "step": 579
+    },
+    {
+      "epoch": 0.06738309613708975,
+      "grad_norm": 0.34433478116989136,
+      "learning_rate": 0.0001,
+      "loss": 1.7748,
+      "step": 580
+    },
+    {
+      "epoch": 0.06749927388905025,
+      "grad_norm": 0.3812878727912903,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 581
+    },
+    {
+      "epoch": 0.06761545164101075,
+      "grad_norm": 0.344705730676651,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 582
+    },
+    {
+      "epoch": 0.06773162939297124,
+      "grad_norm": 0.3333836793899536,
+      "learning_rate": 0.0001,
+      "loss": 1.6683,
+      "step": 583
+    },
+    {
+      "epoch": 0.06784780714493174,
+      "grad_norm": 0.35086333751678467,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 584
+    },
+    {
+      "epoch": 0.06796398489689225,
+      "grad_norm": 0.3325541019439697,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 585
+    },
+    {
+      "epoch": 0.06808016264885275,
+      "grad_norm": 0.35840341448783875,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 586
+    },
+    {
+      "epoch": 0.06819634040081324,
+      "grad_norm": 0.33432942628860474,
+      "learning_rate": 0.0001,
+      "loss": 1.8359,
+      "step": 587
+    },
+    {
+      "epoch": 0.06831251815277374,
+      "grad_norm": 0.34798067808151245,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 588
+    },
+    {
+      "epoch": 0.06842869590473424,
+      "grad_norm": 0.319845974445343,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 589
+    },
+    {
+      "epoch": 0.06854487365669475,
+      "grad_norm": 0.37829452753067017,
+      "learning_rate": 0.0001,
+      "loss": 1.8517,
+      "step": 590
+    },
+    {
+      "epoch": 0.06866105140865525,
+      "grad_norm": 0.35298144817352295,
+      "learning_rate": 0.0001,
+      "loss": 1.801,
+      "step": 591
+    },
+    {
+      "epoch": 0.06877722916061574,
+      "grad_norm": 0.40371862053871155,
+      "learning_rate": 0.0001,
+      "loss": 1.5935,
+      "step": 592
+    },
+    {
+      "epoch": 0.06889340691257624,
+      "grad_norm": 0.3450127840042114,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 593
+    },
+    {
+      "epoch": 0.06900958466453674,
+      "grad_norm": 0.3465590178966522,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 594
+    },
+    {
+      "epoch": 0.06912576241649725,
+      "grad_norm": 0.3387679159641266,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 595
+    },
+    {
+      "epoch": 0.06924194016845774,
+      "grad_norm": 0.3386651277542114,
+      "learning_rate": 0.0001,
+      "loss": 1.7596,
+      "step": 596
+    },
+    {
+      "epoch": 0.06935811792041824,
+      "grad_norm": 0.3551906645298004,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 597
+    },
+    {
+      "epoch": 0.06947429567237874,
+      "grad_norm": 0.37498459219932556,
+      "learning_rate": 0.0001,
+      "loss": 1.8622,
+      "step": 598
+    },
+    {
+      "epoch": 0.06959047342433924,
+      "grad_norm": 0.3397311866283417,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 599
+    },
+    {
+      "epoch": 0.06970665117629973,
+      "grad_norm": 0.34366878867149353,
+      "learning_rate": 0.0001,
+      "loss": 1.7838,
+      "step": 600
+    },
+    {
+      "epoch": 0.06982282892826024,
+      "grad_norm": 0.3466067910194397,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 601
+    },
+    {
+      "epoch": 0.06993900668022074,
+      "grad_norm": 0.35218632221221924,
+      "learning_rate": 0.0001,
+      "loss": 1.7995,
+      "step": 602
+    },
+    {
+      "epoch": 0.07005518443218124,
+      "grad_norm": 0.3305307924747467,
+      "learning_rate": 0.0001,
+      "loss": 1.7099,
+      "step": 603
+    },
+    {
+      "epoch": 0.07017136218414173,
+      "grad_norm": 0.3424377143383026,
+      "learning_rate": 0.0001,
+      "loss": 1.8505,
+      "step": 604
+    },
+    {
+      "epoch": 0.07028753993610223,
+      "grad_norm": 0.33201947808265686,
+      "learning_rate": 0.0001,
+      "loss": 1.7887,
+      "step": 605
+    },
+    {
+      "epoch": 0.07040371768806274,
+      "grad_norm": 0.3492118716239929,
+      "learning_rate": 0.0001,
+      "loss": 1.8554,
+      "step": 606
+    },
+    {
+      "epoch": 0.07051989544002324,
+      "grad_norm": 0.3389701843261719,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 607
+    },
+    {
+      "epoch": 0.07063607319198374,
+      "grad_norm": 0.3567183315753937,
+      "learning_rate": 0.0001,
+      "loss": 1.8379,
+      "step": 608
+    },
+    {
+      "epoch": 0.07075225094394423,
+      "grad_norm": 0.342669814825058,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 609
+    },
+    {
+      "epoch": 0.07086842869590473,
+      "grad_norm": 0.37096837162971497,
+      "learning_rate": 0.0001,
+      "loss": 1.8548,
+      "step": 610
+    },
+    {
+      "epoch": 0.07098460644786524,
+      "grad_norm": 0.3211749494075775,
+      "learning_rate": 0.0001,
+      "loss": 1.7913,
+      "step": 611
+    },
+    {
+      "epoch": 0.07110078419982574,
+      "grad_norm": 0.3553162217140198,
+      "learning_rate": 0.0001,
+      "loss": 1.8448,
+      "step": 612
+    },
+    {
+      "epoch": 0.07121696195178623,
+      "grad_norm": 0.3596995174884796,
+      "learning_rate": 0.0001,
+      "loss": 1.7092,
+      "step": 613
+    },
+    {
+      "epoch": 0.07133313970374673,
+      "grad_norm": 0.32127490639686584,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 614
+    },
+    {
+      "epoch": 0.07144931745570723,
+      "grad_norm": 0.361713707447052,
+      "learning_rate": 0.0001,
+      "loss": 1.8077,
+      "step": 615
+    },
+    {
+      "epoch": 0.07156549520766774,
+      "grad_norm": 0.3255080282688141,
+      "learning_rate": 0.0001,
+      "loss": 1.7618,
+      "step": 616
+    },
+    {
+      "epoch": 0.07168167295962823,
+      "grad_norm": 0.3420798182487488,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 617
+    },
+    {
+      "epoch": 0.07179785071158873,
+      "grad_norm": 0.39319682121276855,
+      "learning_rate": 0.0001,
+      "loss": 1.797,
+      "step": 618
+    },
+    {
+      "epoch": 0.07191402846354923,
+      "grad_norm": 0.33742591738700867,
+      "learning_rate": 0.0001,
+      "loss": 1.4727,
+      "step": 619
+    },
+    {
+      "epoch": 0.07203020621550973,
+      "grad_norm": 0.3327065408229828,
+      "learning_rate": 0.0001,
+      "loss": 1.7358,
+      "step": 620
+    },
+    {
+      "epoch": 0.07214638396747022,
+      "grad_norm": 0.3351439833641052,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 621
+    },
+    {
+      "epoch": 0.07226256171943073,
+      "grad_norm": 0.3432386815547943,
+      "learning_rate": 0.0001,
+      "loss": 1.7511,
+      "step": 622
+    },
+    {
+      "epoch": 0.07237873947139123,
+      "grad_norm": 0.35164207220077515,
+      "learning_rate": 0.0001,
+      "loss": 1.7889,
+      "step": 623
+    },
+    {
+      "epoch": 0.07249491722335173,
+      "grad_norm": 0.34505143761634827,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 624
+    },
+    {
+      "epoch": 0.07261109497531222,
+      "grad_norm": 0.34696313738822937,
+      "learning_rate": 0.0001,
+      "loss": 1.7991,
+      "step": 625
+    },
+    {
+      "epoch": 0.07272727272727272,
+      "grad_norm": 0.33465588092803955,
+      "learning_rate": 0.0001,
+      "loss": 1.8238,
+      "step": 626
+    },
+    {
+      "epoch": 0.07284345047923323,
+      "grad_norm": 0.3633832335472107,
+      "learning_rate": 0.0001,
+      "loss": 1.808,
+      "step": 627
+    },
+    {
+      "epoch": 0.07295962823119373,
+      "grad_norm": 0.3349045217037201,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 628
+    },
+    {
+      "epoch": 0.07307580598315423,
+      "grad_norm": 0.3547269105911255,
+      "learning_rate": 0.0001,
+      "loss": 1.7475,
+      "step": 629
+    },
+    {
+      "epoch": 0.07319198373511472,
+      "grad_norm": 0.41657668352127075,
+      "learning_rate": 0.0001,
+      "loss": 2.065,
+      "step": 630
+    },
+    {
+      "epoch": 0.07330816148707522,
+      "grad_norm": 0.33551225066185,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 631
+    },
+    {
+      "epoch": 0.07342433923903573,
+      "grad_norm": 0.3287172317504883,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 632
+    },
+    {
+      "epoch": 0.07354051699099623,
+      "grad_norm": 0.3462834060192108,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 633
+    },
+    {
+      "epoch": 0.07365669474295672,
+      "grad_norm": 0.32798144221305847,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 634
+    },
+    {
+      "epoch": 0.07377287249491722,
+      "grad_norm": 0.3188357651233673,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 635
+    },
+    {
+      "epoch": 0.07388905024687772,
+      "grad_norm": 0.34357234835624695,
+      "learning_rate": 0.0001,
+      "loss": 1.7105,
+      "step": 636
+    },
+    {
+      "epoch": 0.07400522799883823,
+      "grad_norm": 0.3531816005706787,
+      "learning_rate": 0.0001,
+      "loss": 1.8203,
+      "step": 637
+    },
+    {
+      "epoch": 0.07412140575079872,
+      "grad_norm": 0.3235273063182831,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 638
+    },
+    {
+      "epoch": 0.07423758350275922,
+      "grad_norm": 0.3293229639530182,
+      "learning_rate": 0.0001,
+      "loss": 1.7596,
+      "step": 639
+    },
+    {
+      "epoch": 0.07435376125471972,
+      "grad_norm": 0.3663218021392822,
+      "learning_rate": 0.0001,
+      "loss": 1.8909,
+      "step": 640
+    },
+    {
+      "epoch": 0.07446993900668022,
+      "grad_norm": 0.33595624566078186,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 641
+    },
+    {
+      "epoch": 0.07458611675864071,
+      "grad_norm": 0.3276788890361786,
+      "learning_rate": 0.0001,
+      "loss": 1.6903,
+      "step": 642
+    },
+    {
+      "epoch": 0.07470229451060122,
+      "grad_norm": 0.3766964077949524,
+      "learning_rate": 0.0001,
+      "loss": 1.9409,
+      "step": 643
+    },
+    {
+      "epoch": 0.07481847226256172,
+      "grad_norm": 0.3381744623184204,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 644
+    },
+    {
+      "epoch": 0.07493465001452222,
+      "grad_norm": 0.34408965706825256,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 645
+    },
+    {
+      "epoch": 0.07505082776648273,
+      "grad_norm": 0.33645275235176086,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 646
+    },
+    {
+      "epoch": 0.07516700551844321,
+      "grad_norm": 0.3625565469264984,
+      "learning_rate": 0.0001,
+      "loss": 1.798,
+      "step": 647
+    },
+    {
+      "epoch": 0.07528318327040372,
+      "grad_norm": 0.3350875973701477,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 648
+    },
+    {
+      "epoch": 0.07539936102236422,
+      "grad_norm": 0.34720659255981445,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 649
+    },
+    {
+      "epoch": 0.07551553877432472,
+      "grad_norm": 0.3373357951641083,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 650
+    },
+    {
+      "epoch": 0.07563171652628521,
+      "grad_norm": 0.3362538516521454,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 651
+    },
+    {
+      "epoch": 0.07574789427824571,
+      "grad_norm": 0.34010645747184753,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 652
+    },
+    {
+      "epoch": 0.07586407203020622,
+      "grad_norm": 0.31828391551971436,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 653
+    },
+    {
+      "epoch": 0.07598024978216672,
+      "grad_norm": 0.32539987564086914,
+      "learning_rate": 0.0001,
+      "loss": 1.7795,
+      "step": 654
+    },
+    {
+      "epoch": 0.07609642753412721,
+      "grad_norm": 0.32553842663764954,
+      "learning_rate": 0.0001,
+      "loss": 1.7163,
+      "step": 655
+    },
+    {
+      "epoch": 0.07621260528608771,
+      "grad_norm": 0.35710620880126953,
+      "learning_rate": 0.0001,
+      "loss": 1.8839,
+      "step": 656
+    },
+    {
+      "epoch": 0.07632878303804821,
+      "grad_norm": 0.33389732241630554,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 657
+    },
+    {
+      "epoch": 0.07644496079000872,
+      "grad_norm": 0.3359222412109375,
+      "learning_rate": 0.0001,
+      "loss": 1.9214,
+      "step": 658
+    },
+    {
+      "epoch": 0.0765611385419692,
+      "grad_norm": 0.3710361123085022,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 659
+    },
+    {
+      "epoch": 0.07667731629392971,
+      "grad_norm": 0.34122204780578613,
+      "learning_rate": 0.0001,
+      "loss": 1.8186,
+      "step": 660
+    },
+    {
+      "epoch": 0.07679349404589021,
+      "grad_norm": 0.3368107080459595,
+      "learning_rate": 0.0001,
+      "loss": 1.8068,
+      "step": 661
+    },
+    {
+      "epoch": 0.07690967179785071,
+      "grad_norm": 0.3254035711288452,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 662
+    },
+    {
+      "epoch": 0.07702584954981122,
+      "grad_norm": 0.3608943223953247,
+      "learning_rate": 0.0001,
+      "loss": 1.9746,
+      "step": 663
+    },
+    {
+      "epoch": 0.0771420273017717,
+      "grad_norm": 0.3355536162853241,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 664
+    },
+    {
+      "epoch": 0.07725820505373221,
+      "grad_norm": 0.3630955219268799,
+      "learning_rate": 0.0001,
+      "loss": 1.7642,
+      "step": 665
+    },
+    {
+      "epoch": 0.07737438280569271,
+      "grad_norm": 0.3425436019897461,
+      "learning_rate": 0.0001,
+      "loss": 1.8083,
+      "step": 666
+    },
+    {
+      "epoch": 0.07749056055765322,
+      "grad_norm": 0.3148491382598877,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 667
+    },
+    {
+      "epoch": 0.0776067383096137,
+      "grad_norm": 0.3225914239883423,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 668
+    },
+    {
+      "epoch": 0.07772291606157421,
+      "grad_norm": 0.37254008650779724,
+      "learning_rate": 0.0001,
+      "loss": 1.8129,
+      "step": 669
+    },
+    {
+      "epoch": 0.07783909381353471,
+      "grad_norm": 0.3412669897079468,
+      "learning_rate": 0.0001,
+      "loss": 1.8073,
+      "step": 670
+    },
+    {
+      "epoch": 0.07795527156549521,
+      "grad_norm": 0.333635151386261,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 671
+    },
+    {
+      "epoch": 0.0780714493174557,
+      "grad_norm": 0.36147254705429077,
+      "learning_rate": 0.0001,
+      "loss": 1.8375,
+      "step": 672
+    },
+    {
+      "epoch": 0.0781876270694162,
+      "grad_norm": 0.3428162932395935,
+      "learning_rate": 0.0001,
+      "loss": 1.7767,
+      "step": 673
+    },
+    {
+      "epoch": 0.07830380482137671,
+      "grad_norm": 0.3516107499599457,
+      "learning_rate": 0.0001,
+      "loss": 1.7804,
+      "step": 674
+    },
+    {
+      "epoch": 0.07841998257333721,
+      "grad_norm": 0.37563571333885193,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 675
+    },
+    {
+      "epoch": 0.0785361603252977,
+      "grad_norm": 0.32837507128715515,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 676
+    },
+    {
+      "epoch": 0.0786523380772582,
+      "grad_norm": 0.34303468465805054,
+      "learning_rate": 0.0001,
+      "loss": 1.8518,
+      "step": 677
+    },
+    {
+      "epoch": 0.0787685158292187,
+      "grad_norm": 0.3297956883907318,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 678
+    },
+    {
+      "epoch": 0.07888469358117921,
+      "grad_norm": 0.34342435002326965,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 679
+    },
+    {
+      "epoch": 0.07900087133313971,
+      "grad_norm": 0.3558964729309082,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 680
+    },
+    {
+      "epoch": 0.0791170490851002,
+      "grad_norm": 0.3384302258491516,
+      "learning_rate": 0.0001,
+      "loss": 1.8352,
+      "step": 681
+    },
+    {
+      "epoch": 0.0792332268370607,
+      "grad_norm": 0.33917075395584106,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 682
+    },
+    {
+      "epoch": 0.0793494045890212,
+      "grad_norm": 0.37113896012306213,
+      "learning_rate": 0.0001,
+      "loss": 1.7594,
+      "step": 683
+    },
+    {
+      "epoch": 0.07946558234098171,
+      "grad_norm": 0.34009361267089844,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 684
+    },
+    {
+      "epoch": 0.0795817600929422,
+      "grad_norm": 0.35961294174194336,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 685
+    },
+    {
+      "epoch": 0.0796979378449027,
+      "grad_norm": 0.3599853813648224,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 686
+    },
+    {
+      "epoch": 0.0798141155968632,
+      "grad_norm": 0.33733391761779785,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 687
+    },
+    {
+      "epoch": 0.0799302933488237,
+      "grad_norm": 0.33123722672462463,
+      "learning_rate": 0.0001,
+      "loss": 1.7556,
+      "step": 688
+    },
+    {
+      "epoch": 0.0800464711007842,
+      "grad_norm": 0.3545984923839569,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 689
+    },
+    {
+      "epoch": 0.0801626488527447,
+      "grad_norm": 0.3520684242248535,
+      "learning_rate": 0.0001,
+      "loss": 1.7908,
+      "step": 690
+    },
+    {
+      "epoch": 0.0802788266047052,
+      "grad_norm": 0.364214152097702,
+      "learning_rate": 0.0001,
+      "loss": 1.8752,
+      "step": 691
+    },
+    {
+      "epoch": 0.0803950043566657,
+      "grad_norm": 0.35055190324783325,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 692
+    },
+    {
+      "epoch": 0.08051118210862619,
+      "grad_norm": 0.3547899127006531,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 693
+    },
+    {
+      "epoch": 0.0806273598605867,
+      "grad_norm": 0.31684044003486633,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 694
+    },
+    {
+      "epoch": 0.0807435376125472,
+      "grad_norm": 0.3523019254207611,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 695
+    },
+    {
+      "epoch": 0.0808597153645077,
+      "grad_norm": 0.379930317401886,
+      "learning_rate": 0.0001,
+      "loss": 1.8274,
+      "step": 696
+    },
+    {
+      "epoch": 0.08097589311646819,
+      "grad_norm": 0.32600757479667664,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 697
+    },
+    {
+      "epoch": 0.08109207086842869,
+      "grad_norm": 0.33486902713775635,
+      "learning_rate": 0.0001,
+      "loss": 1.7029,
+      "step": 698
+    },
+    {
+      "epoch": 0.0812082486203892,
+      "grad_norm": 0.3586275577545166,
+      "learning_rate": 0.0001,
+      "loss": 1.7762,
+      "step": 699
+    },
+    {
+      "epoch": 0.0813244263723497,
+      "grad_norm": 0.34406691789627075,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 700
+    },
+    {
+      "epoch": 0.0814406041243102,
+      "grad_norm": 0.35073554515838623,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 701
+    },
+    {
+      "epoch": 0.08155678187627069,
+      "grad_norm": 0.4370097517967224,
+      "learning_rate": 0.0001,
+      "loss": 1.9901,
+      "step": 702
+    },
+    {
+      "epoch": 0.08167295962823119,
+      "grad_norm": 0.36396339535713196,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 703
+    },
+    {
+      "epoch": 0.0817891373801917,
+      "grad_norm": 0.31740906834602356,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 704
+    },
+    {
+      "epoch": 0.0819053151321522,
+      "grad_norm": 0.36947962641716003,
+      "learning_rate": 0.0001,
+      "loss": 1.8666,
+      "step": 705
+    },
+    {
+      "epoch": 0.08202149288411269,
+      "grad_norm": 0.3637601435184479,
+      "learning_rate": 0.0001,
+      "loss": 1.8011,
+      "step": 706
+    },
+    {
+      "epoch": 0.08213767063607319,
+      "grad_norm": 0.35673728585243225,
+      "learning_rate": 0.0001,
+      "loss": 1.8199,
+      "step": 707
+    },
+    {
+      "epoch": 0.08225384838803369,
+      "grad_norm": 0.3384333550930023,
+      "learning_rate": 0.0001,
+      "loss": 1.7729,
+      "step": 708
+    },
+    {
+      "epoch": 0.0823700261399942,
+      "grad_norm": 0.3344199061393738,
+      "learning_rate": 0.0001,
+      "loss": 1.797,
+      "step": 709
+    },
+    {
+      "epoch": 0.08248620389195468,
+      "grad_norm": 0.3451163172721863,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 710
+    },
+    {
+      "epoch": 0.08260238164391519,
+      "grad_norm": 0.3421926200389862,
+      "learning_rate": 0.0001,
+      "loss": 1.8762,
+      "step": 711
+    },
+    {
+      "epoch": 0.08271855939587569,
+      "grad_norm": 0.32630541920661926,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 712
+    },
+    {
+      "epoch": 0.0828347371478362,
+      "grad_norm": 0.35678330063819885,
+      "learning_rate": 0.0001,
+      "loss": 1.915,
+      "step": 713
+    },
+    {
+      "epoch": 0.08295091489979668,
+      "grad_norm": 0.3382868766784668,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 714
+    },
+    {
+      "epoch": 0.08306709265175719,
+      "grad_norm": 0.3485075831413269,
+      "learning_rate": 0.0001,
+      "loss": 1.7365,
+      "step": 715
+    },
+    {
+      "epoch": 0.08318327040371769,
+      "grad_norm": 0.31177419424057007,
+      "learning_rate": 0.0001,
+      "loss": 1.6356,
+      "step": 716
+    },
+    {
+      "epoch": 0.08329944815567819,
+      "grad_norm": 0.34105831384658813,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 717
+    },
+    {
+      "epoch": 0.0834156259076387,
+      "grad_norm": 0.3517579138278961,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 718
+    },
+    {
+      "epoch": 0.08353180365959918,
+      "grad_norm": 0.3053590953350067,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 719
+    },
+    {
+      "epoch": 0.08364798141155969,
+      "grad_norm": 0.35734882950782776,
+      "learning_rate": 0.0001,
+      "loss": 1.779,
+      "step": 720
+    },
+    {
+      "epoch": 0.08376415916352019,
+      "grad_norm": 0.33821558952331543,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 721
+    },
+    {
+      "epoch": 0.08388033691548069,
+      "grad_norm": 0.3388650417327881,
+      "learning_rate": 0.0001,
+      "loss": 1.7397,
+      "step": 722
+    },
+    {
+      "epoch": 0.08399651466744118,
+      "grad_norm": 0.36601775884628296,
+      "learning_rate": 0.0001,
+      "loss": 1.8082,
+      "step": 723
+    },
+    {
+      "epoch": 0.08411269241940168,
+      "grad_norm": 0.34071359038352966,
+      "learning_rate": 0.0001,
+      "loss": 1.7245,
+      "step": 724
+    },
+    {
+      "epoch": 0.08422887017136219,
+      "grad_norm": 0.34587469696998596,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 725
+    },
+    {
+      "epoch": 0.08434504792332269,
+      "grad_norm": 0.3538304269313812,
+      "learning_rate": 0.0001,
+      "loss": 1.973,
+      "step": 726
+    },
+    {
+      "epoch": 0.08446122567528318,
+      "grad_norm": 0.3469333350658417,
+      "learning_rate": 0.0001,
+      "loss": 1.8881,
+      "step": 727
+    },
+    {
+      "epoch": 0.08457740342724368,
+      "grad_norm": 0.31623193621635437,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 728
+    },
+    {
+      "epoch": 0.08469358117920418,
+      "grad_norm": 0.3332175016403198,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 729
+    },
+    {
+      "epoch": 0.08480975893116469,
+      "grad_norm": 0.35500526428222656,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 730
+    },
+    {
+      "epoch": 0.08492593668312517,
+      "grad_norm": 0.35975468158721924,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 731
+    },
+    {
+      "epoch": 0.08504211443508568,
+      "grad_norm": 0.33805137872695923,
+      "learning_rate": 0.0001,
+      "loss": 1.7452,
+      "step": 732
+    },
+    {
+      "epoch": 0.08515829218704618,
+      "grad_norm": 0.3446786105632782,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 733
+    },
+    {
+      "epoch": 0.08527446993900668,
+      "grad_norm": 0.3499710261821747,
+      "learning_rate": 0.0001,
+      "loss": 1.8734,
+      "step": 734
+    },
+    {
+      "epoch": 0.08539064769096719,
+      "grad_norm": 0.33205246925354004,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 735
+    },
+    {
+      "epoch": 0.08550682544292768,
+      "grad_norm": 0.3275506794452667,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 736
+    },
+    {
+      "epoch": 0.08562300319488818,
+      "grad_norm": 0.291427344083786,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 737
+    },
+    {
+      "epoch": 0.08573918094684868,
+      "grad_norm": 0.3342317044734955,
+      "learning_rate": 0.0001,
+      "loss": 1.8063,
+      "step": 738
+    },
+    {
+      "epoch": 0.08585535869880918,
+      "grad_norm": 0.3613170087337494,
+      "learning_rate": 0.0001,
+      "loss": 1.8471,
+      "step": 739
+    },
+    {
+      "epoch": 0.08597153645076967,
+      "grad_norm": 0.35327574610710144,
+      "learning_rate": 0.0001,
+      "loss": 1.8022,
+      "step": 740
+    },
+    {
+      "epoch": 0.08608771420273018,
+      "grad_norm": 0.32750821113586426,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 741
+    },
+    {
+      "epoch": 0.08620389195469068,
+      "grad_norm": 0.37629932165145874,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 742
+    },
+    {
+      "epoch": 0.08632006970665118,
+      "grad_norm": 0.3730964958667755,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 743
+    },
+    {
+      "epoch": 0.08643624745861167,
+      "grad_norm": 0.36657699942588806,
+      "learning_rate": 0.0001,
+      "loss": 1.8363,
+      "step": 744
+    },
+    {
+      "epoch": 0.08655242521057217,
+      "grad_norm": 0.3720366358757019,
+      "learning_rate": 0.0001,
+      "loss": 1.9311,
+      "step": 745
+    },
+    {
+      "epoch": 0.08666860296253268,
+      "grad_norm": 0.34417784214019775,
+      "learning_rate": 0.0001,
+      "loss": 1.902,
+      "step": 746
+    },
+    {
+      "epoch": 0.08678478071449318,
+      "grad_norm": 0.34215980768203735,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 747
+    },
+    {
+      "epoch": 0.08690095846645367,
+      "grad_norm": 0.37166187167167664,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 748
+    },
+    {
+      "epoch": 0.08701713621841417,
+      "grad_norm": 0.3440292179584503,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 749
+    },
+    {
+      "epoch": 0.08713331397037467,
+      "grad_norm": 0.38115811347961426,
+      "learning_rate": 0.0001,
+      "loss": 1.8192,
+      "step": 750
+    },
+    {
+      "epoch": 0.08724949172233518,
+      "grad_norm": 0.38101625442504883,
+      "learning_rate": 0.0001,
+      "loss": 1.8616,
+      "step": 751
+    },
+    {
+      "epoch": 0.08736566947429568,
+      "grad_norm": 0.3463743031024933,
+      "learning_rate": 0.0001,
+      "loss": 1.7568,
+      "step": 752
+    },
+    {
+      "epoch": 0.08748184722625617,
+      "grad_norm": 0.36165106296539307,
+      "learning_rate": 0.0001,
+      "loss": 1.8383,
+      "step": 753
+    },
+    {
+      "epoch": 0.08759802497821667,
+      "grad_norm": 0.31763550639152527,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 754
+    },
+    {
+      "epoch": 0.08771420273017717,
+      "grad_norm": 0.35172948241233826,
+      "learning_rate": 0.0001,
+      "loss": 1.8296,
+      "step": 755
+    },
+    {
+      "epoch": 0.08783038048213768,
+      "grad_norm": 0.3546104431152344,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 756
+    },
+    {
+      "epoch": 0.08794655823409817,
+      "grad_norm": 0.3244597911834717,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 757
+    },
+    {
+      "epoch": 0.08806273598605867,
+      "grad_norm": 0.33493107557296753,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 758
+    },
+    {
+      "epoch": 0.08817891373801917,
+      "grad_norm": 0.3546777665615082,
+      "learning_rate": 0.0001,
+      "loss": 1.7593,
+      "step": 759
+    },
+    {
+      "epoch": 0.08829509148997967,
+      "grad_norm": 0.3528088629245758,
+      "learning_rate": 0.0001,
+      "loss": 1.8915,
+      "step": 760
+    },
+    {
+      "epoch": 0.08841126924194016,
+      "grad_norm": 0.32234612107276917,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 761
+    },
+    {
+      "epoch": 0.08852744699390067,
+      "grad_norm": 0.35947954654693604,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 762
+    },
+    {
+      "epoch": 0.08864362474586117,
+      "grad_norm": 0.34787270426750183,
+      "learning_rate": 0.0001,
+      "loss": 1.7837,
+      "step": 763
+    },
+    {
+      "epoch": 0.08875980249782167,
+      "grad_norm": 0.3539670407772064,
+      "learning_rate": 0.0001,
+      "loss": 1.8225,
+      "step": 764
+    },
+    {
+      "epoch": 0.08887598024978216,
+      "grad_norm": 0.3532291352748871,
+      "learning_rate": 0.0001,
+      "loss": 1.7474,
+      "step": 765
+    },
+    {
+      "epoch": 0.08899215800174266,
+      "grad_norm": 0.35020220279693604,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 766
+    },
+    {
+      "epoch": 0.08910833575370317,
+      "grad_norm": 0.3528810441493988,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 767
+    },
+    {
+      "epoch": 0.08922451350566367,
+      "grad_norm": 0.38562941551208496,
+      "learning_rate": 0.0001,
+      "loss": 1.7691,
+      "step": 768
+    },
+    {
+      "epoch": 0.08934069125762417,
+      "grad_norm": 0.3530423641204834,
+      "learning_rate": 0.0001,
+      "loss": 1.8129,
+      "step": 769
+    },
+    {
+      "epoch": 0.08945686900958466,
+      "grad_norm": 0.35964086651802063,
+      "learning_rate": 0.0001,
+      "loss": 1.8651,
+      "step": 770
+    },
+    {
+      "epoch": 0.08957304676154516,
+      "grad_norm": 0.3369607627391815,
+      "learning_rate": 0.0001,
+      "loss": 1.7238,
+      "step": 771
+    },
+    {
+      "epoch": 0.08968922451350567,
+      "grad_norm": 0.3420158326625824,
+      "learning_rate": 0.0001,
+      "loss": 1.7426,
+      "step": 772
+    },
+    {
+      "epoch": 0.08980540226546617,
+      "grad_norm": 0.3466860353946686,
+      "learning_rate": 0.0001,
+      "loss": 1.7715,
+      "step": 773
+    },
+    {
+      "epoch": 0.08992158001742666,
+      "grad_norm": 0.33295726776123047,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 774
+    },
+    {
+      "epoch": 0.09003775776938716,
+      "grad_norm": 0.2985909581184387,
+      "learning_rate": 0.0001,
+      "loss": 1.5965,
+      "step": 775
+    },
+    {
+      "epoch": 0.09015393552134766,
+      "grad_norm": 0.334733247756958,
+      "learning_rate": 0.0001,
+      "loss": 1.8092,
+      "step": 776
+    },
+    {
+      "epoch": 0.09027011327330817,
+      "grad_norm": 0.3320635259151459,
+      "learning_rate": 0.0001,
+      "loss": 1.6475,
+      "step": 777
+    },
+    {
+      "epoch": 0.09038629102526866,
+      "grad_norm": 0.33000028133392334,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 778
+    },
+    {
+      "epoch": 0.09050246877722916,
+      "grad_norm": 0.3425371050834656,
+      "learning_rate": 0.0001,
+      "loss": 1.8284,
+      "step": 779
+    },
+    {
+      "epoch": 0.09061864652918966,
+      "grad_norm": 0.3471533954143524,
+      "learning_rate": 0.0001,
+      "loss": 1.824,
+      "step": 780
+    },
+    {
+      "epoch": 0.09073482428115016,
+      "grad_norm": 0.34877267479896545,
+      "learning_rate": 0.0001,
+      "loss": 1.7793,
+      "step": 781
+    },
+    {
+      "epoch": 0.09085100203311065,
+      "grad_norm": 0.33842357993125916,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 782
+    },
+    {
+      "epoch": 0.09096717978507116,
+      "grad_norm": 0.3466620445251465,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 783
+    },
+    {
+      "epoch": 0.09108335753703166,
+      "grad_norm": 0.3401154577732086,
+      "learning_rate": 0.0001,
+      "loss": 1.8291,
+      "step": 784
+    },
+    {
+      "epoch": 0.09119953528899216,
+      "grad_norm": 0.3380959630012512,
+      "learning_rate": 0.0001,
+      "loss": 1.7823,
+      "step": 785
+    },
+    {
+      "epoch": 0.09131571304095265,
+      "grad_norm": 0.3219566345214844,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 786
+    },
+    {
+      "epoch": 0.09143189079291315,
+      "grad_norm": 0.36787810921669006,
+      "learning_rate": 0.0001,
+      "loss": 1.8798,
+      "step": 787
+    },
+    {
+      "epoch": 0.09154806854487366,
+      "grad_norm": 0.3317922055721283,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 788
+    },
+    {
+      "epoch": 0.09166424629683416,
+      "grad_norm": 0.32760554552078247,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 789
+    },
+    {
+      "epoch": 0.09178042404879466,
+      "grad_norm": 0.3528217375278473,
+      "learning_rate": 0.0001,
+      "loss": 1.8356,
+      "step": 790
+    },
+    {
+      "epoch": 0.09189660180075515,
+      "grad_norm": 0.3439396321773529,
+      "learning_rate": 0.0001,
+      "loss": 1.8907,
+      "step": 791
+    },
+    {
+      "epoch": 0.09201277955271565,
+      "grad_norm": 0.32898107171058655,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 792
+    },
+    {
+      "epoch": 0.09212895730467616,
+      "grad_norm": 0.35222840309143066,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 793
+    },
+    {
+      "epoch": 0.09224513505663666,
+      "grad_norm": 0.35788872838020325,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 794
+    },
+    {
+      "epoch": 0.09236131280859715,
+      "grad_norm": 0.34906426072120667,
+      "learning_rate": 0.0001,
+      "loss": 1.7877,
+      "step": 795
+    },
+    {
+      "epoch": 0.09247749056055765,
+      "grad_norm": 0.3351164162158966,
+      "learning_rate": 0.0001,
+      "loss": 1.7754,
+      "step": 796
+    },
+    {
+      "epoch": 0.09259366831251815,
+      "grad_norm": 0.34004396200180054,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 797
+    },
+    {
+      "epoch": 0.09270984606447866,
+      "grad_norm": 0.3525846302509308,
+      "learning_rate": 0.0001,
+      "loss": 1.852,
+      "step": 798
+    },
+    {
+      "epoch": 0.09282602381643915,
+      "grad_norm": 0.3507898449897766,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 799
+    },
+    {
+      "epoch": 0.09294220156839965,
+      "grad_norm": 0.33934998512268066,
+      "learning_rate": 0.0001,
+      "loss": 1.8192,
+      "step": 800
+    },
+    {
+      "epoch": 0.09305837932036015,
+      "grad_norm": 0.3526393473148346,
+      "learning_rate": 0.0001,
+      "loss": 1.7328,
+      "step": 801
+    },
+    {
+      "epoch": 0.09317455707232065,
+      "grad_norm": 0.3338293135166168,
+      "learning_rate": 0.0001,
+      "loss": 1.8076,
+      "step": 802
+    },
+    {
+      "epoch": 0.09329073482428114,
+      "grad_norm": 0.3807794451713562,
+      "learning_rate": 0.0001,
+      "loss": 1.9431,
+      "step": 803
+    },
+    {
+      "epoch": 0.09340691257624165,
+      "grad_norm": 0.33969688415527344,
+      "learning_rate": 0.0001,
+      "loss": 1.8415,
+      "step": 804
+    },
+    {
+      "epoch": 0.09352309032820215,
+      "grad_norm": 0.3313851058483124,
+      "learning_rate": 0.0001,
+      "loss": 1.7324,
+      "step": 805
+    },
+    {
+      "epoch": 0.09363926808016265,
+      "grad_norm": 0.3267270624637604,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 806
+    },
+    {
+      "epoch": 0.09375544583212315,
+      "grad_norm": 0.3477146327495575,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 807
+    },
+    {
+      "epoch": 0.09387162358408364,
+      "grad_norm": 0.36601418256759644,
+      "learning_rate": 0.0001,
+      "loss": 1.8103,
+      "step": 808
+    },
+    {
+      "epoch": 0.09398780133604415,
+      "grad_norm": 0.33894267678260803,
+      "learning_rate": 0.0001,
+      "loss": 1.7657,
+      "step": 809
+    },
+    {
+      "epoch": 0.09410397908800465,
+      "grad_norm": 0.35825762152671814,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 810
+    },
+    {
+      "epoch": 0.09422015683996515,
+      "grad_norm": 0.3400390148162842,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 811
+    },
+    {
+      "epoch": 0.09433633459192564,
+      "grad_norm": 0.32821226119995117,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 812
+    },
+    {
+      "epoch": 0.09445251234388614,
+      "grad_norm": 0.3435446619987488,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 813
+    },
+    {
+      "epoch": 0.09456869009584665,
+      "grad_norm": 0.37115979194641113,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 814
+    },
+    {
+      "epoch": 0.09468486784780715,
+      "grad_norm": 0.3451048731803894,
+      "learning_rate": 0.0001,
+      "loss": 1.6982,
+      "step": 815
+    },
+    {
+      "epoch": 0.09480104559976764,
+      "grad_norm": 0.3507342040538788,
+      "learning_rate": 0.0001,
+      "loss": 1.8472,
+      "step": 816
+    },
+    {
+      "epoch": 0.09491722335172814,
+      "grad_norm": 0.32728418707847595,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 817
+    },
+    {
+      "epoch": 0.09503340110368864,
+      "grad_norm": 0.3312884271144867,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 818
+    },
+    {
+      "epoch": 0.09514957885564915,
+      "grad_norm": 0.3303951025009155,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 819
+    },
+    {
+      "epoch": 0.09526575660760964,
+      "grad_norm": 0.3290101885795593,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 820
+    },
+    {
+      "epoch": 0.09538193435957014,
+      "grad_norm": 0.36628472805023193,
+      "learning_rate": 0.0001,
+      "loss": 1.7772,
+      "step": 821
+    },
+    {
+      "epoch": 0.09549811211153064,
+      "grad_norm": 0.3897988796234131,
+      "learning_rate": 0.0001,
+      "loss": 1.808,
+      "step": 822
+    },
+    {
+      "epoch": 0.09561428986349114,
+      "grad_norm": 0.3413255512714386,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 823
+    },
+    {
+      "epoch": 0.09573046761545165,
+      "grad_norm": 0.4127778112888336,
+      "learning_rate": 0.0001,
+      "loss": 1.8115,
+      "step": 824
+    },
+    {
+      "epoch": 0.09584664536741214,
+      "grad_norm": 0.3477415442466736,
+      "learning_rate": 0.0001,
+      "loss": 1.6128,
+      "step": 825
+    },
+    {
+      "epoch": 0.09596282311937264,
+      "grad_norm": 0.32524755597114563,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 826
+    },
+    {
+      "epoch": 0.09607900087133314,
+      "grad_norm": 0.36483168601989746,
+      "learning_rate": 0.0001,
+      "loss": 1.8151,
+      "step": 827
+    },
+    {
+      "epoch": 0.09619517862329365,
+      "grad_norm": 0.34030166268348694,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 828
+    },
+    {
+      "epoch": 0.09631135637525413,
+      "grad_norm": 0.3241477906703949,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 829
+    },
+    {
+      "epoch": 0.09642753412721464,
+      "grad_norm": 0.3371472954750061,
+      "learning_rate": 0.0001,
+      "loss": 1.7769,
+      "step": 830
+    },
+    {
+      "epoch": 0.09654371187917514,
+      "grad_norm": 0.3401608169078827,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 831
+    },
+    {
+      "epoch": 0.09665988963113564,
+      "grad_norm": 0.32106491923332214,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 832
+    },
+    {
+      "epoch": 0.09677606738309613,
+      "grad_norm": 0.34740883111953735,
+      "learning_rate": 0.0001,
+      "loss": 1.8344,
+      "step": 833
+    },
+    {
+      "epoch": 0.09689224513505663,
+      "grad_norm": 0.3772992789745331,
+      "learning_rate": 0.0001,
+      "loss": 1.7999,
+      "step": 834
+    },
+    {
+      "epoch": 0.09700842288701714,
+      "grad_norm": 0.3412041664123535,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 835
+    },
+    {
+      "epoch": 0.09712460063897764,
+      "grad_norm": 0.34832799434661865,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 836
+    },
+    {
+      "epoch": 0.09724077839093813,
+      "grad_norm": 0.35847747325897217,
+      "learning_rate": 0.0001,
+      "loss": 1.8176,
+      "step": 837
+    },
+    {
+      "epoch": 0.09735695614289863,
+      "grad_norm": 0.36482107639312744,
+      "learning_rate": 0.0001,
+      "loss": 1.8864,
+      "step": 838
+    },
+    {
+      "epoch": 0.09747313389485913,
+      "grad_norm": 0.40612149238586426,
+      "learning_rate": 0.0001,
+      "loss": 1.8004,
+      "step": 839
+    },
+    {
+      "epoch": 0.09758931164681964,
+      "grad_norm": 0.3522476255893707,
+      "learning_rate": 0.0001,
+      "loss": 1.8136,
+      "step": 840
+    },
+    {
+      "epoch": 0.09770548939878014,
+      "grad_norm": 0.37533196806907654,
+      "learning_rate": 0.0001,
+      "loss": 1.8091,
+      "step": 841
+    },
+    {
+      "epoch": 0.09782166715074063,
+      "grad_norm": 0.3364860415458679,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 842
+    },
+    {
+      "epoch": 0.09793784490270113,
+      "grad_norm": 0.34495505690574646,
+      "learning_rate": 0.0001,
+      "loss": 1.8331,
+      "step": 843
+    },
+    {
+      "epoch": 0.09805402265466164,
+      "grad_norm": 0.3652923107147217,
+      "learning_rate": 0.0001,
+      "loss": 1.8655,
+      "step": 844
+    },
+    {
+      "epoch": 0.09817020040662214,
+      "grad_norm": 0.35473141074180603,
+      "learning_rate": 0.0001,
+      "loss": 1.907,
+      "step": 845
+    },
+    {
+      "epoch": 0.09828637815858263,
+      "grad_norm": 0.36820659041404724,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 846
+    },
+    {
+      "epoch": 0.09840255591054313,
+      "grad_norm": 0.34225067496299744,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 847
+    },
+    {
+      "epoch": 0.09851873366250363,
+      "grad_norm": 0.3609171509742737,
+      "learning_rate": 0.0001,
+      "loss": 1.7822,
+      "step": 848
+    },
+    {
+      "epoch": 0.09863491141446414,
+      "grad_norm": 0.3501855731010437,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 849
+    },
+    {
+      "epoch": 0.09875108916642462,
+      "grad_norm": 0.32687613368034363,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 850
+    },
+    {
+      "epoch": 0.09886726691838513,
+      "grad_norm": 0.34075310826301575,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 851
+    },
+    {
+      "epoch": 0.09898344467034563,
+      "grad_norm": 0.3613976836204529,
+      "learning_rate": 0.0001,
+      "loss": 1.9505,
+      "step": 852
+    },
+    {
+      "epoch": 0.09909962242230613,
+      "grad_norm": 0.33815905451774597,
+      "learning_rate": 0.0001,
+      "loss": 1.7972,
+      "step": 853
+    },
+    {
+      "epoch": 0.09921580017426662,
+      "grad_norm": 0.35017895698547363,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 854
+    },
+    {
+      "epoch": 0.09933197792622712,
+      "grad_norm": 0.3758436441421509,
+      "learning_rate": 0.0001,
+      "loss": 1.6596,
+      "step": 855
+    },
+    {
+      "epoch": 0.09944815567818763,
+      "grad_norm": 0.32959380745887756,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 856
+    },
+    {
+      "epoch": 0.09956433343014813,
+      "grad_norm": 0.3575064539909363,
+      "learning_rate": 0.0001,
+      "loss": 1.8109,
+      "step": 857
+    },
+    {
+      "epoch": 0.09968051118210862,
+      "grad_norm": 0.3391128182411194,
+      "learning_rate": 0.0001,
+      "loss": 1.7841,
+      "step": 858
+    },
+    {
+      "epoch": 0.09979668893406912,
+      "grad_norm": 0.33528846502304077,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 859
+    },
+    {
+      "epoch": 0.09991286668602962,
+      "grad_norm": 0.32930776476860046,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 860
+    },
+    {
+      "epoch": 0.10002904443799013,
+      "grad_norm": 0.34488263726234436,
+      "learning_rate": 0.0001,
+      "loss": 1.7357,
+      "step": 861
+    },
+    {
+      "epoch": 0.10014522218995063,
+      "grad_norm": 0.36875104904174805,
+      "learning_rate": 0.0001,
+      "loss": 1.8439,
+      "step": 862
+    },
+    {
+      "epoch": 0.10026139994191112,
+      "grad_norm": 0.3462197184562683,
+      "learning_rate": 0.0001,
+      "loss": 1.7203,
+      "step": 863
+    },
+    {
+      "epoch": 0.10037757769387162,
+      "grad_norm": 0.38343024253845215,
+      "learning_rate": 0.0001,
+      "loss": 2.0117,
+      "step": 864
+    },
+    {
+      "epoch": 0.10049375544583213,
+      "grad_norm": 0.3336431682109833,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 865
+    },
+    {
+      "epoch": 0.10060993319779263,
+      "grad_norm": 0.3539637327194214,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 866
+    },
+    {
+      "epoch": 0.10072611094975312,
+      "grad_norm": 0.38433191180229187,
+      "learning_rate": 0.0001,
+      "loss": 1.8894,
+      "step": 867
+    },
+    {
+      "epoch": 0.10084228870171362,
+      "grad_norm": 0.34169065952301025,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 868
+    },
+    {
+      "epoch": 0.10095846645367412,
+      "grad_norm": 0.3778311610221863,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 869
+    },
+    {
+      "epoch": 0.10107464420563463,
+      "grad_norm": 0.3747129440307617,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 870
+    },
+    {
+      "epoch": 0.10119082195759511,
+      "grad_norm": 0.34259381890296936,
+      "learning_rate": 0.0001,
+      "loss": 1.7926,
+      "step": 871
+    },
+    {
+      "epoch": 0.10130699970955562,
+      "grad_norm": 0.35702332854270935,
+      "learning_rate": 0.0001,
+      "loss": 1.7456,
+      "step": 872
+    },
+    {
+      "epoch": 0.10142317746151612,
+      "grad_norm": 0.35375505685806274,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 873
+    },
+    {
+      "epoch": 0.10153935521347662,
+      "grad_norm": 0.35189199447631836,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 874
+    },
+    {
+      "epoch": 0.10165553296543711,
+      "grad_norm": 0.3281596601009369,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 875
+    },
+    {
+      "epoch": 0.10177171071739761,
+      "grad_norm": 0.34264156222343445,
+      "learning_rate": 0.0001,
+      "loss": 1.8293,
+      "step": 876
+    },
+    {
+      "epoch": 0.10188788846935812,
+      "grad_norm": 0.3711477518081665,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 877
+    },
+    {
+      "epoch": 0.10200406622131862,
+      "grad_norm": 0.31812041997909546,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 878
+    },
+    {
+      "epoch": 0.10212024397327912,
+      "grad_norm": 0.3646787405014038,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 879
+    },
+    {
+      "epoch": 0.10223642172523961,
+      "grad_norm": 0.3644947111606598,
+      "learning_rate": 0.0001,
+      "loss": 1.8658,
+      "step": 880
+    },
+    {
+      "epoch": 0.10235259947720012,
+      "grad_norm": 0.345739483833313,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 881
+    },
+    {
+      "epoch": 0.10246877722916062,
+      "grad_norm": 0.3633933663368225,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 882
+    },
+    {
+      "epoch": 0.10258495498112112,
+      "grad_norm": 0.35784977674484253,
+      "learning_rate": 0.0001,
+      "loss": 1.7953,
+      "step": 883
+    },
+    {
+      "epoch": 0.10270113273308161,
+      "grad_norm": 0.34802740812301636,
+      "learning_rate": 0.0001,
+      "loss": 1.7537,
+      "step": 884
+    },
+    {
+      "epoch": 0.10281731048504211,
+      "grad_norm": 0.334532231092453,
+      "learning_rate": 0.0001,
+      "loss": 1.6814,
+      "step": 885
+    },
+    {
+      "epoch": 0.10293348823700262,
+      "grad_norm": 0.37249666452407837,
+      "learning_rate": 0.0001,
+      "loss": 1.838,
+      "step": 886
+    },
+    {
+      "epoch": 0.10304966598896312,
+      "grad_norm": 0.37367671728134155,
+      "learning_rate": 0.0001,
+      "loss": 1.8214,
+      "step": 887
+    },
+    {
+      "epoch": 0.10316584374092361,
+      "grad_norm": 0.32655900716781616,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 888
+    },
+    {
+      "epoch": 0.10328202149288411,
+      "grad_norm": 0.33447375893592834,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 889
+    },
+    {
+      "epoch": 0.10339819924484461,
+      "grad_norm": 0.3553028404712677,
+      "learning_rate": 0.0001,
+      "loss": 1.9126,
+      "step": 890
+    },
+    {
+      "epoch": 0.10351437699680512,
+      "grad_norm": 0.3429213762283325,
+      "learning_rate": 0.0001,
+      "loss": 1.7968,
+      "step": 891
+    },
+    {
+      "epoch": 0.1036305547487656,
+      "grad_norm": 0.3589216470718384,
+      "learning_rate": 0.0001,
+      "loss": 1.87,
+      "step": 892
+    },
+    {
+      "epoch": 0.10374673250072611,
+      "grad_norm": 0.35310253500938416,
+      "learning_rate": 0.0001,
+      "loss": 1.7481,
+      "step": 893
+    },
+    {
+      "epoch": 0.10386291025268661,
+      "grad_norm": 0.36510980129241943,
+      "learning_rate": 0.0001,
+      "loss": 1.8996,
+      "step": 894
+    },
+    {
+      "epoch": 0.10397908800464711,
+      "grad_norm": 0.3604600429534912,
+      "learning_rate": 0.0001,
+      "loss": 1.8,
+      "step": 895
+    },
+    {
+      "epoch": 0.10409526575660762,
+      "grad_norm": 0.3504015803337097,
+      "learning_rate": 0.0001,
+      "loss": 1.706,
+      "step": 896
+    },
+    {
+      "epoch": 0.1042114435085681,
+      "grad_norm": 0.3666890561580658,
+      "learning_rate": 0.0001,
+      "loss": 1.9185,
+      "step": 897
+    },
+    {
+      "epoch": 0.10432762126052861,
+      "grad_norm": 0.32218024134635925,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 898
+    },
+    {
+      "epoch": 0.10444379901248911,
+      "grad_norm": 0.32854628562927246,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 899
+    },
+    {
+      "epoch": 0.10455997676444961,
+      "grad_norm": 0.34168779850006104,
+      "learning_rate": 0.0001,
+      "loss": 1.7293,
+      "step": 900
+    },
+    {
+      "epoch": 0.1046761545164101,
+      "grad_norm": 0.3389854431152344,
+      "learning_rate": 0.0001,
+      "loss": 1.7061,
+      "step": 901
+    },
+    {
+      "epoch": 0.1047923322683706,
+      "grad_norm": 0.3675687611103058,
+      "learning_rate": 0.0001,
+      "loss": 1.9126,
+      "step": 902
+    },
+    {
+      "epoch": 0.10490851002033111,
+      "grad_norm": 0.33231931924819946,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 903
+    },
+    {
+      "epoch": 0.10502468777229161,
+      "grad_norm": 0.3507463037967682,
+      "learning_rate": 0.0001,
+      "loss": 1.7924,
+      "step": 904
+    },
+    {
+      "epoch": 0.1051408655242521,
+      "grad_norm": 0.35986390709877014,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 905
+    },
+    {
+      "epoch": 0.1052570432762126,
+      "grad_norm": 0.36210981011390686,
+      "learning_rate": 0.0001,
+      "loss": 1.8232,
+      "step": 906
+    },
+    {
+      "epoch": 0.1053732210281731,
+      "grad_norm": 0.3577278256416321,
+      "learning_rate": 0.0001,
+      "loss": 1.7484,
+      "step": 907
+    },
+    {
+      "epoch": 0.10548939878013361,
+      "grad_norm": 0.3454272449016571,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 908
+    },
+    {
+      "epoch": 0.1056055765320941,
+      "grad_norm": 0.34774741530418396,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 909
+    },
+    {
+      "epoch": 0.1057217542840546,
+      "grad_norm": 0.3502449095249176,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 910
+    },
+    {
+      "epoch": 0.1058379320360151,
+      "grad_norm": 0.36542028188705444,
+      "learning_rate": 0.0001,
+      "loss": 1.8321,
+      "step": 911
+    },
+    {
+      "epoch": 0.1059541097879756,
+      "grad_norm": 0.35387906432151794,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 912
+    },
+    {
+      "epoch": 0.10607028753993611,
+      "grad_norm": 0.3482963442802429,
+      "learning_rate": 0.0001,
+      "loss": 1.7885,
+      "step": 913
+    },
+    {
+      "epoch": 0.1061864652918966,
+      "grad_norm": 0.3488115668296814,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 914
+    },
+    {
+      "epoch": 0.1063026430438571,
+      "grad_norm": 0.3296864926815033,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 915
+    },
+    {
+      "epoch": 0.1064188207958176,
+      "grad_norm": 0.34132471680641174,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 916
+    },
+    {
+      "epoch": 0.1065349985477781,
+      "grad_norm": 0.3466493487358093,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 917
+    },
+    {
+      "epoch": 0.1066511762997386,
+      "grad_norm": 0.3528394401073456,
+      "learning_rate": 0.0001,
+      "loss": 1.847,
+      "step": 918
+    },
+    {
+      "epoch": 0.1067673540516991,
+      "grad_norm": 0.32509297132492065,
+      "learning_rate": 0.0001,
+      "loss": 1.681,
+      "step": 919
+    },
+    {
+      "epoch": 0.1068835318036596,
+      "grad_norm": 0.3523486852645874,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 920
+    },
+    {
+      "epoch": 0.1069997095556201,
+      "grad_norm": 0.3355209529399872,
+      "learning_rate": 0.0001,
+      "loss": 1.6507,
+      "step": 921
+    },
+    {
+      "epoch": 0.10711588730758059,
+      "grad_norm": 0.3474280834197998,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 922
+    },
+    {
+      "epoch": 0.1072320650595411,
+      "grad_norm": 0.359402596950531,
+      "learning_rate": 0.0001,
+      "loss": 1.8466,
+      "step": 923
+    },
+    {
+      "epoch": 0.1073482428115016,
+      "grad_norm": 0.3309195339679718,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 924
+    },
+    {
+      "epoch": 0.1074644205634621,
+      "grad_norm": 0.3497809171676636,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 925
+    },
+    {
+      "epoch": 0.10758059831542259,
+      "grad_norm": 0.34211966395378113,
+      "learning_rate": 0.0001,
+      "loss": 1.7996,
+      "step": 926
+    },
+    {
+      "epoch": 0.1076967760673831,
+      "grad_norm": 0.3558671772480011,
+      "learning_rate": 0.0001,
+      "loss": 1.8548,
+      "step": 927
+    },
+    {
+      "epoch": 0.1078129538193436,
+      "grad_norm": 0.3554135859012604,
+      "learning_rate": 0.0001,
+      "loss": 1.873,
+      "step": 928
+    },
+    {
+      "epoch": 0.1079291315713041,
+      "grad_norm": 0.35786765813827515,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 929
+    },
+    {
+      "epoch": 0.1080453093232646,
+      "grad_norm": 0.3519211709499359,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 930
+    },
+    {
+      "epoch": 0.10816148707522509,
+      "grad_norm": 0.37536898255348206,
+      "learning_rate": 0.0001,
+      "loss": 1.8926,
+      "step": 931
+    },
+    {
+      "epoch": 0.1082776648271856,
+      "grad_norm": 0.3422173857688904,
+      "learning_rate": 0.0001,
+      "loss": 1.8172,
+      "step": 932
+    },
+    {
+      "epoch": 0.1083938425791461,
+      "grad_norm": 0.35913097858428955,
+      "learning_rate": 0.0001,
+      "loss": 1.8137,
+      "step": 933
+    },
+    {
+      "epoch": 0.1085100203311066,
+      "grad_norm": 0.3589080274105072,
+      "learning_rate": 0.0001,
+      "loss": 1.7949,
+      "step": 934
+    },
+    {
+      "epoch": 0.10862619808306709,
+      "grad_norm": 0.37230291962623596,
+      "learning_rate": 0.0001,
+      "loss": 1.8148,
+      "step": 935
+    },
+    {
+      "epoch": 0.10874237583502759,
+      "grad_norm": 0.3313770294189453,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 936
+    },
+    {
+      "epoch": 0.1088585535869881,
+      "grad_norm": 0.344086229801178,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 937
+    },
+    {
+      "epoch": 0.1089747313389486,
+      "grad_norm": 0.3584043085575104,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 938
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 0.34426915645599365,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 939
+    },
+    {
+      "epoch": 0.10920708684286959,
+      "grad_norm": 0.340386301279068,
+      "learning_rate": 0.0001,
+      "loss": 1.7741,
+      "step": 940
+    },
+    {
+      "epoch": 0.10932326459483009,
+      "grad_norm": 0.3348495662212372,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 941
+    },
+    {
+      "epoch": 0.1094394423467906,
+      "grad_norm": 0.35593461990356445,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 942
+    },
+    {
+      "epoch": 0.10955562009875108,
+      "grad_norm": 0.3561510145664215,
+      "learning_rate": 0.0001,
+      "loss": 1.8853,
+      "step": 943
+    },
+    {
+      "epoch": 0.10967179785071159,
+      "grad_norm": 0.36038845777511597,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 944
+    },
+    {
+      "epoch": 0.10978797560267209,
+      "grad_norm": 0.3256826400756836,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 945
+    },
+    {
+      "epoch": 0.10990415335463259,
+      "grad_norm": 0.3304516673088074,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 946
+    },
+    {
+      "epoch": 0.11002033110659308,
+      "grad_norm": 0.34142622351646423,
+      "learning_rate": 0.0001,
+      "loss": 1.7302,
+      "step": 947
+    },
+    {
+      "epoch": 0.11013650885855358,
+      "grad_norm": 0.3722458481788635,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 948
+    },
+    {
+      "epoch": 0.11025268661051409,
+      "grad_norm": 0.35679808259010315,
+      "learning_rate": 0.0001,
+      "loss": 1.7921,
+      "step": 949
+    },
+    {
+      "epoch": 0.11036886436247459,
+      "grad_norm": 0.3507538139820099,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 950
+    },
+    {
+      "epoch": 0.11048504211443509,
+      "grad_norm": 0.34563156962394714,
+      "learning_rate": 0.0001,
+      "loss": 1.7972,
+      "step": 951
+    },
+    {
+      "epoch": 0.11060121986639558,
+      "grad_norm": 0.35172975063323975,
+      "learning_rate": 0.0001,
+      "loss": 1.8513,
+      "step": 952
+    },
+    {
+      "epoch": 0.11071739761835608,
+      "grad_norm": 0.3704832196235657,
+      "learning_rate": 0.0001,
+      "loss": 1.7626,
+      "step": 953
+    },
+    {
+      "epoch": 0.11083357537031659,
+      "grad_norm": 0.3332252502441406,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 954
+    },
+    {
+      "epoch": 0.11094975312227709,
+      "grad_norm": 0.33239227533340454,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 955
+    },
+    {
+      "epoch": 0.11106593087423758,
+      "grad_norm": 0.35018518567085266,
+      "learning_rate": 0.0001,
+      "loss": 1.8144,
+      "step": 956
+    },
+    {
+      "epoch": 0.11118210862619808,
+      "grad_norm": 0.34933680295944214,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 957
+    },
+    {
+      "epoch": 0.11129828637815858,
+      "grad_norm": 0.38236263394355774,
+      "learning_rate": 0.0001,
+      "loss": 1.8272,
+      "step": 958
+    },
+    {
+      "epoch": 0.11141446413011909,
+      "grad_norm": 0.3630216717720032,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 959
+    },
+    {
+      "epoch": 0.11153064188207958,
+      "grad_norm": 0.3446943461894989,
+      "learning_rate": 0.0001,
+      "loss": 1.8034,
+      "step": 960
+    },
+    {
+      "epoch": 0.11164681963404008,
+      "grad_norm": 0.40010347962379456,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 961
+    },
+    {
+      "epoch": 0.11176299738600058,
+      "grad_norm": 0.35464581847190857,
+      "learning_rate": 0.0001,
+      "loss": 1.7943,
+      "step": 962
+    },
+    {
+      "epoch": 0.11187917513796108,
+      "grad_norm": 0.38472074270248413,
+      "learning_rate": 0.0001,
+      "loss": 1.8237,
+      "step": 963
+    },
+    {
+      "epoch": 0.11199535288992157,
+      "grad_norm": 0.3539893329143524,
+      "learning_rate": 0.0001,
+      "loss": 1.7777,
+      "step": 964
+    },
+    {
+      "epoch": 0.11211153064188208,
+      "grad_norm": 0.3627575933933258,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 965
+    },
+    {
+      "epoch": 0.11222770839384258,
+      "grad_norm": 0.3388507664203644,
+      "learning_rate": 0.0001,
+      "loss": 1.7009,
+      "step": 966
+    },
+    {
+      "epoch": 0.11234388614580308,
+      "grad_norm": 0.3800560534000397,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 967
+    },
+    {
+      "epoch": 0.11246006389776358,
+      "grad_norm": 0.3563110828399658,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 968
+    },
+    {
+      "epoch": 0.11257624164972407,
+      "grad_norm": 0.34514880180358887,
+      "learning_rate": 0.0001,
+      "loss": 1.7658,
+      "step": 969
+    },
+    {
+      "epoch": 0.11269241940168458,
+      "grad_norm": 0.35564523935317993,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 970
+    },
+    {
+      "epoch": 0.11280859715364508,
+      "grad_norm": 0.35224324464797974,
+      "learning_rate": 0.0001,
+      "loss": 1.8569,
+      "step": 971
+    },
+    {
+      "epoch": 0.11292477490560558,
+      "grad_norm": 0.344664990901947,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 972
+    },
+    {
+      "epoch": 0.11304095265756607,
+      "grad_norm": 0.3399982452392578,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 973
+    },
+    {
+      "epoch": 0.11315713040952657,
+      "grad_norm": 0.3541862368583679,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 974
+    },
+    {
+      "epoch": 0.11327330816148708,
+      "grad_norm": 0.35132747888565063,
+      "learning_rate": 0.0001,
+      "loss": 1.7683,
+      "step": 975
+    },
+    {
+      "epoch": 0.11338948591344758,
+      "grad_norm": 0.3748994469642639,
+      "learning_rate": 0.0001,
+      "loss": 1.6813,
+      "step": 976
+    },
+    {
+      "epoch": 0.11350566366540807,
+      "grad_norm": 0.3247508406639099,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 977
+    },
+    {
+      "epoch": 0.11362184141736857,
+      "grad_norm": 0.33404844999313354,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 978
+    },
+    {
+      "epoch": 0.11373801916932907,
+      "grad_norm": 0.36675605177879333,
+      "learning_rate": 0.0001,
+      "loss": 1.7969,
+      "step": 979
+    },
+    {
+      "epoch": 0.11385419692128958,
+      "grad_norm": 0.37571048736572266,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 980
+    },
+    {
+      "epoch": 0.11397037467325007,
+      "grad_norm": 0.3701353967189789,
+      "learning_rate": 0.0001,
+      "loss": 1.8767,
+      "step": 981
+    },
+    {
+      "epoch": 0.11408655242521057,
+      "grad_norm": 0.34536612033843994,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 982
+    },
+    {
+      "epoch": 0.11420273017717107,
+      "grad_norm": 0.43408679962158203,
+      "learning_rate": 0.0001,
+      "loss": 1.9732,
+      "step": 983
+    },
+    {
+      "epoch": 0.11431890792913157,
+      "grad_norm": 0.34038683772087097,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 984
+    },
+    {
+      "epoch": 0.11443508568109208,
+      "grad_norm": 0.3477528691291809,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 985
+    },
+    {
+      "epoch": 0.11455126343305257,
+      "grad_norm": 0.3574272096157074,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 986
+    },
+    {
+      "epoch": 0.11466744118501307,
+      "grad_norm": 0.3423398733139038,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 987
+    },
+    {
+      "epoch": 0.11478361893697357,
+      "grad_norm": 0.35041436553001404,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 988
+    },
+    {
+      "epoch": 0.11489979668893407,
+      "grad_norm": 0.3532884418964386,
+      "learning_rate": 0.0001,
+      "loss": 1.8723,
+      "step": 989
+    },
+    {
+      "epoch": 0.11501597444089456,
+      "grad_norm": 0.3406219482421875,
+      "learning_rate": 0.0001,
+      "loss": 1.6606,
+      "step": 990
+    },
+    {
+      "epoch": 0.11513215219285507,
+      "grad_norm": 0.33296358585357666,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 991
+    },
+    {
+      "epoch": 0.11524832994481557,
+      "grad_norm": 0.35060951113700867,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 992
+    },
+    {
+      "epoch": 0.11536450769677607,
+      "grad_norm": 0.369406521320343,
+      "learning_rate": 0.0001,
+      "loss": 1.7544,
+      "step": 993
+    },
+    {
+      "epoch": 0.11548068544873656,
+      "grad_norm": 0.3404718041419983,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 994
+    },
+    {
+      "epoch": 0.11559686320069706,
+      "grad_norm": 0.3357236385345459,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 995
+    },
+    {
+      "epoch": 0.11571304095265757,
+      "grad_norm": 0.3525453805923462,
+      "learning_rate": 0.0001,
+      "loss": 1.8039,
+      "step": 996
+    },
+    {
+      "epoch": 0.11582921870461807,
+      "grad_norm": 0.35840287804603577,
+      "learning_rate": 0.0001,
+      "loss": 1.8253,
+      "step": 997
+    },
+    {
+      "epoch": 0.11594539645657856,
+      "grad_norm": 0.36817336082458496,
+      "learning_rate": 0.0001,
+      "loss": 1.8989,
+      "step": 998
+    },
+    {
+      "epoch": 0.11606157420853906,
+      "grad_norm": 0.349050372838974,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 999
+    },
+    {
+      "epoch": 0.11617775196049956,
+      "grad_norm": 0.37279772758483887,
+      "learning_rate": 0.0001,
+      "loss": 1.8076,
+      "step": 1000
+    },
+    {
+      "epoch": 0.11629392971246007,
+      "grad_norm": 0.36874422430992126,
+      "learning_rate": 0.0001,
+      "loss": 1.7866,
+      "step": 1001
+    },
+    {
+      "epoch": 0.11641010746442057,
+      "grad_norm": 0.3826844394207001,
+      "learning_rate": 0.0001,
+      "loss": 1.7474,
+      "step": 1002
+    },
+    {
+      "epoch": 0.11652628521638106,
+      "grad_norm": 0.34281986951828003,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 1003
+    },
+    {
+      "epoch": 0.11664246296834156,
+      "grad_norm": 0.3320811688899994,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 1004
+    },
+    {
+      "epoch": 0.11675864072030206,
+      "grad_norm": 0.3438369631767273,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 1005
+    },
+    {
+      "epoch": 0.11687481847226257,
+      "grad_norm": 0.35468512773513794,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 1006
+    },
+    {
+      "epoch": 0.11699099622422306,
+      "grad_norm": 0.3840721547603607,
+      "learning_rate": 0.0001,
+      "loss": 1.9277,
+      "step": 1007
+    },
+    {
+      "epoch": 0.11710717397618356,
+      "grad_norm": 0.3772999346256256,
+      "learning_rate": 0.0001,
+      "loss": 1.8193,
+      "step": 1008
+    },
+    {
+      "epoch": 0.11722335172814406,
+      "grad_norm": 0.32721465826034546,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 1009
+    },
+    {
+      "epoch": 0.11733952948010457,
+      "grad_norm": 0.3579975664615631,
+      "learning_rate": 0.0001,
+      "loss": 1.8259,
+      "step": 1010
+    },
+    {
+      "epoch": 0.11745570723206505,
+      "grad_norm": 0.35289841890335083,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 1011
+    },
+    {
+      "epoch": 0.11757188498402556,
+      "grad_norm": 0.35894694924354553,
+      "learning_rate": 0.0001,
+      "loss": 1.7533,
+      "step": 1012
+    },
+    {
+      "epoch": 0.11768806273598606,
+      "grad_norm": 0.35734298825263977,
+      "learning_rate": 0.0001,
+      "loss": 1.8021,
+      "step": 1013
+    },
+    {
+      "epoch": 0.11780424048794656,
+      "grad_norm": 0.3388814330101013,
+      "learning_rate": 0.0001,
+      "loss": 1.7065,
+      "step": 1014
+    },
+    {
+      "epoch": 0.11792041823990705,
+      "grad_norm": 0.3703312277793884,
+      "learning_rate": 0.0001,
+      "loss": 1.7811,
+      "step": 1015
+    },
+    {
+      "epoch": 0.11803659599186755,
+      "grad_norm": 0.3490375280380249,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 1016
+    },
+    {
+      "epoch": 0.11815277374382806,
+      "grad_norm": 0.36077988147735596,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 1017
+    },
+    {
+      "epoch": 0.11826895149578856,
+      "grad_norm": 0.366590291261673,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 1018
+    },
+    {
+      "epoch": 0.11838512924774906,
+      "grad_norm": 0.3408995568752289,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 1019
+    },
+    {
+      "epoch": 0.11850130699970955,
+      "grad_norm": 0.3684650659561157,
+      "learning_rate": 0.0001,
+      "loss": 1.7644,
+      "step": 1020
+    },
+    {
+      "epoch": 0.11861748475167005,
+      "grad_norm": 0.37491750717163086,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 1021
+    },
+    {
+      "epoch": 0.11873366250363056,
+      "grad_norm": 0.34040531516075134,
+      "learning_rate": 0.0001,
+      "loss": 1.7995,
+      "step": 1022
+    },
+    {
+      "epoch": 0.11884984025559106,
+      "grad_norm": 0.3536158800125122,
+      "learning_rate": 0.0001,
+      "loss": 1.811,
+      "step": 1023
+    },
+    {
+      "epoch": 0.11896601800755155,
+      "grad_norm": 0.38111457228660583,
+      "learning_rate": 0.0001,
+      "loss": 1.8178,
+      "step": 1024
+    },
+    {
+      "epoch": 0.11908219575951205,
+      "grad_norm": 0.39054590463638306,
+      "learning_rate": 0.0001,
+      "loss": 1.8101,
+      "step": 1025
+    },
+    {
+      "epoch": 0.11919837351147256,
+      "grad_norm": 0.37817317247390747,
+      "learning_rate": 0.0001,
+      "loss": 1.8686,
+      "step": 1026
+    },
+    {
+      "epoch": 0.11931455126343306,
+      "grad_norm": 0.38268011808395386,
+      "learning_rate": 0.0001,
+      "loss": 1.8499,
+      "step": 1027
+    },
+    {
+      "epoch": 0.11943072901539355,
+      "grad_norm": 0.3847000002861023,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 1028
+    },
+    {
+      "epoch": 0.11954690676735405,
+      "grad_norm": 0.3735695481300354,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 1029
+    },
+    {
+      "epoch": 0.11966308451931455,
+      "grad_norm": 0.3305729031562805,
+      "learning_rate": 0.0001,
+      "loss": 1.7738,
+      "step": 1030
+    },
+    {
+      "epoch": 0.11977926227127506,
+      "grad_norm": 0.3482242822647095,
+      "learning_rate": 0.0001,
+      "loss": 1.8291,
+      "step": 1031
+    },
+    {
+      "epoch": 0.11989544002323554,
+      "grad_norm": 0.359893798828125,
+      "learning_rate": 0.0001,
+      "loss": 1.7426,
+      "step": 1032
+    },
+    {
+      "epoch": 0.12001161777519605,
+      "grad_norm": 0.36758822202682495,
+      "learning_rate": 0.0001,
+      "loss": 1.9225,
+      "step": 1033
+    },
+    {
+      "epoch": 0.12012779552715655,
+      "grad_norm": 0.3210867643356323,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 1034
+    },
+    {
+      "epoch": 0.12024397327911705,
+      "grad_norm": 0.3599686622619629,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 1035
+    },
+    {
+      "epoch": 0.12036015103107754,
+      "grad_norm": 0.3894933760166168,
+      "learning_rate": 0.0001,
+      "loss": 1.7807,
+      "step": 1036
+    },
+    {
+      "epoch": 0.12047632878303804,
+      "grad_norm": 0.3600466847419739,
+      "learning_rate": 0.0001,
+      "loss": 1.7304,
+      "step": 1037
+    },
+    {
+      "epoch": 0.12059250653499855,
+      "grad_norm": 0.3773282766342163,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 1038
+    },
+    {
+      "epoch": 0.12070868428695905,
+      "grad_norm": 0.342816561460495,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 1039
+    },
+    {
+      "epoch": 0.12082486203891955,
+      "grad_norm": 0.34455814957618713,
+      "learning_rate": 0.0001,
+      "loss": 1.8846,
+      "step": 1040
+    },
+    {
+      "epoch": 0.12094103979088004,
+      "grad_norm": 0.3425041437149048,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 1041
+    },
+    {
+      "epoch": 0.12105721754284055,
+      "grad_norm": 0.3462069034576416,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 1042
+    },
+    {
+      "epoch": 0.12117339529480105,
+      "grad_norm": 0.3301829397678375,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 1043
+    },
+    {
+      "epoch": 0.12128957304676155,
+      "grad_norm": 0.32111644744873047,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 1044
+    },
+    {
+      "epoch": 0.12140575079872204,
+      "grad_norm": 0.3677181005477905,
+      "learning_rate": 0.0001,
+      "loss": 1.7678,
+      "step": 1045
+    },
+    {
+      "epoch": 0.12152192855068254,
+      "grad_norm": 0.3687067925930023,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 1046
+    },
+    {
+      "epoch": 0.12163810630264305,
+      "grad_norm": 0.34119775891304016,
+      "learning_rate": 0.0001,
+      "loss": 1.7466,
+      "step": 1047
+    },
+    {
+      "epoch": 0.12175428405460355,
+      "grad_norm": 0.3656751811504364,
+      "learning_rate": 0.0001,
+      "loss": 1.6437,
+      "step": 1048
+    },
+    {
+      "epoch": 0.12187046180656404,
+      "grad_norm": 0.35760533809661865,
+      "learning_rate": 0.0001,
+      "loss": 1.8831,
+      "step": 1049
+    },
+    {
+      "epoch": 0.12198663955852454,
+      "grad_norm": 0.3624860942363739,
+      "learning_rate": 0.0001,
+      "loss": 1.7621,
+      "step": 1050
+    },
+    {
+      "epoch": 0.12210281731048504,
+      "grad_norm": 0.3888024091720581,
+      "learning_rate": 0.0001,
+      "loss": 1.6873,
+      "step": 1051
+    },
+    {
+      "epoch": 0.12221899506244555,
+      "grad_norm": 0.39359578490257263,
+      "learning_rate": 0.0001,
+      "loss": 1.9219,
+      "step": 1052
+    },
+    {
+      "epoch": 0.12233517281440603,
+      "grad_norm": 0.3813614249229431,
+      "learning_rate": 0.0001,
+      "loss": 1.7615,
+      "step": 1053
+    },
+    {
+      "epoch": 0.12245135056636654,
+      "grad_norm": 0.3501231074333191,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 1054
+    },
+    {
+      "epoch": 0.12256752831832704,
+      "grad_norm": 0.32617613673210144,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 1055
+    },
+    {
+      "epoch": 0.12268370607028754,
+      "grad_norm": 0.30450668931007385,
+      "learning_rate": 0.0001,
+      "loss": 1.4726,
+      "step": 1056
+    },
+    {
+      "epoch": 0.12279988382224805,
+      "grad_norm": 0.33408787846565247,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 1057
+    },
+    {
+      "epoch": 0.12291606157420853,
+      "grad_norm": 0.3771173357963562,
+      "learning_rate": 0.0001,
+      "loss": 1.8417,
+      "step": 1058
+    },
+    {
+      "epoch": 0.12303223932616904,
+      "grad_norm": 0.3441436290740967,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 1059
+    },
+    {
+      "epoch": 0.12314841707812954,
+      "grad_norm": 0.3676146864891052,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 1060
+    },
+    {
+      "epoch": 0.12326459483009004,
+      "grad_norm": 0.35352498292922974,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 1061
+    },
+    {
+      "epoch": 0.12338077258205053,
+      "grad_norm": 0.43162113428115845,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 1062
+    },
+    {
+      "epoch": 0.12349695033401104,
+      "grad_norm": 0.370995432138443,
+      "learning_rate": 0.0001,
+      "loss": 1.7928,
+      "step": 1063
+    },
+    {
+      "epoch": 0.12361312808597154,
+      "grad_norm": 0.35732775926589966,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 1064
+    },
+    {
+      "epoch": 0.12372930583793204,
+      "grad_norm": 0.40261656045913696,
+      "learning_rate": 0.0001,
+      "loss": 2.0031,
+      "step": 1065
+    },
+    {
+      "epoch": 0.12384548358989253,
+      "grad_norm": 0.34159186482429504,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 1066
+    },
+    {
+      "epoch": 0.12396166134185303,
+      "grad_norm": 0.3457348942756653,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 1067
+    },
+    {
+      "epoch": 0.12407783909381354,
+      "grad_norm": 0.36517080664634705,
+      "learning_rate": 0.0001,
+      "loss": 1.7111,
+      "step": 1068
+    },
+    {
+      "epoch": 0.12419401684577404,
+      "grad_norm": 0.3570208251476288,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 1069
+    },
+    {
+      "epoch": 0.12431019459773453,
+      "grad_norm": 0.33659952878952026,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 1070
+    },
+    {
+      "epoch": 0.12442637234969503,
+      "grad_norm": 0.3287290930747986,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 1071
+    },
+    {
+      "epoch": 0.12454255010165553,
+      "grad_norm": 0.3403220772743225,
+      "learning_rate": 0.0001,
+      "loss": 1.8731,
+      "step": 1072
+    },
+    {
+      "epoch": 0.12465872785361604,
+      "grad_norm": 0.38080114126205444,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 1073
+    },
+    {
+      "epoch": 0.12477490560557654,
+      "grad_norm": 0.3732984662055969,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 1074
+    },
+    {
+      "epoch": 0.12489108335753703,
+      "grad_norm": 0.3515911102294922,
+      "learning_rate": 0.0001,
+      "loss": 1.7473,
+      "step": 1075
+    },
+    {
+      "epoch": 0.12500726110949753,
+      "grad_norm": 0.36280357837677,
+      "learning_rate": 0.0001,
+      "loss": 1.852,
+      "step": 1076
+    },
+    {
+      "epoch": 0.12512343886145802,
+      "grad_norm": 0.4288184344768524,
+      "learning_rate": 0.0001,
+      "loss": 1.9649,
+      "step": 1077
+    },
+    {
+      "epoch": 0.12523961661341854,
+      "grad_norm": 0.3732617199420929,
+      "learning_rate": 0.0001,
+      "loss": 1.9636,
+      "step": 1078
+    },
+    {
+      "epoch": 0.12535579436537903,
+      "grad_norm": 0.3731965124607086,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 1079
+    },
+    {
+      "epoch": 0.12547197211733954,
+      "grad_norm": 0.33766472339630127,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 1080
+    },
+    {
+      "epoch": 0.12558814986930003,
+      "grad_norm": 0.41299498081207275,
+      "learning_rate": 0.0001,
+      "loss": 2.008,
+      "step": 1081
+    },
+    {
+      "epoch": 0.12570432762126052,
+      "grad_norm": 0.332131952047348,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 1082
+    },
+    {
+      "epoch": 0.12582050537322104,
+      "grad_norm": 0.3629733920097351,
+      "learning_rate": 0.0001,
+      "loss": 1.8256,
+      "step": 1083
+    },
+    {
+      "epoch": 0.12593668312518153,
+      "grad_norm": 0.35666128993034363,
+      "learning_rate": 0.0001,
+      "loss": 1.6795,
+      "step": 1084
+    },
+    {
+      "epoch": 0.12605286087714201,
+      "grad_norm": 0.3693120777606964,
+      "learning_rate": 0.0001,
+      "loss": 1.8276,
+      "step": 1085
+    },
+    {
+      "epoch": 0.12616903862910253,
+      "grad_norm": 0.38100001215934753,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 1086
+    },
+    {
+      "epoch": 0.12628521638106302,
+      "grad_norm": 0.35735470056533813,
+      "learning_rate": 0.0001,
+      "loss": 1.8506,
+      "step": 1087
+    },
+    {
+      "epoch": 0.12640139413302354,
+      "grad_norm": 0.3401797413825989,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 1088
+    },
+    {
+      "epoch": 0.12651757188498403,
+      "grad_norm": 0.34798330068588257,
+      "learning_rate": 0.0001,
+      "loss": 1.8221,
+      "step": 1089
+    },
+    {
+      "epoch": 0.12663374963694451,
+      "grad_norm": 0.3499447703361511,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 1090
+    },
+    {
+      "epoch": 0.12674992738890503,
+      "grad_norm": 0.3714812099933624,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 1091
+    },
+    {
+      "epoch": 0.12686610514086552,
+      "grad_norm": 0.3492056429386139,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 1092
+    },
+    {
+      "epoch": 0.12698228289282604,
+      "grad_norm": 0.3401804566383362,
+      "learning_rate": 0.0001,
+      "loss": 1.7797,
+      "step": 1093
+    },
+    {
+      "epoch": 0.12709846064478653,
+      "grad_norm": 0.33436286449432373,
+      "learning_rate": 0.0001,
+      "loss": 1.7528,
+      "step": 1094
+    },
+    {
+      "epoch": 0.12721463839674702,
+      "grad_norm": 0.35085633397102356,
+      "learning_rate": 0.0001,
+      "loss": 1.7508,
+      "step": 1095
+    },
+    {
+      "epoch": 0.12733081614870753,
+      "grad_norm": 0.3480139970779419,
+      "learning_rate": 0.0001,
+      "loss": 1.7582,
+      "step": 1096
+    },
+    {
+      "epoch": 0.12744699390066802,
+      "grad_norm": 0.3390267789363861,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 1097
+    },
+    {
+      "epoch": 0.1275631716526285,
+      "grad_norm": 0.3470866084098816,
+      "learning_rate": 0.0001,
+      "loss": 1.7661,
+      "step": 1098
+    },
+    {
+      "epoch": 0.12767934940458903,
+      "grad_norm": 0.37509381771087646,
+      "learning_rate": 0.0001,
+      "loss": 1.761,
+      "step": 1099
+    },
+    {
+      "epoch": 0.12779552715654952,
+      "grad_norm": 0.35059359669685364,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 1100
+    },
+    {
+      "epoch": 0.12791170490851003,
+      "grad_norm": 0.38559988141059875,
+      "learning_rate": 0.0001,
+      "loss": 1.9108,
+      "step": 1101
+    },
+    {
+      "epoch": 0.12802788266047052,
+      "grad_norm": 0.3470633924007416,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 1102
+    },
+    {
+      "epoch": 0.128144060412431,
+      "grad_norm": 0.3547315001487732,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 1103
+    },
+    {
+      "epoch": 0.12826023816439153,
+      "grad_norm": 0.3657218813896179,
+      "learning_rate": 0.0001,
+      "loss": 1.8437,
+      "step": 1104
+    },
+    {
+      "epoch": 0.12837641591635202,
+      "grad_norm": 0.3548438847064972,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 1105
+    },
+    {
+      "epoch": 0.1284925936683125,
+      "grad_norm": 0.3553130030632019,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 1106
+    },
+    {
+      "epoch": 0.12860877142027302,
+      "grad_norm": 0.35484910011291504,
+      "learning_rate": 0.0001,
+      "loss": 1.7819,
+      "step": 1107
+    },
+    {
+      "epoch": 0.1287249491722335,
+      "grad_norm": 0.3301905691623688,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 1108
+    },
+    {
+      "epoch": 0.12884112692419403,
+      "grad_norm": 0.3722817301750183,
+      "learning_rate": 0.0001,
+      "loss": 1.8194,
+      "step": 1109
+    },
+    {
+      "epoch": 0.12895730467615452,
+      "grad_norm": 0.408559113740921,
+      "learning_rate": 0.0001,
+      "loss": 1.8952,
+      "step": 1110
+    },
+    {
+      "epoch": 0.129073482428115,
+      "grad_norm": 0.3764549791812897,
+      "learning_rate": 0.0001,
+      "loss": 1.8921,
+      "step": 1111
+    },
+    {
+      "epoch": 0.12918966018007552,
+      "grad_norm": 0.3443762958049774,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 1112
+    },
+    {
+      "epoch": 0.129305837932036,
+      "grad_norm": 0.37455350160598755,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 1113
+    },
+    {
+      "epoch": 0.12942201568399653,
+      "grad_norm": 0.35170894861221313,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 1114
+    },
+    {
+      "epoch": 0.12953819343595702,
+      "grad_norm": 0.33991748094558716,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1296543711879175,
+      "grad_norm": 0.3735693693161011,
+      "learning_rate": 0.0001,
+      "loss": 1.928,
+      "step": 1116
+    },
+    {
+      "epoch": 0.12977054893987802,
+      "grad_norm": 0.34423983097076416,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 1117
+    },
+    {
+      "epoch": 0.1298867266918385,
+      "grad_norm": 0.3688075542449951,
+      "learning_rate": 0.0001,
+      "loss": 1.8482,
+      "step": 1118
+    },
+    {
+      "epoch": 0.130002904443799,
+      "grad_norm": 0.3608585000038147,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 1119
+    },
+    {
+      "epoch": 0.13011908219575952,
+      "grad_norm": 0.38905078172683716,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 1120
+    },
+    {
+      "epoch": 0.13023525994772,
+      "grad_norm": 0.3615328371524811,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 1121
+    },
+    {
+      "epoch": 0.13035143769968052,
+      "grad_norm": 0.35324275493621826,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 1122
+    },
+    {
+      "epoch": 0.130467615451641,
+      "grad_norm": 0.3279106318950653,
+      "learning_rate": 0.0001,
+      "loss": 1.6848,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1305837932036015,
+      "grad_norm": 0.37770041823387146,
+      "learning_rate": 0.0001,
+      "loss": 1.9185,
+      "step": 1124
+    },
+    {
+      "epoch": 0.13069997095556202,
+      "grad_norm": 0.3365723192691803,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1308161487075225,
+      "grad_norm": 0.35211873054504395,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 1126
+    },
+    {
+      "epoch": 0.13093232645948302,
+      "grad_norm": 0.3364546000957489,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1310485042114435,
+      "grad_norm": 0.33624467253685,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 1128
+    },
+    {
+      "epoch": 0.131164681963404,
+      "grad_norm": 0.34484627842903137,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 1129
+    },
+    {
+      "epoch": 0.13128085971536452,
+      "grad_norm": 0.3245522677898407,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 1130
+    },
+    {
+      "epoch": 0.131397037467325,
+      "grad_norm": 0.3584153652191162,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 1131
+    },
+    {
+      "epoch": 0.1315132152192855,
+      "grad_norm": 0.35809946060180664,
+      "learning_rate": 0.0001,
+      "loss": 1.7665,
+      "step": 1132
+    },
+    {
+      "epoch": 0.131629392971246,
+      "grad_norm": 0.3418474793434143,
+      "learning_rate": 0.0001,
+      "loss": 1.6811,
+      "step": 1133
+    },
+    {
+      "epoch": 0.1317455707232065,
+      "grad_norm": 0.3623388707637787,
+      "learning_rate": 0.0001,
+      "loss": 1.7171,
+      "step": 1134
+    },
+    {
+      "epoch": 0.13186174847516702,
+      "grad_norm": 0.3962494730949402,
+      "learning_rate": 0.0001,
+      "loss": 1.9956,
+      "step": 1135
+    },
+    {
+      "epoch": 0.1319779262271275,
+      "grad_norm": 0.34545156359672546,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 1136
+    },
+    {
+      "epoch": 0.132094103979088,
+      "grad_norm": 0.3415702283382416,
+      "learning_rate": 0.0001,
+      "loss": 1.7397,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1322102817310485,
+      "grad_norm": 0.41906920075416565,
+      "learning_rate": 0.0001,
+      "loss": 1.8653,
+      "step": 1138
+    },
+    {
+      "epoch": 0.132326459483009,
+      "grad_norm": 0.3778826892375946,
+      "learning_rate": 0.0001,
+      "loss": 1.696,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1324426372349695,
+      "grad_norm": 0.3520076870918274,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 1140
+    },
+    {
+      "epoch": 0.13255881498693,
+      "grad_norm": 0.38221481442451477,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 1141
+    },
+    {
+      "epoch": 0.1326749927388905,
+      "grad_norm": 0.34587883949279785,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 1142
+    },
+    {
+      "epoch": 0.132791170490851,
+      "grad_norm": 0.35534965991973877,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1329073482428115,
+      "grad_norm": 0.36971423029899597,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 1144
+    },
+    {
+      "epoch": 0.133023525994772,
+      "grad_norm": 0.3652137219905853,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 1145
+    },
+    {
+      "epoch": 0.1331397037467325,
+      "grad_norm": 0.340309739112854,
+      "learning_rate": 0.0001,
+      "loss": 1.75,
+      "step": 1146
+    },
+    {
+      "epoch": 0.133255881498693,
+      "grad_norm": 0.3661729693412781,
+      "learning_rate": 0.0001,
+      "loss": 1.7802,
+      "step": 1147
+    },
+    {
+      "epoch": 0.1333720592506535,
+      "grad_norm": 0.3509800434112549,
+      "learning_rate": 0.0001,
+      "loss": 1.7748,
+      "step": 1148
+    },
+    {
+      "epoch": 0.133488237002614,
+      "grad_norm": 0.34903261065483093,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 1149
+    },
+    {
+      "epoch": 0.1336044147545745,
+      "grad_norm": 0.3474218249320984,
+      "learning_rate": 0.0001,
+      "loss": 1.8942,
+      "step": 1150
+    },
+    {
+      "epoch": 0.133720592506535,
+      "grad_norm": 0.3618185818195343,
+      "learning_rate": 0.0001,
+      "loss": 1.7351,
+      "step": 1151
+    },
+    {
+      "epoch": 0.1338367702584955,
+      "grad_norm": 0.3447827696800232,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 1152
+    },
+    {
+      "epoch": 0.13395294801045599,
+      "grad_norm": 0.35929834842681885,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1340691257624165,
+      "grad_norm": 0.378379762172699,
+      "learning_rate": 0.0001,
+      "loss": 1.8789,
+      "step": 1154
+    },
+    {
+      "epoch": 0.134185303514377,
+      "grad_norm": 0.3777737319469452,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 1155
+    },
+    {
+      "epoch": 0.1343014812663375,
+      "grad_norm": 0.542852520942688,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 1156
+    },
+    {
+      "epoch": 0.134417659018298,
+      "grad_norm": 0.3426233232021332,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 1157
+    },
+    {
+      "epoch": 0.13453383677025849,
+      "grad_norm": 0.34573081135749817,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 1158
+    },
+    {
+      "epoch": 0.134650014522219,
+      "grad_norm": 0.35474127531051636,
+      "learning_rate": 0.0001,
+      "loss": 1.7912,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1347661922741795,
+      "grad_norm": 0.3295106589794159,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 1160
+    },
+    {
+      "epoch": 0.13488237002613998,
+      "grad_norm": 0.380728542804718,
+      "learning_rate": 0.0001,
+      "loss": 1.8133,
+      "step": 1161
+    },
+    {
+      "epoch": 0.1349985477781005,
+      "grad_norm": 0.38181304931640625,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 1162
+    },
+    {
+      "epoch": 0.135114725530061,
+      "grad_norm": 0.35895970463752747,
+      "learning_rate": 0.0001,
+      "loss": 1.6405,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1352309032820215,
+      "grad_norm": 0.3574581742286682,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 1164
+    },
+    {
+      "epoch": 0.135347081033982,
+      "grad_norm": 0.3409847319126129,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 1165
+    },
+    {
+      "epoch": 0.13546325878594248,
+      "grad_norm": 0.36997130513191223,
+      "learning_rate": 0.0001,
+      "loss": 1.8488,
+      "step": 1166
+    },
+    {
+      "epoch": 0.135579436537903,
+      "grad_norm": 0.34550440311431885,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 1167
+    },
+    {
+      "epoch": 0.1356956142898635,
+      "grad_norm": 0.36882877349853516,
+      "learning_rate": 0.0001,
+      "loss": 1.7311,
+      "step": 1168
+    },
+    {
+      "epoch": 0.135811792041824,
+      "grad_norm": 0.34376415610313416,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1359279697937845,
+      "grad_norm": 0.35478463768959045,
+      "learning_rate": 0.0001,
+      "loss": 1.836,
+      "step": 1170
+    },
+    {
+      "epoch": 0.13604414754574498,
+      "grad_norm": 0.3668883442878723,
+      "learning_rate": 0.0001,
+      "loss": 1.8465,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1361603252977055,
+      "grad_norm": 0.33655285835266113,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 1172
+    },
+    {
+      "epoch": 0.136276503049666,
+      "grad_norm": 0.34968823194503784,
+      "learning_rate": 0.0001,
+      "loss": 1.6423,
+      "step": 1173
+    },
+    {
+      "epoch": 0.13639268080162648,
+      "grad_norm": 0.37849414348602295,
+      "learning_rate": 0.0001,
+      "loss": 1.832,
+      "step": 1174
+    },
+    {
+      "epoch": 0.136508858553587,
+      "grad_norm": 0.3569866120815277,
+      "learning_rate": 0.0001,
+      "loss": 1.7077,
+      "step": 1175
+    },
+    {
+      "epoch": 0.13662503630554748,
+      "grad_norm": 0.34238025546073914,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 1176
+    },
+    {
+      "epoch": 0.136741214057508,
+      "grad_norm": 0.35172078013420105,
+      "learning_rate": 0.0001,
+      "loss": 1.7818,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1368573918094685,
+      "grad_norm": 0.35582435131073,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 1178
+    },
+    {
+      "epoch": 0.13697356956142898,
+      "grad_norm": 0.36293935775756836,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1370897473133895,
+      "grad_norm": 0.4134126901626587,
+      "learning_rate": 0.0001,
+      "loss": 1.8926,
+      "step": 1180
+    },
+    {
+      "epoch": 0.13720592506534998,
+      "grad_norm": 0.34876781702041626,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1373221028173105,
+      "grad_norm": 0.35906705260276794,
+      "learning_rate": 0.0001,
+      "loss": 1.8285,
+      "step": 1182
+    },
+    {
+      "epoch": 0.137438280569271,
+      "grad_norm": 0.3706406354904175,
+      "learning_rate": 0.0001,
+      "loss": 1.8965,
+      "step": 1183
+    },
+    {
+      "epoch": 0.13755445832123148,
+      "grad_norm": 0.3602171242237091,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 1184
+    },
+    {
+      "epoch": 0.137670636073192,
+      "grad_norm": 0.3797997534275055,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 1185
+    },
+    {
+      "epoch": 0.13778681382515248,
+      "grad_norm": 0.3507281541824341,
+      "learning_rate": 0.0001,
+      "loss": 1.7615,
+      "step": 1186
+    },
+    {
+      "epoch": 0.13790299157711297,
+      "grad_norm": 0.3676402270793915,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 1187
+    },
+    {
+      "epoch": 0.1380191693290735,
+      "grad_norm": 0.36422815918922424,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 1188
+    },
+    {
+      "epoch": 0.13813534708103398,
+      "grad_norm": 0.3710031807422638,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1382515248329945,
+      "grad_norm": 0.34949737787246704,
+      "learning_rate": 0.0001,
+      "loss": 1.6865,
+      "step": 1190
+    },
+    {
+      "epoch": 0.13836770258495498,
+      "grad_norm": 0.38215914368629456,
+      "learning_rate": 0.0001,
+      "loss": 1.9185,
+      "step": 1191
+    },
+    {
+      "epoch": 0.13848388033691547,
+      "grad_norm": 0.3533918857574463,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 1192
+    },
+    {
+      "epoch": 0.138600058088876,
+      "grad_norm": 0.36544063687324524,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 1193
+    },
+    {
+      "epoch": 0.13871623584083648,
+      "grad_norm": 0.3464662730693817,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13883241359279697,
+      "grad_norm": 0.32799604535102844,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13894859134475748,
+      "grad_norm": 0.37455034255981445,
+      "learning_rate": 0.0001,
+      "loss": 1.895,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13906476909671797,
+      "grad_norm": 0.3819703459739685,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1391809468486785,
+      "grad_norm": 0.33819907903671265,
+      "learning_rate": 0.0001,
+      "loss": 1.7259,
+      "step": 1198
+    },
+    {
+      "epoch": 0.13929712460063898,
+      "grad_norm": 0.36918461322784424,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 1199
+    },
+    {
+      "epoch": 0.13941330235259947,
+      "grad_norm": 0.37889495491981506,
+      "learning_rate": 0.0001,
+      "loss": 1.8902,
+      "step": 1200
+    },
+    {
+      "epoch": 0.13952948010455998,
+      "grad_norm": 0.34494003653526306,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 1201
+    },
+    {
+      "epoch": 0.13964565785652047,
+      "grad_norm": 0.3527129292488098,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 1202
+    },
+    {
+      "epoch": 0.139761835608481,
+      "grad_norm": 0.35137444734573364,
+      "learning_rate": 0.0001,
+      "loss": 1.8853,
+      "step": 1203
+    },
+    {
+      "epoch": 0.13987801336044148,
+      "grad_norm": 0.3461925685405731,
+      "learning_rate": 0.0001,
+      "loss": 1.7065,
+      "step": 1204
+    },
+    {
+      "epoch": 0.13999419111240197,
+      "grad_norm": 0.36129575967788696,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 1205
+    },
+    {
+      "epoch": 0.14011036886436248,
+      "grad_norm": 0.34536105394363403,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 1206
+    },
+    {
+      "epoch": 0.14022654661632297,
+      "grad_norm": 0.35485297441482544,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 1207
+    },
+    {
+      "epoch": 0.14034272436828346,
+      "grad_norm": 0.3653600811958313,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 1208
+    },
+    {
+      "epoch": 0.14045890212024398,
+      "grad_norm": 0.36751556396484375,
+      "learning_rate": 0.0001,
+      "loss": 1.9142,
+      "step": 1209
+    },
+    {
+      "epoch": 0.14057507987220447,
+      "grad_norm": 0.3849860727787018,
+      "learning_rate": 0.0001,
+      "loss": 1.7801,
+      "step": 1210
+    },
+    {
+      "epoch": 0.14069125762416498,
+      "grad_norm": 0.3888527452945709,
+      "learning_rate": 0.0001,
+      "loss": 1.7732,
+      "step": 1211
+    },
+    {
+      "epoch": 0.14080743537612547,
+      "grad_norm": 0.3473045825958252,
+      "learning_rate": 0.0001,
+      "loss": 1.7279,
+      "step": 1212
+    },
+    {
+      "epoch": 0.14092361312808596,
+      "grad_norm": 0.31697720289230347,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 1213
+    },
+    {
+      "epoch": 0.14103979088004648,
+      "grad_norm": 0.36347395181655884,
+      "learning_rate": 0.0001,
+      "loss": 1.8358,
+      "step": 1214
+    },
+    {
+      "epoch": 0.14115596863200697,
+      "grad_norm": 0.34787067770957947,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 1215
+    },
+    {
+      "epoch": 0.14127214638396748,
+      "grad_norm": 0.35520100593566895,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 1216
+    },
+    {
+      "epoch": 0.14138832413592797,
+      "grad_norm": 0.3633323609828949,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 1217
+    },
+    {
+      "epoch": 0.14150450188788846,
+      "grad_norm": 0.3367605209350586,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 1218
+    },
+    {
+      "epoch": 0.14162067963984898,
+      "grad_norm": 0.36629173159599304,
+      "learning_rate": 0.0001,
+      "loss": 1.8747,
+      "step": 1219
+    },
+    {
+      "epoch": 0.14173685739180947,
+      "grad_norm": 0.3563576340675354,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 1220
+    },
+    {
+      "epoch": 0.14185303514376996,
+      "grad_norm": 0.32516106963157654,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 1221
+    },
+    {
+      "epoch": 0.14196921289573047,
+      "grad_norm": 0.3476703464984894,
+      "learning_rate": 0.0001,
+      "loss": 1.7661,
+      "step": 1222
+    },
+    {
+      "epoch": 0.14208539064769096,
+      "grad_norm": 0.3780508041381836,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 1223
+    },
+    {
+      "epoch": 0.14220156839965148,
+      "grad_norm": 0.3513147234916687,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 1224
+    },
+    {
+      "epoch": 0.14231774615161197,
+      "grad_norm": 0.384937584400177,
+      "learning_rate": 0.0001,
+      "loss": 1.8593,
+      "step": 1225
+    },
+    {
+      "epoch": 0.14243392390357246,
+      "grad_norm": 0.37124475836753845,
+      "learning_rate": 0.0001,
+      "loss": 1.8555,
+      "step": 1226
+    },
+    {
+      "epoch": 0.14255010165553297,
+      "grad_norm": 0.3864074647426605,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 1227
+    },
+    {
+      "epoch": 0.14266627940749346,
+      "grad_norm": 0.34239131212234497,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 1228
+    },
+    {
+      "epoch": 0.14278245715945395,
+      "grad_norm": 0.3898089528083801,
+      "learning_rate": 0.0001,
+      "loss": 1.8215,
+      "step": 1229
+    },
+    {
+      "epoch": 0.14289863491141447,
+      "grad_norm": 0.3413831889629364,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 1230
+    },
+    {
+      "epoch": 0.14301481266337496,
+      "grad_norm": 0.34374555945396423,
+      "learning_rate": 0.0001,
+      "loss": 1.7268,
+      "step": 1231
+    },
+    {
+      "epoch": 0.14313099041533547,
+      "grad_norm": 0.3335866928100586,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 1232
+    },
+    {
+      "epoch": 0.14324716816729596,
+      "grad_norm": 0.34190499782562256,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 1233
+    },
+    {
+      "epoch": 0.14336334591925645,
+      "grad_norm": 0.3815557062625885,
+      "learning_rate": 0.0001,
+      "loss": 1.9198,
+      "step": 1234
+    },
+    {
+      "epoch": 0.14347952367121697,
+      "grad_norm": 0.3535671830177307,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 1235
+    },
+    {
+      "epoch": 0.14359570142317746,
+      "grad_norm": 0.38413700461387634,
+      "learning_rate": 0.0001,
+      "loss": 1.8972,
+      "step": 1236
+    },
+    {
+      "epoch": 0.14371187917513797,
+      "grad_norm": 0.3618411719799042,
+      "learning_rate": 0.0001,
+      "loss": 1.6467,
+      "step": 1237
+    },
+    {
+      "epoch": 0.14382805692709846,
+      "grad_norm": 0.34210920333862305,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 1238
+    },
+    {
+      "epoch": 0.14394423467905895,
+      "grad_norm": 0.3580038249492645,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 1239
+    },
+    {
+      "epoch": 0.14406041243101947,
+      "grad_norm": 0.36096274852752686,
+      "learning_rate": 0.0001,
+      "loss": 1.7914,
+      "step": 1240
+    },
+    {
+      "epoch": 0.14417659018297996,
+      "grad_norm": 0.3829607665538788,
+      "learning_rate": 0.0001,
+      "loss": 1.8332,
+      "step": 1241
+    },
+    {
+      "epoch": 0.14429276793494045,
+      "grad_norm": 0.3591415286064148,
+      "learning_rate": 0.0001,
+      "loss": 1.7212,
+      "step": 1242
+    },
+    {
+      "epoch": 0.14440894568690096,
+      "grad_norm": 0.36007246375083923,
+      "learning_rate": 0.0001,
+      "loss": 1.7336,
+      "step": 1243
+    },
+    {
+      "epoch": 0.14452512343886145,
+      "grad_norm": 0.3532988131046295,
+      "learning_rate": 0.0001,
+      "loss": 1.8504,
+      "step": 1244
+    },
+    {
+      "epoch": 0.14464130119082197,
+      "grad_norm": 0.37223345041275024,
+      "learning_rate": 0.0001,
+      "loss": 1.7329,
+      "step": 1245
+    },
+    {
+      "epoch": 0.14475747894278246,
+      "grad_norm": 0.34656253457069397,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 1246
+    },
+    {
+      "epoch": 0.14487365669474295,
+      "grad_norm": 0.3697267472743988,
+      "learning_rate": 0.0001,
+      "loss": 1.9745,
+      "step": 1247
+    },
+    {
+      "epoch": 0.14498983444670346,
+      "grad_norm": 0.3211793899536133,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 1248
+    },
+    {
+      "epoch": 0.14510601219866395,
+      "grad_norm": 0.350603848695755,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 1249
+    },
+    {
+      "epoch": 0.14522218995062444,
+      "grad_norm": 0.3709065318107605,
+      "learning_rate": 0.0001,
+      "loss": 1.7782,
+      "step": 1250
+    },
+    {
+      "epoch": 0.14533836770258496,
+      "grad_norm": 0.367072194814682,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 1251
+    },
+    {
+      "epoch": 0.14545454545454545,
+      "grad_norm": 0.35711991786956787,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 1252
+    },
+    {
+      "epoch": 0.14557072320650596,
+      "grad_norm": 0.37235334515571594,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 1253
+    },
+    {
+      "epoch": 0.14568690095846645,
+      "grad_norm": 0.36526671051979065,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 1254
+    },
+    {
+      "epoch": 0.14580307871042694,
+      "grad_norm": 0.35159561038017273,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 1255
+    },
+    {
+      "epoch": 0.14591925646238746,
+      "grad_norm": 0.35681840777397156,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 1256
+    },
+    {
+      "epoch": 0.14603543421434795,
+      "grad_norm": 0.3531685471534729,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 1257
+    },
+    {
+      "epoch": 0.14615161196630846,
+      "grad_norm": 0.3450145423412323,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 1258
+    },
+    {
+      "epoch": 0.14626778971826895,
+      "grad_norm": 0.3472040891647339,
+      "learning_rate": 0.0001,
+      "loss": 1.7006,
+      "step": 1259
+    },
+    {
+      "epoch": 0.14638396747022944,
+      "grad_norm": 0.35504022240638733,
+      "learning_rate": 0.0001,
+      "loss": 1.8701,
+      "step": 1260
+    },
+    {
+      "epoch": 0.14650014522218996,
+      "grad_norm": 0.3664184808731079,
+      "learning_rate": 0.0001,
+      "loss": 1.7835,
+      "step": 1261
+    },
+    {
+      "epoch": 0.14661632297415045,
+      "grad_norm": 0.32701820135116577,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 1262
+    },
+    {
+      "epoch": 0.14673250072611094,
+      "grad_norm": 0.34154212474823,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 1263
+    },
+    {
+      "epoch": 0.14684867847807145,
+      "grad_norm": 0.3491269052028656,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 1264
+    },
+    {
+      "epoch": 0.14696485623003194,
+      "grad_norm": 0.351469486951828,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 1265
+    },
+    {
+      "epoch": 0.14708103398199246,
+      "grad_norm": 0.3573369085788727,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 1266
+    },
+    {
+      "epoch": 0.14719721173395295,
+      "grad_norm": 0.35862183570861816,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 1267
+    },
+    {
+      "epoch": 0.14731338948591344,
+      "grad_norm": 0.3771260976791382,
+      "learning_rate": 0.0001,
+      "loss": 1.8338,
+      "step": 1268
+    },
+    {
+      "epoch": 0.14742956723787395,
+      "grad_norm": 0.36234867572784424,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 1269
+    },
+    {
+      "epoch": 0.14754574498983444,
+      "grad_norm": 0.35228458046913147,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 1270
+    },
+    {
+      "epoch": 0.14766192274179496,
+      "grad_norm": 0.3665032982826233,
+      "learning_rate": 0.0001,
+      "loss": 1.772,
+      "step": 1271
+    },
+    {
+      "epoch": 0.14777810049375545,
+      "grad_norm": 0.3650833070278168,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 1272
+    },
+    {
+      "epoch": 0.14789427824571594,
+      "grad_norm": 0.36604440212249756,
+      "learning_rate": 0.0001,
+      "loss": 1.648,
+      "step": 1273
+    },
+    {
+      "epoch": 0.14801045599767645,
+      "grad_norm": 0.360029935836792,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 1274
+    },
+    {
+      "epoch": 0.14812663374963694,
+      "grad_norm": 0.3815794885158539,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 1275
+    },
+    {
+      "epoch": 0.14824281150159743,
+      "grad_norm": 0.36898788809776306,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 1276
+    },
+    {
+      "epoch": 0.14835898925355795,
+      "grad_norm": 0.3591456115245819,
+      "learning_rate": 0.0001,
+      "loss": 1.8453,
+      "step": 1277
+    },
+    {
+      "epoch": 0.14847516700551844,
+      "grad_norm": 0.3744908273220062,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 1278
+    },
+    {
+      "epoch": 0.14859134475747895,
+      "grad_norm": 0.3827156722545624,
+      "learning_rate": 0.0001,
+      "loss": 1.837,
+      "step": 1279
+    },
+    {
+      "epoch": 0.14870752250943944,
+      "grad_norm": 0.34295526146888733,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 1280
+    },
+    {
+      "epoch": 0.14882370026139993,
+      "grad_norm": 0.36582034826278687,
+      "learning_rate": 0.0001,
+      "loss": 1.8293,
+      "step": 1281
+    },
+    {
+      "epoch": 0.14893987801336045,
+      "grad_norm": 0.34312430024147034,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 1282
+    },
+    {
+      "epoch": 0.14905605576532094,
+      "grad_norm": 0.3591785728931427,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 1283
+    },
+    {
+      "epoch": 0.14917223351728143,
+      "grad_norm": 0.3512047231197357,
+      "learning_rate": 0.0001,
+      "loss": 1.7607,
+      "step": 1284
+    },
+    {
+      "epoch": 0.14928841126924194,
+      "grad_norm": 0.35547614097595215,
+      "learning_rate": 0.0001,
+      "loss": 1.8423,
+      "step": 1285
+    },
+    {
+      "epoch": 0.14940458902120243,
+      "grad_norm": 0.3671259582042694,
+      "learning_rate": 0.0001,
+      "loss": 1.7298,
+      "step": 1286
+    },
+    {
+      "epoch": 0.14952076677316295,
+      "grad_norm": 0.3785369098186493,
+      "learning_rate": 0.0001,
+      "loss": 1.7921,
+      "step": 1287
+    },
+    {
+      "epoch": 0.14963694452512344,
+      "grad_norm": 0.3537178635597229,
+      "learning_rate": 0.0001,
+      "loss": 1.6989,
+      "step": 1288
+    },
+    {
+      "epoch": 0.14975312227708393,
+      "grad_norm": 0.35491520166397095,
+      "learning_rate": 0.0001,
+      "loss": 1.7895,
+      "step": 1289
+    },
+    {
+      "epoch": 0.14986930002904444,
+      "grad_norm": 0.3567078709602356,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 1290
+    },
+    {
+      "epoch": 0.14998547778100493,
+      "grad_norm": 0.35973167419433594,
+      "learning_rate": 0.0001,
+      "loss": 1.8044,
+      "step": 1291
+    },
+    {
+      "epoch": 0.15010165553296545,
+      "grad_norm": 0.3756448030471802,
+      "learning_rate": 0.0001,
+      "loss": 1.9249,
+      "step": 1292
+    },
+    {
+      "epoch": 0.15021783328492594,
+      "grad_norm": 0.3457126021385193,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 1293
+    },
+    {
+      "epoch": 0.15033401103688643,
+      "grad_norm": 0.3493401110172272,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 1294
+    },
+    {
+      "epoch": 0.15045018878884694,
+      "grad_norm": 0.38052037358283997,
+      "learning_rate": 0.0001,
+      "loss": 1.8351,
+      "step": 1295
+    },
+    {
+      "epoch": 0.15056636654080743,
+      "grad_norm": 0.326164186000824,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 1296
+    },
+    {
+      "epoch": 0.15068254429276792,
+      "grad_norm": 0.37613871693611145,
+      "learning_rate": 0.0001,
+      "loss": 1.7088,
+      "step": 1297
+    },
+    {
+      "epoch": 0.15079872204472844,
+      "grad_norm": 0.40426647663116455,
+      "learning_rate": 0.0001,
+      "loss": 1.8863,
+      "step": 1298
+    },
+    {
+      "epoch": 0.15091489979668893,
+      "grad_norm": 0.3410615026950836,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 1299
+    },
+    {
+      "epoch": 0.15103107754864945,
+      "grad_norm": 0.36211374402046204,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 1300
+    },
+    {
+      "epoch": 0.15114725530060993,
+      "grad_norm": 0.39096057415008545,
+      "learning_rate": 0.0001,
+      "loss": 1.8656,
+      "step": 1301
+    },
+    {
+      "epoch": 0.15126343305257042,
+      "grad_norm": 0.3603553771972656,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 1302
+    },
+    {
+      "epoch": 0.15137961080453094,
+      "grad_norm": 0.3809981048107147,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 1303
+    },
+    {
+      "epoch": 0.15149578855649143,
+      "grad_norm": 0.35240018367767334,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 1304
+    },
+    {
+      "epoch": 0.15161196630845195,
+      "grad_norm": 0.3340260088443756,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 1305
+    },
+    {
+      "epoch": 0.15172814406041243,
+      "grad_norm": 0.3688841164112091,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 1306
+    },
+    {
+      "epoch": 0.15184432181237292,
+      "grad_norm": 0.36552947759628296,
+      "learning_rate": 0.0001,
+      "loss": 1.6414,
+      "step": 1307
+    },
+    {
+      "epoch": 0.15196049956433344,
+      "grad_norm": 0.3461878001689911,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 1308
+    },
+    {
+      "epoch": 0.15207667731629393,
+      "grad_norm": 0.35376498103141785,
+      "learning_rate": 0.0001,
+      "loss": 1.7873,
+      "step": 1309
+    },
+    {
+      "epoch": 0.15219285506825442,
+      "grad_norm": 0.37778156995773315,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 1310
+    },
+    {
+      "epoch": 0.15230903282021493,
+      "grad_norm": 0.3635299503803253,
+      "learning_rate": 0.0001,
+      "loss": 1.7731,
+      "step": 1311
+    },
+    {
+      "epoch": 0.15242521057217542,
+      "grad_norm": 0.37583014369010925,
+      "learning_rate": 0.0001,
+      "loss": 1.8543,
+      "step": 1312
+    },
+    {
+      "epoch": 0.15254138832413594,
+      "grad_norm": 0.37398314476013184,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 1313
+    },
+    {
+      "epoch": 0.15265756607609643,
+      "grad_norm": 0.37770211696624756,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 1314
+    },
+    {
+      "epoch": 0.15277374382805692,
+      "grad_norm": 0.3403077721595764,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 1315
+    },
+    {
+      "epoch": 0.15288992158001743,
+      "grad_norm": 0.3504425287246704,
+      "learning_rate": 0.0001,
+      "loss": 1.7012,
+      "step": 1316
+    },
+    {
+      "epoch": 0.15300609933197792,
+      "grad_norm": 0.3553392291069031,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1531222770839384,
+      "grad_norm": 0.3788677752017975,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 1318
+    },
+    {
+      "epoch": 0.15323845483589893,
+      "grad_norm": 0.3600800037384033,
+      "learning_rate": 0.0001,
+      "loss": 1.7677,
+      "step": 1319
+    },
+    {
+      "epoch": 0.15335463258785942,
+      "grad_norm": 0.38489770889282227,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 1320
+    },
+    {
+      "epoch": 0.15347081033981994,
+      "grad_norm": 0.3559502959251404,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 1321
+    },
+    {
+      "epoch": 0.15358698809178042,
+      "grad_norm": 0.3516106605529785,
+      "learning_rate": 0.0001,
+      "loss": 1.7595,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1537031658437409,
+      "grad_norm": 0.35282179713249207,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 1323
+    },
+    {
+      "epoch": 0.15381934359570143,
+      "grad_norm": 0.3448900878429413,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 1324
+    },
+    {
+      "epoch": 0.15393552134766192,
+      "grad_norm": 0.3468807339668274,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 1325
+    },
+    {
+      "epoch": 0.15405169909962244,
+      "grad_norm": 0.3733367621898651,
+      "learning_rate": 0.0001,
+      "loss": 1.7449,
+      "step": 1326
+    },
+    {
+      "epoch": 0.15416787685158292,
+      "grad_norm": 0.3257894217967987,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 1327
+    },
+    {
+      "epoch": 0.1542840546035434,
+      "grad_norm": 0.3381756842136383,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 1328
+    },
+    {
+      "epoch": 0.15440023235550393,
+      "grad_norm": 0.3703326880931854,
+      "learning_rate": 0.0001,
+      "loss": 1.8643,
+      "step": 1329
+    },
+    {
+      "epoch": 0.15451641010746442,
+      "grad_norm": 0.3577940762042999,
+      "learning_rate": 0.0001,
+      "loss": 1.75,
+      "step": 1330
+    },
+    {
+      "epoch": 0.1546325878594249,
+      "grad_norm": 0.38018643856048584,
+      "learning_rate": 0.0001,
+      "loss": 1.8148,
+      "step": 1331
+    },
+    {
+      "epoch": 0.15474876561138542,
+      "grad_norm": 0.3477984368801117,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 1332
+    },
+    {
+      "epoch": 0.1548649433633459,
+      "grad_norm": 0.35512134432792664,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 1333
+    },
+    {
+      "epoch": 0.15498112111530643,
+      "grad_norm": 0.34343427419662476,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 1334
+    },
+    {
+      "epoch": 0.15509729886726692,
+      "grad_norm": 0.36737141013145447,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 1335
+    },
+    {
+      "epoch": 0.1552134766192274,
+      "grad_norm": 0.3362652659416199,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 1336
+    },
+    {
+      "epoch": 0.15532965437118793,
+      "grad_norm": 0.3790622651576996,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 1337
+    },
+    {
+      "epoch": 0.15544583212314841,
+      "grad_norm": 0.3796531856060028,
+      "learning_rate": 0.0001,
+      "loss": 1.7659,
+      "step": 1338
+    },
+    {
+      "epoch": 0.1555620098751089,
+      "grad_norm": 0.3766029477119446,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 1339
+    },
+    {
+      "epoch": 0.15567818762706942,
+      "grad_norm": 0.33512985706329346,
+      "learning_rate": 0.0001,
+      "loss": 1.7591,
+      "step": 1340
+    },
+    {
+      "epoch": 0.1557943653790299,
+      "grad_norm": 0.35915273427963257,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 1341
+    },
+    {
+      "epoch": 0.15591054313099043,
+      "grad_norm": 0.3571474254131317,
+      "learning_rate": 0.0001,
+      "loss": 1.7366,
+      "step": 1342
+    },
+    {
+      "epoch": 0.15602672088295091,
+      "grad_norm": 0.36832332611083984,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 1343
+    },
+    {
+      "epoch": 0.1561428986349114,
+      "grad_norm": 0.34558358788490295,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 1344
+    },
+    {
+      "epoch": 0.15625907638687192,
+      "grad_norm": 0.3483864665031433,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1563752541388324,
+      "grad_norm": 0.3357907831668854,
+      "learning_rate": 0.0001,
+      "loss": 1.7206,
+      "step": 1346
+    },
+    {
+      "epoch": 0.15649143189079293,
+      "grad_norm": 0.37820810079574585,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 1347
+    },
+    {
+      "epoch": 0.15660760964275341,
+      "grad_norm": 0.4033626616001129,
+      "learning_rate": 0.0001,
+      "loss": 1.8405,
+      "step": 1348
+    },
+    {
+      "epoch": 0.1567237873947139,
+      "grad_norm": 0.3691973090171814,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 1349
+    },
+    {
+      "epoch": 0.15683996514667442,
+      "grad_norm": 0.3847627639770508,
+      "learning_rate": 0.0001,
+      "loss": 1.8025,
+      "step": 1350
+    },
+    {
+      "epoch": 0.1569561428986349,
+      "grad_norm": 0.4157688617706299,
+      "learning_rate": 0.0001,
+      "loss": 1.8269,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1570723206505954,
+      "grad_norm": 0.35120290517807007,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 1352
+    },
+    {
+      "epoch": 0.15718849840255592,
+      "grad_norm": 0.3579387664794922,
+      "learning_rate": 0.0001,
+      "loss": 1.7129,
+      "step": 1353
+    },
+    {
+      "epoch": 0.1573046761545164,
+      "grad_norm": 0.3572061359882355,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 1354
+    },
+    {
+      "epoch": 0.15742085390647692,
+      "grad_norm": 0.38334596157073975,
+      "learning_rate": 0.0001,
+      "loss": 1.8802,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1575370316584374,
+      "grad_norm": 0.3814426064491272,
+      "learning_rate": 0.0001,
+      "loss": 1.8669,
+      "step": 1356
+    },
+    {
+      "epoch": 0.1576532094103979,
+      "grad_norm": 0.3972860276699066,
+      "learning_rate": 0.0001,
+      "loss": 1.8,
+      "step": 1357
+    },
+    {
+      "epoch": 0.15776938716235842,
+      "grad_norm": 0.41393837332725525,
+      "learning_rate": 0.0001,
+      "loss": 1.8818,
+      "step": 1358
+    },
+    {
+      "epoch": 0.1578855649143189,
+      "grad_norm": 0.35115018486976624,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 1359
+    },
+    {
+      "epoch": 0.15800174266627942,
+      "grad_norm": 0.36894285678863525,
+      "learning_rate": 0.0001,
+      "loss": 1.8274,
+      "step": 1360
+    },
+    {
+      "epoch": 0.1581179204182399,
+      "grad_norm": 0.344990074634552,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1582340981702004,
+      "grad_norm": 0.33179888129234314,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 1362
+    },
+    {
+      "epoch": 0.15835027592216092,
+      "grad_norm": 0.3482305407524109,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 1363
+    },
+    {
+      "epoch": 0.1584664536741214,
+      "grad_norm": 0.34571194648742676,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 1364
+    },
+    {
+      "epoch": 0.1585826314260819,
+      "grad_norm": 0.3417724668979645,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 1365
+    },
+    {
+      "epoch": 0.1586988091780424,
+      "grad_norm": 0.36330536007881165,
+      "learning_rate": 0.0001,
+      "loss": 1.7957,
+      "step": 1366
+    },
+    {
+      "epoch": 0.1588149869300029,
+      "grad_norm": 0.3438703417778015,
+      "learning_rate": 0.0001,
+      "loss": 1.8058,
+      "step": 1367
+    },
+    {
+      "epoch": 0.15893116468196342,
+      "grad_norm": 0.35250309109687805,
+      "learning_rate": 0.0001,
+      "loss": 1.8591,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1590473424339239,
+      "grad_norm": 0.3651219606399536,
+      "learning_rate": 0.0001,
+      "loss": 1.6975,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1591635201858844,
+      "grad_norm": 0.3463347256183624,
+      "learning_rate": 0.0001,
+      "loss": 1.6508,
+      "step": 1370
+    },
+    {
+      "epoch": 0.1592796979378449,
+      "grad_norm": 0.3516363799571991,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 1371
+    },
+    {
+      "epoch": 0.1593958756898054,
+      "grad_norm": 0.37753501534461975,
+      "learning_rate": 0.0001,
+      "loss": 1.8072,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1595120534417659,
+      "grad_norm": 0.3713725507259369,
+      "learning_rate": 0.0001,
+      "loss": 1.7824,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1596282311937264,
+      "grad_norm": 0.35233864188194275,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 1374
+    },
+    {
+      "epoch": 0.1597444089456869,
+      "grad_norm": 0.3708571195602417,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 1375
+    },
+    {
+      "epoch": 0.1598605866976474,
+      "grad_norm": 0.4148230254650116,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 1376
+    },
+    {
+      "epoch": 0.1599767644496079,
+      "grad_norm": 0.3738183081150055,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 1377
+    },
+    {
+      "epoch": 0.1600929422015684,
+      "grad_norm": 0.40050017833709717,
+      "learning_rate": 0.0001,
+      "loss": 1.8349,
+      "step": 1378
+    },
+    {
+      "epoch": 0.1602091199535289,
+      "grad_norm": 0.3961089551448822,
+      "learning_rate": 0.0001,
+      "loss": 1.8826,
+      "step": 1379
+    },
+    {
+      "epoch": 0.1603252977054894,
+      "grad_norm": 0.38461560010910034,
+      "learning_rate": 0.0001,
+      "loss": 1.7562,
+      "step": 1380
+    },
+    {
+      "epoch": 0.1604414754574499,
+      "grad_norm": 0.3477541208267212,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 1381
+    },
+    {
+      "epoch": 0.1605576532094104,
+      "grad_norm": 0.3460700213909149,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 1382
+    },
+    {
+      "epoch": 0.1606738309613709,
+      "grad_norm": 0.3566724956035614,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 1383
+    },
+    {
+      "epoch": 0.1607900087133314,
+      "grad_norm": 0.36470624804496765,
+      "learning_rate": 0.0001,
+      "loss": 1.8717,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1609061864652919,
+      "grad_norm": 0.3292492926120758,
+      "learning_rate": 0.0001,
+      "loss": 1.6557,
+      "step": 1385
+    },
+    {
+      "epoch": 0.16102236421725238,
+      "grad_norm": 0.36960333585739136,
+      "learning_rate": 0.0001,
+      "loss": 1.7142,
+      "step": 1386
+    },
+    {
+      "epoch": 0.1611385419692129,
+      "grad_norm": 0.36718302965164185,
+      "learning_rate": 0.0001,
+      "loss": 1.8368,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1612547197211734,
+      "grad_norm": 0.3775370419025421,
+      "learning_rate": 0.0001,
+      "loss": 1.7769,
+      "step": 1388
+    },
+    {
+      "epoch": 0.1613708974731339,
+      "grad_norm": 0.3697305917739868,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 1389
+    },
+    {
+      "epoch": 0.1614870752250944,
+      "grad_norm": 0.37467798590660095,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 1390
+    },
+    {
+      "epoch": 0.16160325297705488,
+      "grad_norm": 0.3733084499835968,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 1391
+    },
+    {
+      "epoch": 0.1617194307290154,
+      "grad_norm": 0.3742946982383728,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 1392
+    },
+    {
+      "epoch": 0.1618356084809759,
+      "grad_norm": 0.3777306079864502,
+      "learning_rate": 0.0001,
+      "loss": 1.7787,
+      "step": 1393
+    },
+    {
+      "epoch": 0.16195178623293638,
+      "grad_norm": 0.3901897370815277,
+      "learning_rate": 0.0001,
+      "loss": 1.8356,
+      "step": 1394
+    },
+    {
+      "epoch": 0.1620679639848969,
+      "grad_norm": 0.38239678740501404,
+      "learning_rate": 0.0001,
+      "loss": 1.9155,
+      "step": 1395
+    },
+    {
+      "epoch": 0.16218414173685738,
+      "grad_norm": 0.3667686879634857,
+      "learning_rate": 0.0001,
+      "loss": 1.7969,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1623003194888179,
+      "grad_norm": 0.38420993089675903,
+      "learning_rate": 0.0001,
+      "loss": 1.7801,
+      "step": 1397
+    },
+    {
+      "epoch": 0.1624164972407784,
+      "grad_norm": 0.36830443143844604,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 1398
+    },
+    {
+      "epoch": 0.16253267499273888,
+      "grad_norm": 0.35537177324295044,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1626488527446994,
+      "grad_norm": 0.35182371735572815,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 1400
+    },
+    {
+      "epoch": 0.16276503049665988,
+      "grad_norm": 0.36817070841789246,
+      "learning_rate": 0.0001,
+      "loss": 1.773,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1628812082486204,
+      "grad_norm": 0.3580371141433716,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 1402
+    },
+    {
+      "epoch": 0.1629973860005809,
+      "grad_norm": 0.37469223141670227,
+      "learning_rate": 0.0001,
+      "loss": 1.7603,
+      "step": 1403
+    },
+    {
+      "epoch": 0.16311356375254138,
+      "grad_norm": 0.39260217547416687,
+      "learning_rate": 0.0001,
+      "loss": 1.9347,
+      "step": 1404
+    },
+    {
+      "epoch": 0.1632297415045019,
+      "grad_norm": 0.37076276540756226,
+      "learning_rate": 0.0001,
+      "loss": 1.755,
+      "step": 1405
+    },
+    {
+      "epoch": 0.16334591925646239,
+      "grad_norm": 0.34179428219795227,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 1406
+    },
+    {
+      "epoch": 0.16346209700842287,
+      "grad_norm": 0.359271377325058,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1635782747603834,
+      "grad_norm": 0.3883320093154907,
+      "learning_rate": 0.0001,
+      "loss": 1.8176,
+      "step": 1408
+    },
+    {
+      "epoch": 0.16369445251234388,
+      "grad_norm": 0.35614779591560364,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 1409
+    },
+    {
+      "epoch": 0.1638106302643044,
+      "grad_norm": 0.3655150830745697,
+      "learning_rate": 0.0001,
+      "loss": 1.8285,
+      "step": 1410
+    },
+    {
+      "epoch": 0.16392680801626489,
+      "grad_norm": 0.37644287943840027,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 1411
+    },
+    {
+      "epoch": 0.16404298576822537,
+      "grad_norm": 0.33691367506980896,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 1412
+    },
+    {
+      "epoch": 0.1641591635201859,
+      "grad_norm": 0.3288310766220093,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 1413
+    },
+    {
+      "epoch": 0.16427534127214638,
+      "grad_norm": 0.37377673387527466,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 1414
+    },
+    {
+      "epoch": 0.1643915190241069,
+      "grad_norm": 0.3379480242729187,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 1415
+    },
+    {
+      "epoch": 0.16450769677606739,
+      "grad_norm": 0.3620838522911072,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 1416
+    },
+    {
+      "epoch": 0.16462387452802787,
+      "grad_norm": 0.3558915853500366,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 1417
+    },
+    {
+      "epoch": 0.1647400522799884,
+      "grad_norm": 0.36311277747154236,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 1418
+    },
+    {
+      "epoch": 0.16485623003194888,
+      "grad_norm": 0.41347232460975647,
+      "learning_rate": 0.0001,
+      "loss": 1.834,
+      "step": 1419
+    },
+    {
+      "epoch": 0.16497240778390937,
+      "grad_norm": 0.3624221086502075,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 1420
+    },
+    {
+      "epoch": 0.1650885855358699,
+      "grad_norm": 0.3721679449081421,
+      "learning_rate": 0.0001,
+      "loss": 1.7655,
+      "step": 1421
+    },
+    {
+      "epoch": 0.16520476328783038,
+      "grad_norm": 0.3526135981082916,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 1422
+    },
+    {
+      "epoch": 0.1653209410397909,
+      "grad_norm": 0.4029957354068756,
+      "learning_rate": 0.0001,
+      "loss": 2.0216,
+      "step": 1423
+    },
+    {
+      "epoch": 0.16543711879175138,
+      "grad_norm": 0.36750248074531555,
+      "learning_rate": 0.0001,
+      "loss": 1.8155,
+      "step": 1424
+    },
+    {
+      "epoch": 0.16555329654371187,
+      "grad_norm": 0.358469158411026,
+      "learning_rate": 0.0001,
+      "loss": 1.7874,
+      "step": 1425
+    },
+    {
+      "epoch": 0.1656694742956724,
+      "grad_norm": 0.3403339684009552,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 1426
+    },
+    {
+      "epoch": 0.16578565204763288,
+      "grad_norm": 0.3741562068462372,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 1427
+    },
+    {
+      "epoch": 0.16590182979959336,
+      "grad_norm": 0.34802475571632385,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 1428
+    },
+    {
+      "epoch": 0.16601800755155388,
+      "grad_norm": 0.3652847707271576,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 1429
+    },
+    {
+      "epoch": 0.16613418530351437,
+      "grad_norm": 0.37153634428977966,
+      "learning_rate": 0.0001,
+      "loss": 1.8237,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1662503630554749,
+      "grad_norm": 0.36057737469673157,
+      "learning_rate": 0.0001,
+      "loss": 1.8384,
+      "step": 1431
+    },
+    {
+      "epoch": 0.16636654080743538,
+      "grad_norm": 0.3642016053199768,
+      "learning_rate": 0.0001,
+      "loss": 1.8367,
+      "step": 1432
+    },
+    {
+      "epoch": 0.16648271855939586,
+      "grad_norm": 0.36787521839141846,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 1433
+    },
+    {
+      "epoch": 0.16659889631135638,
+      "grad_norm": 0.3940580189228058,
+      "learning_rate": 0.0001,
+      "loss": 1.9512,
+      "step": 1434
+    },
+    {
+      "epoch": 0.16671507406331687,
+      "grad_norm": 0.3518722355365753,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 1435
+    },
+    {
+      "epoch": 0.1668312518152774,
+      "grad_norm": 0.3771706223487854,
+      "learning_rate": 0.0001,
+      "loss": 1.7006,
+      "step": 1436
+    },
+    {
+      "epoch": 0.16694742956723788,
+      "grad_norm": 0.3364204466342926,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 1437
+    },
+    {
+      "epoch": 0.16706360731919837,
+      "grad_norm": 0.36092275381088257,
+      "learning_rate": 0.0001,
+      "loss": 1.8134,
+      "step": 1438
+    },
+    {
+      "epoch": 0.16717978507115888,
+      "grad_norm": 0.3495999276638031,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 1439
+    },
+    {
+      "epoch": 0.16729596282311937,
+      "grad_norm": 0.35832324624061584,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 1440
+    },
+    {
+      "epoch": 0.16741214057507986,
+      "grad_norm": 0.38260895013809204,
+      "learning_rate": 0.0001,
+      "loss": 1.8781,
+      "step": 1441
+    },
+    {
+      "epoch": 0.16752831832704038,
+      "grad_norm": 0.37501585483551025,
+      "learning_rate": 0.0001,
+      "loss": 1.7943,
+      "step": 1442
+    },
+    {
+      "epoch": 0.16764449607900087,
+      "grad_norm": 0.34276142716407776,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 1443
+    },
+    {
+      "epoch": 0.16776067383096138,
+      "grad_norm": 0.32856735587120056,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 1444
+    },
+    {
+      "epoch": 0.16787685158292187,
+      "grad_norm": 0.35961073637008667,
+      "learning_rate": 0.0001,
+      "loss": 1.7903,
+      "step": 1445
+    },
+    {
+      "epoch": 0.16799302933488236,
+      "grad_norm": 0.3753131628036499,
+      "learning_rate": 0.0001,
+      "loss": 1.8137,
+      "step": 1446
+    },
+    {
+      "epoch": 0.16810920708684288,
+      "grad_norm": 0.3495393693447113,
+      "learning_rate": 0.0001,
+      "loss": 1.7824,
+      "step": 1447
+    },
+    {
+      "epoch": 0.16822538483880337,
+      "grad_norm": 0.35757923126220703,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 1448
+    },
+    {
+      "epoch": 0.16834156259076388,
+      "grad_norm": 0.3703269064426422,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 1449
+    },
+    {
+      "epoch": 0.16845774034272437,
+      "grad_norm": 0.3545929789543152,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 1450
+    },
+    {
+      "epoch": 0.16857391809468486,
+      "grad_norm": 0.3971767723560333,
+      "learning_rate": 0.0001,
+      "loss": 1.8604,
+      "step": 1451
+    },
+    {
+      "epoch": 0.16869009584664538,
+      "grad_norm": 0.3627997040748596,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 1452
+    },
+    {
+      "epoch": 0.16880627359860587,
+      "grad_norm": 0.37545061111450195,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 1453
+    },
+    {
+      "epoch": 0.16892245135056635,
+      "grad_norm": 0.4215669631958008,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 1454
+    },
+    {
+      "epoch": 0.16903862910252687,
+      "grad_norm": 0.3428536355495453,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 1455
+    },
+    {
+      "epoch": 0.16915480685448736,
+      "grad_norm": 0.3856685161590576,
+      "learning_rate": 0.0001,
+      "loss": 1.8013,
+      "step": 1456
+    },
+    {
+      "epoch": 0.16927098460644788,
+      "grad_norm": 0.3408958315849304,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 1457
+    },
+    {
+      "epoch": 0.16938716235840837,
+      "grad_norm": 0.3705407381057739,
+      "learning_rate": 0.0001,
+      "loss": 1.7433,
+      "step": 1458
+    },
+    {
+      "epoch": 0.16950334011036886,
+      "grad_norm": 0.3559417128562927,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 1459
+    },
+    {
+      "epoch": 0.16961951786232937,
+      "grad_norm": 0.39479878544807434,
+      "learning_rate": 0.0001,
+      "loss": 1.8522,
+      "step": 1460
+    },
+    {
+      "epoch": 0.16973569561428986,
+      "grad_norm": 0.36740705370903015,
+      "learning_rate": 0.0001,
+      "loss": 1.7336,
+      "step": 1461
+    },
+    {
+      "epoch": 0.16985187336625035,
+      "grad_norm": 0.37933793663978577,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 1462
+    },
+    {
+      "epoch": 0.16996805111821087,
+      "grad_norm": 0.3855569064617157,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 1463
+    },
+    {
+      "epoch": 0.17008422887017136,
+      "grad_norm": 0.36735615134239197,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 1464
+    },
+    {
+      "epoch": 0.17020040662213187,
+      "grad_norm": 0.3651230037212372,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 1465
+    },
+    {
+      "epoch": 0.17031658437409236,
+      "grad_norm": 0.3359425663948059,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 1466
+    },
+    {
+      "epoch": 0.17043276212605285,
+      "grad_norm": 0.37476974725723267,
+      "learning_rate": 0.0001,
+      "loss": 1.7478,
+      "step": 1467
+    },
+    {
+      "epoch": 0.17054893987801337,
+      "grad_norm": 0.3349404036998749,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 1468
+    },
+    {
+      "epoch": 0.17066511762997386,
+      "grad_norm": 0.375108003616333,
+      "learning_rate": 0.0001,
+      "loss": 1.7672,
+      "step": 1469
+    },
+    {
+      "epoch": 0.17078129538193437,
+      "grad_norm": 0.36245962977409363,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 1470
+    },
+    {
+      "epoch": 0.17089747313389486,
+      "grad_norm": 0.3471023142337799,
+      "learning_rate": 0.0001,
+      "loss": 1.4378,
+      "step": 1471
+    },
+    {
+      "epoch": 0.17101365088585535,
+      "grad_norm": 0.37973228096961975,
+      "learning_rate": 0.0001,
+      "loss": 1.8489,
+      "step": 1472
+    },
+    {
+      "epoch": 0.17112982863781587,
+      "grad_norm": 0.3702830374240875,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 1473
+    },
+    {
+      "epoch": 0.17124600638977636,
+      "grad_norm": 0.37454211711883545,
+      "learning_rate": 0.0001,
+      "loss": 1.8488,
+      "step": 1474
+    },
+    {
+      "epoch": 0.17136218414173685,
+      "grad_norm": 0.3523843586444855,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 1475
+    },
+    {
+      "epoch": 0.17147836189369736,
+      "grad_norm": 0.38084614276885986,
+      "learning_rate": 0.0001,
+      "loss": 1.8845,
+      "step": 1476
+    },
+    {
+      "epoch": 0.17159453964565785,
+      "grad_norm": 0.35280105471611023,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 1477
+    },
+    {
+      "epoch": 0.17171071739761837,
+      "grad_norm": 0.34946003556251526,
+      "learning_rate": 0.0001,
+      "loss": 1.5721,
+      "step": 1478
+    },
+    {
+      "epoch": 0.17182689514957886,
+      "grad_norm": 0.3549077808856964,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 1479
+    },
+    {
+      "epoch": 0.17194307290153935,
+      "grad_norm": 0.34895646572113037,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 1480
+    },
+    {
+      "epoch": 0.17205925065349986,
+      "grad_norm": 0.3998420536518097,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 1481
+    },
+    {
+      "epoch": 0.17217542840546035,
+      "grad_norm": 0.3754946291446686,
+      "learning_rate": 0.0001,
+      "loss": 1.6843,
+      "step": 1482
+    },
+    {
+      "epoch": 0.17229160615742084,
+      "grad_norm": 0.36129212379455566,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 1483
+    },
+    {
+      "epoch": 0.17240778390938136,
+      "grad_norm": 0.3542667329311371,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 1484
+    },
+    {
+      "epoch": 0.17252396166134185,
+      "grad_norm": 0.3390235900878906,
+      "learning_rate": 0.0001,
+      "loss": 1.5409,
+      "step": 1485
+    },
+    {
+      "epoch": 0.17264013941330236,
+      "grad_norm": 0.36193352937698364,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 1486
+    },
+    {
+      "epoch": 0.17275631716526285,
+      "grad_norm": 0.3606919050216675,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 1487
+    },
+    {
+      "epoch": 0.17287249491722334,
+      "grad_norm": 0.4732804000377655,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 1488
+    },
+    {
+      "epoch": 0.17298867266918386,
+      "grad_norm": 0.3804911971092224,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 1489
+    },
+    {
+      "epoch": 0.17310485042114435,
+      "grad_norm": 0.37315547466278076,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 1490
+    },
+    {
+      "epoch": 0.17322102817310486,
+      "grad_norm": 0.3711428642272949,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 1491
+    },
+    {
+      "epoch": 0.17333720592506535,
+      "grad_norm": 0.33480215072631836,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 1492
+    },
+    {
+      "epoch": 0.17345338367702584,
+      "grad_norm": 0.3833867907524109,
+      "learning_rate": 0.0001,
+      "loss": 1.7718,
+      "step": 1493
+    },
+    {
+      "epoch": 0.17356956142898636,
+      "grad_norm": 0.35482731461524963,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 1494
+    },
+    {
+      "epoch": 0.17368573918094685,
+      "grad_norm": 0.35908132791519165,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 1495
+    },
+    {
+      "epoch": 0.17380191693290734,
+      "grad_norm": 0.34301432967185974,
+      "learning_rate": 0.0001,
+      "loss": 1.6988,
+      "step": 1496
+    },
+    {
+      "epoch": 0.17391809468486785,
+      "grad_norm": 0.34541556239128113,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 1497
+    },
+    {
+      "epoch": 0.17403427243682834,
+      "grad_norm": 0.36009690165519714,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 1498
+    },
+    {
+      "epoch": 0.17415045018878886,
+      "grad_norm": 0.3848399221897125,
+      "learning_rate": 0.0001,
+      "loss": 1.9549,
+      "step": 1499
+    },
+    {
+      "epoch": 0.17426662794074935,
+      "grad_norm": 0.35505911707878113,
+      "learning_rate": 0.0001,
+      "loss": 1.7141,
+      "step": 1500
+    },
+    {
+      "epoch": 0.17438280569270984,
+      "grad_norm": 0.3311626613140106,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 1501
+    },
+    {
+      "epoch": 0.17449898344467035,
+      "grad_norm": 0.37923452258110046,
+      "learning_rate": 0.0001,
+      "loss": 1.7772,
+      "step": 1502
+    },
+    {
+      "epoch": 0.17461516119663084,
+      "grad_norm": 0.3624334931373596,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 1503
+    },
+    {
+      "epoch": 0.17473133894859136,
+      "grad_norm": 0.3661384880542755,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 1504
+    },
+    {
+      "epoch": 0.17484751670055185,
+      "grad_norm": 0.3476395308971405,
+      "learning_rate": 0.0001,
+      "loss": 1.8493,
+      "step": 1505
+    },
+    {
+      "epoch": 0.17496369445251234,
+      "grad_norm": 0.3515165448188782,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 1506
+    },
+    {
+      "epoch": 0.17507987220447285,
+      "grad_norm": 0.36239245533943176,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 1507
+    },
+    {
+      "epoch": 0.17519604995643334,
+      "grad_norm": 0.3474177420139313,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 1508
+    },
+    {
+      "epoch": 0.17531222770839383,
+      "grad_norm": 0.34168288111686707,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 1509
+    },
+    {
+      "epoch": 0.17542840546035435,
+      "grad_norm": 0.3697127103805542,
+      "learning_rate": 0.0001,
+      "loss": 1.7274,
+      "step": 1510
+    },
+    {
+      "epoch": 0.17554458321231484,
+      "grad_norm": 0.36415886878967285,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 1511
+    },
+    {
+      "epoch": 0.17566076096427535,
+      "grad_norm": 0.3710338771343231,
+      "learning_rate": 0.0001,
+      "loss": 1.8252,
+      "step": 1512
+    },
+    {
+      "epoch": 0.17577693871623584,
+      "grad_norm": 0.3623411953449249,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 1513
+    },
+    {
+      "epoch": 0.17589311646819633,
+      "grad_norm": 0.3631919026374817,
+      "learning_rate": 0.0001,
+      "loss": 1.7089,
+      "step": 1514
+    },
+    {
+      "epoch": 0.17600929422015685,
+      "grad_norm": 0.37533360719680786,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 1515
+    },
+    {
+      "epoch": 0.17612547197211734,
+      "grad_norm": 0.36933374404907227,
+      "learning_rate": 0.0001,
+      "loss": 1.8289,
+      "step": 1516
+    },
+    {
+      "epoch": 0.17624164972407783,
+      "grad_norm": 0.3541377782821655,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 1517
+    },
+    {
+      "epoch": 0.17635782747603834,
+      "grad_norm": 0.38855910301208496,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 1518
+    },
+    {
+      "epoch": 0.17647400522799883,
+      "grad_norm": 0.3846849203109741,
+      "learning_rate": 0.0001,
+      "loss": 1.81,
+      "step": 1519
+    },
+    {
+      "epoch": 0.17659018297995935,
+      "grad_norm": 0.4078463912010193,
+      "learning_rate": 0.0001,
+      "loss": 1.8295,
+      "step": 1520
+    },
+    {
+      "epoch": 0.17670636073191984,
+      "grad_norm": 0.3415427803993225,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 1521
+    },
+    {
+      "epoch": 0.17682253848388033,
+      "grad_norm": 0.36249756813049316,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 1522
+    },
+    {
+      "epoch": 0.17693871623584084,
+      "grad_norm": 0.3513442277908325,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 1523
+    },
+    {
+      "epoch": 0.17705489398780133,
+      "grad_norm": 0.42712968587875366,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 1524
+    },
+    {
+      "epoch": 0.17717107173976185,
+      "grad_norm": 0.3688381612300873,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 1525
+    },
+    {
+      "epoch": 0.17728724949172234,
+      "grad_norm": 0.38176625967025757,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 1526
+    },
+    {
+      "epoch": 0.17740342724368283,
+      "grad_norm": 0.3425157070159912,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 1527
+    },
+    {
+      "epoch": 0.17751960499564334,
+      "grad_norm": 0.362857848405838,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 1528
+    },
+    {
+      "epoch": 0.17763578274760383,
+      "grad_norm": 0.3613092303276062,
+      "learning_rate": 0.0001,
+      "loss": 1.7633,
+      "step": 1529
+    },
+    {
+      "epoch": 0.17775196049956432,
+      "grad_norm": 0.35249873995780945,
+      "learning_rate": 0.0001,
+      "loss": 1.7397,
+      "step": 1530
+    },
+    {
+      "epoch": 0.17786813825152484,
+      "grad_norm": 0.3414962887763977,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 1531
+    },
+    {
+      "epoch": 0.17798431600348533,
+      "grad_norm": 0.3660277724266052,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 1532
+    },
+    {
+      "epoch": 0.17810049375544584,
+      "grad_norm": 0.344135046005249,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 1533
+    },
+    {
+      "epoch": 0.17821667150740633,
+      "grad_norm": 0.3552722632884979,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 1534
+    },
+    {
+      "epoch": 0.17833284925936682,
+      "grad_norm": 0.3565710186958313,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 1535
+    },
+    {
+      "epoch": 0.17844902701132734,
+      "grad_norm": 0.34754517674446106,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 1536
+    },
+    {
+      "epoch": 0.17856520476328783,
+      "grad_norm": 0.377332478761673,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 1537
+    },
+    {
+      "epoch": 0.17868138251524834,
+      "grad_norm": 0.34879541397094727,
+      "learning_rate": 0.0001,
+      "loss": 1.6475,
+      "step": 1538
+    },
+    {
+      "epoch": 0.17879756026720883,
+      "grad_norm": 0.3693576753139496,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 1539
+    },
+    {
+      "epoch": 0.17891373801916932,
+      "grad_norm": 0.3447043001651764,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 1540
+    },
+    {
+      "epoch": 0.17902991577112984,
+      "grad_norm": 0.3493332862854004,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 1541
+    },
+    {
+      "epoch": 0.17914609352309033,
+      "grad_norm": 0.3783632814884186,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 1542
+    },
+    {
+      "epoch": 0.17926227127505082,
+      "grad_norm": 0.33536890149116516,
+      "learning_rate": 0.0001,
+      "loss": 1.7176,
+      "step": 1543
+    },
+    {
+      "epoch": 0.17937844902701133,
+      "grad_norm": 0.3568625748157501,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 1544
+    },
+    {
+      "epoch": 0.17949462677897182,
+      "grad_norm": 0.3375276029109955,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 1545
+    },
+    {
+      "epoch": 0.17961080453093234,
+      "grad_norm": 0.3463954031467438,
+      "learning_rate": 0.0001,
+      "loss": 1.7331,
+      "step": 1546
+    },
+    {
+      "epoch": 0.17972698228289283,
+      "grad_norm": 0.38217151165008545,
+      "learning_rate": 0.0001,
+      "loss": 1.7775,
+      "step": 1547
+    },
+    {
+      "epoch": 0.17984316003485332,
+      "grad_norm": 0.3745064437389374,
+      "learning_rate": 0.0001,
+      "loss": 1.789,
+      "step": 1548
+    },
+    {
+      "epoch": 0.17995933778681383,
+      "grad_norm": 0.3647879660129547,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 1549
+    },
+    {
+      "epoch": 0.18007551553877432,
+      "grad_norm": 0.34927716851234436,
+      "learning_rate": 0.0001,
+      "loss": 1.7875,
+      "step": 1550
+    },
+    {
+      "epoch": 0.1801916932907348,
+      "grad_norm": 0.35309305787086487,
+      "learning_rate": 0.0001,
+      "loss": 1.7382,
+      "step": 1551
+    },
+    {
+      "epoch": 0.18030787104269533,
+      "grad_norm": 0.385146826505661,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 1552
+    },
+    {
+      "epoch": 0.18042404879465582,
+      "grad_norm": 0.345344603061676,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 1553
+    },
+    {
+      "epoch": 0.18054022654661633,
+      "grad_norm": 0.3495193123817444,
+      "learning_rate": 0.0001,
+      "loss": 1.87,
+      "step": 1554
+    },
+    {
+      "epoch": 0.18065640429857682,
+      "grad_norm": 0.36495864391326904,
+      "learning_rate": 0.0001,
+      "loss": 1.8289,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1807725820505373,
+      "grad_norm": 0.34785377979278564,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 1556
+    },
+    {
+      "epoch": 0.18088875980249783,
+      "grad_norm": 0.36701300740242004,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 1557
+    },
+    {
+      "epoch": 0.18100493755445832,
+      "grad_norm": 0.3620215356349945,
+      "learning_rate": 0.0001,
+      "loss": 1.6724,
+      "step": 1558
+    },
+    {
+      "epoch": 0.18112111530641883,
+      "grad_norm": 0.34096696972846985,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 1559
+    },
+    {
+      "epoch": 0.18123729305837932,
+      "grad_norm": 0.3897305428981781,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1813534708103398,
+      "grad_norm": 0.3481612503528595,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 1561
+    },
+    {
+      "epoch": 0.18146964856230033,
+      "grad_norm": 0.3717382550239563,
+      "learning_rate": 0.0001,
+      "loss": 1.781,
+      "step": 1562
+    },
+    {
+      "epoch": 0.18158582631426082,
+      "grad_norm": 0.3604913353919983,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1817020040662213,
+      "grad_norm": 0.36964964866638184,
+      "learning_rate": 0.0001,
+      "loss": 1.7833,
+      "step": 1564
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 0.35544395446777344,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 1565
+    },
+    {
+      "epoch": 0.1819343595701423,
+      "grad_norm": 0.36883848905563354,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 1566
+    },
+    {
+      "epoch": 0.18205053732210283,
+      "grad_norm": 0.36678242683410645,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 1567
+    },
+    {
+      "epoch": 0.18216671507406332,
+      "grad_norm": 0.3713185489177704,
+      "learning_rate": 0.0001,
+      "loss": 1.8655,
+      "step": 1568
+    },
+    {
+      "epoch": 0.1822828928260238,
+      "grad_norm": 0.32219791412353516,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 1569
+    },
+    {
+      "epoch": 0.18239907057798432,
+      "grad_norm": 0.3948063254356384,
+      "learning_rate": 0.0001,
+      "loss": 1.7762,
+      "step": 1570
+    },
+    {
+      "epoch": 0.1825152483299448,
+      "grad_norm": 0.35549992322921753,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 1571
+    },
+    {
+      "epoch": 0.1826314260819053,
+      "grad_norm": 0.36282774806022644,
+      "learning_rate": 0.0001,
+      "loss": 1.7864,
+      "step": 1572
+    },
+    {
+      "epoch": 0.18274760383386582,
+      "grad_norm": 0.35417771339416504,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 1573
+    },
+    {
+      "epoch": 0.1828637815858263,
+      "grad_norm": 0.406324565410614,
+      "learning_rate": 0.0001,
+      "loss": 1.9048,
+      "step": 1574
+    },
+    {
+      "epoch": 0.18297995933778682,
+      "grad_norm": 0.3815106153488159,
+      "learning_rate": 0.0001,
+      "loss": 1.7483,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1830961370897473,
+      "grad_norm": 0.3650597929954529,
+      "learning_rate": 0.0001,
+      "loss": 1.8492,
+      "step": 1576
+    },
+    {
+      "epoch": 0.1832123148417078,
+      "grad_norm": 0.3681640326976776,
+      "learning_rate": 0.0001,
+      "loss": 1.8177,
+      "step": 1577
+    },
+    {
+      "epoch": 0.18332849259366832,
+      "grad_norm": 0.3654261529445648,
+      "learning_rate": 0.0001,
+      "loss": 1.747,
+      "step": 1578
+    },
+    {
+      "epoch": 0.1834446703456288,
+      "grad_norm": 0.35047435760498047,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 1579
+    },
+    {
+      "epoch": 0.18356084809758932,
+      "grad_norm": 0.33045920729637146,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1836770258495498,
+      "grad_norm": 0.3612997233867645,
+      "learning_rate": 0.0001,
+      "loss": 1.5902,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1837932036015103,
+      "grad_norm": 0.382269948720932,
+      "learning_rate": 0.0001,
+      "loss": 1.8914,
+      "step": 1582
+    },
+    {
+      "epoch": 0.18390938135347082,
+      "grad_norm": 0.348799467086792,
+      "learning_rate": 0.0001,
+      "loss": 1.7365,
+      "step": 1583
+    },
+    {
+      "epoch": 0.1840255591054313,
+      "grad_norm": 0.3481920063495636,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1841417368573918,
+      "grad_norm": 0.38168272376060486,
+      "learning_rate": 0.0001,
+      "loss": 1.8146,
+      "step": 1585
+    },
+    {
+      "epoch": 0.1842579146093523,
+      "grad_norm": 0.3492324650287628,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 1586
+    },
+    {
+      "epoch": 0.1843740923613128,
+      "grad_norm": 0.38045451045036316,
+      "learning_rate": 0.0001,
+      "loss": 1.7292,
+      "step": 1587
+    },
+    {
+      "epoch": 0.18449027011327332,
+      "grad_norm": 0.3941299617290497,
+      "learning_rate": 0.0001,
+      "loss": 1.9037,
+      "step": 1588
+    },
+    {
+      "epoch": 0.1846064478652338,
+      "grad_norm": 0.3451482355594635,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 1589
+    },
+    {
+      "epoch": 0.1847226256171943,
+      "grad_norm": 0.36182719469070435,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1848388033691548,
+      "grad_norm": 0.3482245206832886,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 1591
+    },
+    {
+      "epoch": 0.1849549811211153,
+      "grad_norm": 0.3704969584941864,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 1592
+    },
+    {
+      "epoch": 0.18507115887307582,
+      "grad_norm": 0.3831556737422943,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 1593
+    },
+    {
+      "epoch": 0.1851873366250363,
+      "grad_norm": 0.3729779124259949,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 1594
+    },
+    {
+      "epoch": 0.1853035143769968,
+      "grad_norm": 0.38106775283813477,
+      "learning_rate": 0.0001,
+      "loss": 1.6608,
+      "step": 1595
+    },
+    {
+      "epoch": 0.18541969212895731,
+      "grad_norm": 0.3817857801914215,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1855358698809178,
+      "grad_norm": 0.35032176971435547,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 1597
+    },
+    {
+      "epoch": 0.1856520476328783,
+      "grad_norm": 0.3457838296890259,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1857682253848388,
+      "grad_norm": 0.37174192070961,
+      "learning_rate": 0.0001,
+      "loss": 1.7605,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1858844031367993,
+      "grad_norm": 0.36577579379081726,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 1600
+    },
+    {
+      "epoch": 0.18600058088875981,
+      "grad_norm": 0.3587372899055481,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 1601
+    },
+    {
+      "epoch": 0.1861167586407203,
+      "grad_norm": 0.3637838363647461,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1862329363926808,
+      "grad_norm": 0.3631284534931183,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 1603
+    },
+    {
+      "epoch": 0.1863491141446413,
+      "grad_norm": 0.3679940104484558,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 1604
+    },
+    {
+      "epoch": 0.1864652918966018,
+      "grad_norm": 0.36848756670951843,
+      "learning_rate": 0.0001,
+      "loss": 1.7786,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1865814696485623,
+      "grad_norm": 0.36538413166999817,
+      "learning_rate": 0.0001,
+      "loss": 1.7454,
+      "step": 1606
+    },
+    {
+      "epoch": 0.1866976474005228,
+      "grad_norm": 0.382051944732666,
+      "learning_rate": 0.0001,
+      "loss": 1.842,
+      "step": 1607
+    },
+    {
+      "epoch": 0.1868138251524833,
+      "grad_norm": 0.33358198404312134,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1869300029044438,
+      "grad_norm": 0.3623516261577606,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1870461806564043,
+      "grad_norm": 0.37408748269081116,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 1610
+    },
+    {
+      "epoch": 0.1871623584083648,
+      "grad_norm": 0.38509878516197205,
+      "learning_rate": 0.0001,
+      "loss": 1.935,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1872785361603253,
+      "grad_norm": 0.39082857966423035,
+      "learning_rate": 0.0001,
+      "loss": 1.7414,
+      "step": 1612
+    },
+    {
+      "epoch": 0.1873947139122858,
+      "grad_norm": 0.3575880825519562,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 1613
+    },
+    {
+      "epoch": 0.1875108916642463,
+      "grad_norm": 0.3531934916973114,
+      "learning_rate": 0.0001,
+      "loss": 1.7231,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1876270694162068,
+      "grad_norm": 0.3784133493900299,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1877432471681673,
+      "grad_norm": 0.37965330481529236,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 1616
+    },
+    {
+      "epoch": 0.1878594249201278,
+      "grad_norm": 0.36382848024368286,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1879756026720883,
+      "grad_norm": 0.3851543068885803,
+      "learning_rate": 0.0001,
+      "loss": 1.7535,
+      "step": 1618
+    },
+    {
+      "epoch": 0.18809178042404878,
+      "grad_norm": 0.3497619926929474,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1882079581760093,
+      "grad_norm": 0.34661221504211426,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1883241359279698,
+      "grad_norm": 0.348257452249527,
+      "learning_rate": 0.0001,
+      "loss": 1.7566,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1884403136799303,
+      "grad_norm": 0.3511572480201721,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 1622
+    },
+    {
+      "epoch": 0.1885564914318908,
+      "grad_norm": 0.3942771553993225,
+      "learning_rate": 0.0001,
+      "loss": 1.815,
+      "step": 1623
+    },
+    {
+      "epoch": 0.18867266918385128,
+      "grad_norm": 0.3480950593948364,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 1624
+    },
+    {
+      "epoch": 0.1887888469358118,
+      "grad_norm": 0.3844303786754608,
+      "learning_rate": 0.0001,
+      "loss": 1.8658,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1889050246877723,
+      "grad_norm": 0.36564818024635315,
+      "learning_rate": 0.0001,
+      "loss": 1.7391,
+      "step": 1626
+    },
+    {
+      "epoch": 0.1890212024397328,
+      "grad_norm": 0.3397161662578583,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 1627
+    },
+    {
+      "epoch": 0.1891373801916933,
+      "grad_norm": 0.32916903495788574,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 1628
+    },
+    {
+      "epoch": 0.18925355794365378,
+      "grad_norm": 0.37873196601867676,
+      "learning_rate": 0.0001,
+      "loss": 1.754,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1893697356956143,
+      "grad_norm": 0.3580092787742615,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 1630
+    },
+    {
+      "epoch": 0.1894859134475748,
+      "grad_norm": 0.38638558983802795,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 1631
+    },
+    {
+      "epoch": 0.18960209119953528,
+      "grad_norm": 0.39580628275871277,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1897182689514958,
+      "grad_norm": 0.3963046967983246,
+      "learning_rate": 0.0001,
+      "loss": 1.8863,
+      "step": 1633
+    },
+    {
+      "epoch": 0.18983444670345628,
+      "grad_norm": 0.33917540311813354,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 1634
+    },
+    {
+      "epoch": 0.1899506244554168,
+      "grad_norm": 0.4500591456890106,
+      "learning_rate": 0.0001,
+      "loss": 1.7972,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1900668022073773,
+      "grad_norm": 0.3522135615348816,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 1636
+    },
+    {
+      "epoch": 0.19018297995933778,
+      "grad_norm": 0.3856748640537262,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 1637
+    },
+    {
+      "epoch": 0.1902991577112983,
+      "grad_norm": 0.4033121168613434,
+      "learning_rate": 0.0001,
+      "loss": 1.8365,
+      "step": 1638
+    },
+    {
+      "epoch": 0.19041533546325878,
+      "grad_norm": 0.3969012200832367,
+      "learning_rate": 0.0001,
+      "loss": 1.9015,
+      "step": 1639
+    },
+    {
+      "epoch": 0.19053151321521927,
+      "grad_norm": 0.3580697178840637,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1906476909671798,
+      "grad_norm": 0.37800854444503784,
+      "learning_rate": 0.0001,
+      "loss": 1.7347,
+      "step": 1641
+    },
+    {
+      "epoch": 0.19076386871914028,
+      "grad_norm": 0.3692120313644409,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1908800464711008,
+      "grad_norm": 0.380416601896286,
+      "learning_rate": 0.0001,
+      "loss": 1.851,
+      "step": 1643
+    },
+    {
+      "epoch": 0.19099622422306128,
+      "grad_norm": 0.3523741364479065,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 1644
+    },
+    {
+      "epoch": 0.19111240197502177,
+      "grad_norm": 0.338820219039917,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1912285797269823,
+      "grad_norm": 0.356183260679245,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 1646
+    },
+    {
+      "epoch": 0.19134475747894278,
+      "grad_norm": 0.34077268838882446,
+      "learning_rate": 0.0001,
+      "loss": 1.7581,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1914609352309033,
+      "grad_norm": 0.3771745562553406,
+      "learning_rate": 0.0001,
+      "loss": 1.7918,
+      "step": 1648
+    },
+    {
+      "epoch": 0.19157711298286378,
+      "grad_norm": 0.3605990409851074,
+      "learning_rate": 0.0001,
+      "loss": 1.6994,
+      "step": 1649
+    },
+    {
+      "epoch": 0.19169329073482427,
+      "grad_norm": 0.3496057987213135,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 1650
+    },
+    {
+      "epoch": 0.1918094684867848,
+      "grad_norm": 0.39877089858055115,
+      "learning_rate": 0.0001,
+      "loss": 1.9872,
+      "step": 1651
+    },
+    {
+      "epoch": 0.19192564623874528,
+      "grad_norm": 0.3556186556816101,
+      "learning_rate": 0.0001,
+      "loss": 1.7769,
+      "step": 1652
+    },
+    {
+      "epoch": 0.19204182399070577,
+      "grad_norm": 0.3624957501888275,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 1653
+    },
+    {
+      "epoch": 0.19215800174266628,
+      "grad_norm": 0.3758990466594696,
+      "learning_rate": 0.0001,
+      "loss": 1.7976,
+      "step": 1654
+    },
+    {
+      "epoch": 0.19227417949462677,
+      "grad_norm": 0.38239556550979614,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1923903572465873,
+      "grad_norm": 0.3543239235877991,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 1656
+    },
+    {
+      "epoch": 0.19250653499854778,
+      "grad_norm": 0.3758254647254944,
+      "learning_rate": 0.0001,
+      "loss": 1.8641,
+      "step": 1657
+    },
+    {
+      "epoch": 0.19262271275050827,
+      "grad_norm": 0.40138089656829834,
+      "learning_rate": 0.0001,
+      "loss": 1.8514,
+      "step": 1658
+    },
+    {
+      "epoch": 0.19273889050246878,
+      "grad_norm": 0.36116254329681396,
+      "learning_rate": 0.0001,
+      "loss": 1.7876,
+      "step": 1659
+    },
+    {
+      "epoch": 0.19285506825442927,
+      "grad_norm": 0.34624597430229187,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 1660
+    },
+    {
+      "epoch": 0.19297124600638976,
+      "grad_norm": 0.3567349910736084,
+      "learning_rate": 0.0001,
+      "loss": 1.7766,
+      "step": 1661
+    },
+    {
+      "epoch": 0.19308742375835028,
+      "grad_norm": 0.35692065954208374,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 1662
+    },
+    {
+      "epoch": 0.19320360151031077,
+      "grad_norm": 0.32495301961898804,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 1663
+    },
+    {
+      "epoch": 0.19331977926227129,
+      "grad_norm": 0.3486839830875397,
+      "learning_rate": 0.0001,
+      "loss": 1.739,
+      "step": 1664
+    },
+    {
+      "epoch": 0.19343595701423177,
+      "grad_norm": 0.3624487817287445,
+      "learning_rate": 0.0001,
+      "loss": 1.783,
+      "step": 1665
+    },
+    {
+      "epoch": 0.19355213476619226,
+      "grad_norm": 0.34828513860702515,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 1666
+    },
+    {
+      "epoch": 0.19366831251815278,
+      "grad_norm": 0.339197039604187,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 1667
+    },
+    {
+      "epoch": 0.19378449027011327,
+      "grad_norm": 0.3725949227809906,
+      "learning_rate": 0.0001,
+      "loss": 1.7417,
+      "step": 1668
+    },
+    {
+      "epoch": 0.19390066802207379,
+      "grad_norm": 0.346892386674881,
+      "learning_rate": 0.0001,
+      "loss": 1.4863,
+      "step": 1669
+    },
+    {
+      "epoch": 0.19401684577403427,
+      "grad_norm": 0.36845695972442627,
+      "learning_rate": 0.0001,
+      "loss": 1.8,
+      "step": 1670
+    },
+    {
+      "epoch": 0.19413302352599476,
+      "grad_norm": 0.33988621830940247,
+      "learning_rate": 0.0001,
+      "loss": 1.664,
+      "step": 1671
+    },
+    {
+      "epoch": 0.19424920127795528,
+      "grad_norm": 0.35175544023513794,
+      "learning_rate": 0.0001,
+      "loss": 1.7329,
+      "step": 1672
+    },
+    {
+      "epoch": 0.19436537902991577,
+      "grad_norm": 0.35789933800697327,
+      "learning_rate": 0.0001,
+      "loss": 1.6697,
+      "step": 1673
+    },
+    {
+      "epoch": 0.19448155678187626,
+      "grad_norm": 0.371448278427124,
+      "learning_rate": 0.0001,
+      "loss": 1.7318,
+      "step": 1674
+    },
+    {
+      "epoch": 0.19459773453383677,
+      "grad_norm": 0.3563764989376068,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 1675
+    },
+    {
+      "epoch": 0.19471391228579726,
+      "grad_norm": 0.38420358300209045,
+      "learning_rate": 0.0001,
+      "loss": 1.8371,
+      "step": 1676
+    },
+    {
+      "epoch": 0.19483009003775778,
+      "grad_norm": 0.3995816707611084,
+      "learning_rate": 0.0001,
+      "loss": 1.8055,
+      "step": 1677
+    },
+    {
+      "epoch": 0.19494626778971827,
+      "grad_norm": 0.3630034923553467,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 1678
+    },
+    {
+      "epoch": 0.19506244554167876,
+      "grad_norm": 0.3776208162307739,
+      "learning_rate": 0.0001,
+      "loss": 1.8491,
+      "step": 1679
+    },
+    {
+      "epoch": 0.19517862329363928,
+      "grad_norm": 0.3581395149230957,
+      "learning_rate": 0.0001,
+      "loss": 1.8839,
+      "step": 1680
+    },
+    {
+      "epoch": 0.19529480104559976,
+      "grad_norm": 0.3381625711917877,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 1681
+    },
+    {
+      "epoch": 0.19541097879756028,
+      "grad_norm": 0.3696902096271515,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 1682
+    },
+    {
+      "epoch": 0.19552715654952077,
+      "grad_norm": 0.3632832169532776,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 1683
+    },
+    {
+      "epoch": 0.19564333430148126,
+      "grad_norm": 0.37248799204826355,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 1684
+    },
+    {
+      "epoch": 0.19575951205344178,
+      "grad_norm": 0.354970782995224,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 1685
+    },
+    {
+      "epoch": 0.19587568980540226,
+      "grad_norm": 0.370175838470459,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 1686
+    },
+    {
+      "epoch": 0.19599186755736275,
+      "grad_norm": 0.38560378551483154,
+      "learning_rate": 0.0001,
+      "loss": 1.8188,
+      "step": 1687
+    },
+    {
+      "epoch": 0.19610804530932327,
+      "grad_norm": 0.3574475049972534,
+      "learning_rate": 0.0001,
+      "loss": 1.7771,
+      "step": 1688
+    },
+    {
+      "epoch": 0.19622422306128376,
+      "grad_norm": 0.39494848251342773,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 1689
+    },
+    {
+      "epoch": 0.19634040081324428,
+      "grad_norm": 0.36588966846466064,
+      "learning_rate": 0.0001,
+      "loss": 1.7938,
+      "step": 1690
+    },
+    {
+      "epoch": 0.19645657856520476,
+      "grad_norm": 0.3459327220916748,
+      "learning_rate": 0.0001,
+      "loss": 1.739,
+      "step": 1691
+    },
+    {
+      "epoch": 0.19657275631716525,
+      "grad_norm": 0.36302775144577026,
+      "learning_rate": 0.0001,
+      "loss": 1.741,
+      "step": 1692
+    },
+    {
+      "epoch": 0.19668893406912577,
+      "grad_norm": 0.3560570776462555,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 1693
+    },
+    {
+      "epoch": 0.19680511182108626,
+      "grad_norm": 0.35664597153663635,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 1694
+    },
+    {
+      "epoch": 0.19692128957304675,
+      "grad_norm": 0.35501745343208313,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 1695
+    },
+    {
+      "epoch": 0.19703746732500727,
+      "grad_norm": 0.36302104592323303,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 1696
+    },
+    {
+      "epoch": 0.19715364507696775,
+      "grad_norm": 0.3904525637626648,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 1697
+    },
+    {
+      "epoch": 0.19726982282892827,
+      "grad_norm": 0.38676998019218445,
+      "learning_rate": 0.0001,
+      "loss": 1.8969,
+      "step": 1698
+    },
+    {
+      "epoch": 0.19738600058088876,
+      "grad_norm": 0.3899071514606476,
+      "learning_rate": 0.0001,
+      "loss": 1.8044,
+      "step": 1699
+    },
+    {
+      "epoch": 0.19750217833284925,
+      "grad_norm": 0.38243070244789124,
+      "learning_rate": 0.0001,
+      "loss": 1.9196,
+      "step": 1700
+    },
+    {
+      "epoch": 0.19761835608480977,
+      "grad_norm": 0.3663628101348877,
+      "learning_rate": 0.0001,
+      "loss": 1.9099,
+      "step": 1701
+    },
+    {
+      "epoch": 0.19773453383677025,
+      "grad_norm": 0.3713766634464264,
+      "learning_rate": 0.0001,
+      "loss": 1.9145,
+      "step": 1702
+    },
+    {
+      "epoch": 0.19785071158873077,
+      "grad_norm": 0.36897388100624084,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 1703
+    },
+    {
+      "epoch": 0.19796688934069126,
+      "grad_norm": 0.36905625462532043,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 1704
+    },
+    {
+      "epoch": 0.19808306709265175,
+      "grad_norm": 0.35968464612960815,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 1705
+    },
+    {
+      "epoch": 0.19819924484461227,
+      "grad_norm": 0.3755891025066376,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 1706
+    },
+    {
+      "epoch": 0.19831542259657275,
+      "grad_norm": 0.3964156210422516,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 1707
+    },
+    {
+      "epoch": 0.19843160034853324,
+      "grad_norm": 0.3733653724193573,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 1708
+    },
+    {
+      "epoch": 0.19854777810049376,
+      "grad_norm": 0.36199355125427246,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 1709
+    },
+    {
+      "epoch": 0.19866395585245425,
+      "grad_norm": 0.37445268034935,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 1710
+    },
+    {
+      "epoch": 0.19878013360441477,
+      "grad_norm": 0.3383803069591522,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 1711
+    },
+    {
+      "epoch": 0.19889631135637525,
+      "grad_norm": 0.3505041301250458,
+      "learning_rate": 0.0001,
+      "loss": 1.7243,
+      "step": 1712
+    },
+    {
+      "epoch": 0.19901248910833574,
+      "grad_norm": 0.3884884715080261,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 1713
+    },
+    {
+      "epoch": 0.19912866686029626,
+      "grad_norm": 0.3483599126338959,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 1714
+    },
+    {
+      "epoch": 0.19924484461225675,
+      "grad_norm": 0.36730775237083435,
+      "learning_rate": 0.0001,
+      "loss": 1.8406,
+      "step": 1715
+    },
+    {
+      "epoch": 0.19936102236421724,
+      "grad_norm": 0.3688085675239563,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 1716
+    },
+    {
+      "epoch": 0.19947720011617776,
+      "grad_norm": 0.3598758578300476,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 1717
+    },
+    {
+      "epoch": 0.19959337786813824,
+      "grad_norm": 0.3627162277698517,
+      "learning_rate": 0.0001,
+      "loss": 1.8198,
+      "step": 1718
+    },
+    {
+      "epoch": 0.19970955562009876,
+      "grad_norm": 0.35776904225349426,
+      "learning_rate": 0.0001,
+      "loss": 1.7034,
+      "step": 1719
+    },
+    {
+      "epoch": 0.19982573337205925,
+      "grad_norm": 0.3551950454711914,
+      "learning_rate": 0.0001,
+      "loss": 1.7584,
+      "step": 1720
+    },
+    {
+      "epoch": 0.19994191112401974,
+      "grad_norm": 0.3886015713214874,
+      "learning_rate": 0.0001,
+      "loss": 1.8489,
+      "step": 1721
+    },
+    {
+      "epoch": 0.20005808887598026,
+      "grad_norm": 0.396438330411911,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 1722
+    },
+    {
+      "epoch": 0.20017426662794074,
+      "grad_norm": 0.38339731097221375,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 1723
+    },
+    {
+      "epoch": 0.20029044437990126,
+      "grad_norm": 0.4310664236545563,
+      "learning_rate": 0.0001,
+      "loss": 1.9913,
+      "step": 1724
+    },
+    {
+      "epoch": 0.20040662213186175,
+      "grad_norm": 0.34265899658203125,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 1725
+    },
+    {
+      "epoch": 0.20052279988382224,
+      "grad_norm": 0.33483079075813293,
+      "learning_rate": 0.0001,
+      "loss": 1.5423,
+      "step": 1726
+    },
+    {
+      "epoch": 0.20063897763578276,
+      "grad_norm": 0.34570086002349854,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 1727
+    },
+    {
+      "epoch": 0.20075515538774324,
+      "grad_norm": 0.36235955357551575,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 1728
+    },
+    {
+      "epoch": 0.20087133313970373,
+      "grad_norm": 0.3711889088153839,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 1729
+    },
+    {
+      "epoch": 0.20098751089166425,
+      "grad_norm": 0.34235769510269165,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 1730
+    },
+    {
+      "epoch": 0.20110368864362474,
+      "grad_norm": 0.35882294178009033,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 1731
+    },
+    {
+      "epoch": 0.20121986639558526,
+      "grad_norm": 0.3647457957267761,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 1732
+    },
+    {
+      "epoch": 0.20133604414754575,
+      "grad_norm": 0.35160166025161743,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 1733
+    },
+    {
+      "epoch": 0.20145222189950623,
+      "grad_norm": 0.38673707842826843,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 1734
+    },
+    {
+      "epoch": 0.20156839965146675,
+      "grad_norm": 0.3855576813220978,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 1735
+    },
+    {
+      "epoch": 0.20168457740342724,
+      "grad_norm": 0.34935352206230164,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 1736
+    },
+    {
+      "epoch": 0.20180075515538776,
+      "grad_norm": 0.3582732379436493,
+      "learning_rate": 0.0001,
+      "loss": 1.7167,
+      "step": 1737
+    },
+    {
+      "epoch": 0.20191693290734825,
+      "grad_norm": 0.35151124000549316,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 1738
+    },
+    {
+      "epoch": 0.20203311065930873,
+      "grad_norm": 0.3719716966152191,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 1739
+    },
+    {
+      "epoch": 0.20214928841126925,
+      "grad_norm": 0.35581451654434204,
+      "learning_rate": 0.0001,
+      "loss": 1.7917,
+      "step": 1740
+    },
+    {
+      "epoch": 0.20226546616322974,
+      "grad_norm": 0.3818107545375824,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 1741
+    },
+    {
+      "epoch": 0.20238164391519023,
+      "grad_norm": 0.3568393588066101,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 1742
+    },
+    {
+      "epoch": 0.20249782166715075,
+      "grad_norm": 0.38510867953300476,
+      "learning_rate": 0.0001,
+      "loss": 1.8498,
+      "step": 1743
+    },
+    {
+      "epoch": 0.20261399941911123,
+      "grad_norm": 0.36341622471809387,
+      "learning_rate": 0.0001,
+      "loss": 1.7387,
+      "step": 1744
+    },
+    {
+      "epoch": 0.20273017717107175,
+      "grad_norm": 0.38483190536499023,
+      "learning_rate": 0.0001,
+      "loss": 1.8348,
+      "step": 1745
+    },
+    {
+      "epoch": 0.20284635492303224,
+      "grad_norm": 0.3468552529811859,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 1746
+    },
+    {
+      "epoch": 0.20296253267499273,
+      "grad_norm": 0.388603150844574,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 1747
+    },
+    {
+      "epoch": 0.20307871042695325,
+      "grad_norm": 0.3520529866218567,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 1748
+    },
+    {
+      "epoch": 0.20319488817891374,
+      "grad_norm": 0.39672860503196716,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 1749
+    },
+    {
+      "epoch": 0.20331106593087422,
+      "grad_norm": 0.3811289668083191,
+      "learning_rate": 0.0001,
+      "loss": 1.6476,
+      "step": 1750
+    },
+    {
+      "epoch": 0.20342724368283474,
+      "grad_norm": 0.3765932321548462,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 1751
+    },
+    {
+      "epoch": 0.20354342143479523,
+      "grad_norm": 0.33500736951828003,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 1752
+    },
+    {
+      "epoch": 0.20365959918675575,
+      "grad_norm": 0.3680993318557739,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 1753
+    },
+    {
+      "epoch": 0.20377577693871624,
+      "grad_norm": 0.3892023265361786,
+      "learning_rate": 0.0001,
+      "loss": 1.8349,
+      "step": 1754
+    },
+    {
+      "epoch": 0.20389195469067672,
+      "grad_norm": 0.3450574576854706,
+      "learning_rate": 0.0001,
+      "loss": 1.7574,
+      "step": 1755
+    },
+    {
+      "epoch": 0.20400813244263724,
+      "grad_norm": 0.37310826778411865,
+      "learning_rate": 0.0001,
+      "loss": 1.8353,
+      "step": 1756
+    },
+    {
+      "epoch": 0.20412431019459773,
+      "grad_norm": 0.3621986210346222,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 1757
+    },
+    {
+      "epoch": 0.20424048794655825,
+      "grad_norm": 0.3675428628921509,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 1758
+    },
+    {
+      "epoch": 0.20435666569851874,
+      "grad_norm": 0.3777763843536377,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 1759
+    },
+    {
+      "epoch": 0.20447284345047922,
+      "grad_norm": 0.3660062849521637,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 1760
+    },
+    {
+      "epoch": 0.20458902120243974,
+      "grad_norm": 0.3744828402996063,
+      "learning_rate": 0.0001,
+      "loss": 1.6227,
+      "step": 1761
+    },
+    {
+      "epoch": 0.20470519895440023,
+      "grad_norm": 0.3762773275375366,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 1762
+    },
+    {
+      "epoch": 0.20482137670636072,
+      "grad_norm": 0.3429649770259857,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 1763
+    },
+    {
+      "epoch": 0.20493755445832124,
+      "grad_norm": 0.35902631282806396,
+      "learning_rate": 0.0001,
+      "loss": 1.7535,
+      "step": 1764
+    },
+    {
+      "epoch": 0.20505373221028173,
+      "grad_norm": 0.37041351199150085,
+      "learning_rate": 0.0001,
+      "loss": 1.8098,
+      "step": 1765
+    },
+    {
+      "epoch": 0.20516990996224224,
+      "grad_norm": 0.3608233332633972,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 1766
+    },
+    {
+      "epoch": 0.20528608771420273,
+      "grad_norm": 0.39614373445510864,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 1767
+    },
+    {
+      "epoch": 0.20540226546616322,
+      "grad_norm": 0.3881773054599762,
+      "learning_rate": 0.0001,
+      "loss": 1.7105,
+      "step": 1768
+    },
+    {
+      "epoch": 0.20551844321812374,
+      "grad_norm": 0.3501490354537964,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 1769
+    },
+    {
+      "epoch": 0.20563462097008423,
+      "grad_norm": 0.36642521619796753,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 1770
+    },
+    {
+      "epoch": 0.20575079872204474,
+      "grad_norm": 0.3428248167037964,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 1771
+    },
+    {
+      "epoch": 0.20586697647400523,
+      "grad_norm": 0.39081087708473206,
+      "learning_rate": 0.0001,
+      "loss": 1.7507,
+      "step": 1772
+    },
+    {
+      "epoch": 0.20598315422596572,
+      "grad_norm": 0.3991013467311859,
+      "learning_rate": 0.0001,
+      "loss": 1.9912,
+      "step": 1773
+    },
+    {
+      "epoch": 0.20609933197792624,
+      "grad_norm": 0.3668970763683319,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 1774
+    },
+    {
+      "epoch": 0.20621550972988673,
+      "grad_norm": 0.36014431715011597,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 1775
+    },
+    {
+      "epoch": 0.20633168748184721,
+      "grad_norm": 0.38143759965896606,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 1776
+    },
+    {
+      "epoch": 0.20644786523380773,
+      "grad_norm": 0.382790207862854,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 1777
+    },
+    {
+      "epoch": 0.20656404298576822,
+      "grad_norm": 0.355749249458313,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 1778
+    },
+    {
+      "epoch": 0.20668022073772874,
+      "grad_norm": 0.3811320662498474,
+      "learning_rate": 0.0001,
+      "loss": 1.8527,
+      "step": 1779
+    },
+    {
+      "epoch": 0.20679639848968923,
+      "grad_norm": 0.3613418638706207,
+      "learning_rate": 0.0001,
+      "loss": 1.7679,
+      "step": 1780
+    },
+    {
+      "epoch": 0.20691257624164971,
+      "grad_norm": 0.36926668882369995,
+      "learning_rate": 0.0001,
+      "loss": 1.7807,
+      "step": 1781
+    },
+    {
+      "epoch": 0.20702875399361023,
+      "grad_norm": 0.3642374277114868,
+      "learning_rate": 0.0001,
+      "loss": 1.6741,
+      "step": 1782
+    },
+    {
+      "epoch": 0.20714493174557072,
+      "grad_norm": 0.4205605685710907,
+      "learning_rate": 0.0001,
+      "loss": 1.8032,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2072611094975312,
+      "grad_norm": 0.3942674994468689,
+      "learning_rate": 0.0001,
+      "loss": 1.7935,
+      "step": 1784
+    },
+    {
+      "epoch": 0.20737728724949173,
+      "grad_norm": 0.3895415663719177,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 1785
+    },
+    {
+      "epoch": 0.20749346500145222,
+      "grad_norm": 0.3891303539276123,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 1786
+    },
+    {
+      "epoch": 0.20760964275341273,
+      "grad_norm": 0.3562510311603546,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 1787
+    },
+    {
+      "epoch": 0.20772582050537322,
+      "grad_norm": 0.35545283555984497,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 1788
+    },
+    {
+      "epoch": 0.2078419982573337,
+      "grad_norm": 0.40169668197631836,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 1789
+    },
+    {
+      "epoch": 0.20795817600929423,
+      "grad_norm": 0.36041274666786194,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 1790
+    },
+    {
+      "epoch": 0.20807435376125472,
+      "grad_norm": 0.36906883120536804,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 1791
+    },
+    {
+      "epoch": 0.20819053151321523,
+      "grad_norm": 0.35365477204322815,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 1792
+    },
+    {
+      "epoch": 0.20830670926517572,
+      "grad_norm": 0.3502259850502014,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 1793
+    },
+    {
+      "epoch": 0.2084228870171362,
+      "grad_norm": 0.3471246659755707,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 1794
+    },
+    {
+      "epoch": 0.20853906476909673,
+      "grad_norm": 0.36685308814048767,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 1795
+    },
+    {
+      "epoch": 0.20865524252105722,
+      "grad_norm": 0.40070536732673645,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2087714202730177,
+      "grad_norm": 0.35480430722236633,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 1797
+    },
+    {
+      "epoch": 0.20888759802497822,
+      "grad_norm": 0.35834869742393494,
+      "learning_rate": 0.0001,
+      "loss": 1.8237,
+      "step": 1798
+    },
+    {
+      "epoch": 0.2090037757769387,
+      "grad_norm": 0.521961510181427,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 1799
+    },
+    {
+      "epoch": 0.20911995352889923,
+      "grad_norm": 0.36445826292037964,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 1800
+    },
+    {
+      "epoch": 0.20923613128085972,
+      "grad_norm": 0.3719666302204132,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2093523090328202,
+      "grad_norm": 0.37383589148521423,
+      "learning_rate": 0.0001,
+      "loss": 1.8157,
+      "step": 1802
+    },
+    {
+      "epoch": 0.20946848678478072,
+      "grad_norm": 0.3597404956817627,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2095846645367412,
+      "grad_norm": 0.3522038161754608,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 1804
+    },
+    {
+      "epoch": 0.2097008422887017,
+      "grad_norm": 0.37592774629592896,
+      "learning_rate": 0.0001,
+      "loss": 1.7834,
+      "step": 1805
+    },
+    {
+      "epoch": 0.20981702004066222,
+      "grad_norm": 0.3839535117149353,
+      "learning_rate": 0.0001,
+      "loss": 1.7384,
+      "step": 1806
+    },
+    {
+      "epoch": 0.2099331977926227,
+      "grad_norm": 0.3627369999885559,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 1807
+    },
+    {
+      "epoch": 0.21004937554458322,
+      "grad_norm": 0.36218005418777466,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 1808
+    },
+    {
+      "epoch": 0.2101655532965437,
+      "grad_norm": 0.3566618263721466,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2102817310485042,
+      "grad_norm": 0.3671146631240845,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 1810
+    },
+    {
+      "epoch": 0.21039790880046472,
+      "grad_norm": 0.3826962113380432,
+      "learning_rate": 0.0001,
+      "loss": 1.6545,
+      "step": 1811
+    },
+    {
+      "epoch": 0.2105140865524252,
+      "grad_norm": 0.36169835925102234,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 1812
+    },
+    {
+      "epoch": 0.21063026430438572,
+      "grad_norm": 0.3707777261734009,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2107464420563462,
+      "grad_norm": 0.3656941056251526,
+      "learning_rate": 0.0001,
+      "loss": 1.704,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2108626198083067,
+      "grad_norm": 0.3701861500740051,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 1815
+    },
+    {
+      "epoch": 0.21097879756026722,
+      "grad_norm": 0.37783530354499817,
+      "learning_rate": 0.0001,
+      "loss": 1.7791,
+      "step": 1816
+    },
+    {
+      "epoch": 0.2110949753122277,
+      "grad_norm": 0.3402620255947113,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2112111530641882,
+      "grad_norm": 0.3584825098514557,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 1818
+    },
+    {
+      "epoch": 0.2113273308161487,
+      "grad_norm": 0.37935203313827515,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2114435085681092,
+      "grad_norm": 0.3971646726131439,
+      "learning_rate": 0.0001,
+      "loss": 1.9348,
+      "step": 1820
+    },
+    {
+      "epoch": 0.21155968632006972,
+      "grad_norm": 0.3793870210647583,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 1821
+    },
+    {
+      "epoch": 0.2116758640720302,
+      "grad_norm": 0.34898293018341064,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 1822
+    },
+    {
+      "epoch": 0.2117920418239907,
+      "grad_norm": 0.3546793758869171,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 1823
+    },
+    {
+      "epoch": 0.2119082195759512,
+      "grad_norm": 0.39553505182266235,
+      "learning_rate": 0.0001,
+      "loss": 1.7607,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2120243973279117,
+      "grad_norm": 0.38580450415611267,
+      "learning_rate": 0.0001,
+      "loss": 1.7025,
+      "step": 1825
+    },
+    {
+      "epoch": 0.21214057507987222,
+      "grad_norm": 0.3958745300769806,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 1826
+    },
+    {
+      "epoch": 0.2122567528318327,
+      "grad_norm": 0.3790923058986664,
+      "learning_rate": 0.0001,
+      "loss": 1.7912,
+      "step": 1827
+    },
+    {
+      "epoch": 0.2123729305837932,
+      "grad_norm": 0.3835214376449585,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 1828
+    },
+    {
+      "epoch": 0.2124891083357537,
+      "grad_norm": 0.36152535676956177,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2126052860877142,
+      "grad_norm": 0.3254132866859436,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 1830
+    },
+    {
+      "epoch": 0.2127214638396747,
+      "grad_norm": 0.38130825757980347,
+      "learning_rate": 0.0001,
+      "loss": 1.8067,
+      "step": 1831
+    },
+    {
+      "epoch": 0.2128376415916352,
+      "grad_norm": 0.36673715710639954,
+      "learning_rate": 0.0001,
+      "loss": 1.7599,
+      "step": 1832
+    },
+    {
+      "epoch": 0.2129538193435957,
+      "grad_norm": 0.39289578795433044,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 1833
+    },
+    {
+      "epoch": 0.2130699970955562,
+      "grad_norm": 0.35290926694869995,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2131861748475167,
+      "grad_norm": 0.36013439297676086,
+      "learning_rate": 0.0001,
+      "loss": 1.8015,
+      "step": 1835
+    },
+    {
+      "epoch": 0.2133023525994772,
+      "grad_norm": 0.39623382687568665,
+      "learning_rate": 0.0001,
+      "loss": 1.7266,
+      "step": 1836
+    },
+    {
+      "epoch": 0.2134185303514377,
+      "grad_norm": 0.37049758434295654,
+      "learning_rate": 0.0001,
+      "loss": 1.8266,
+      "step": 1837
+    },
+    {
+      "epoch": 0.2135347081033982,
+      "grad_norm": 0.37480175495147705,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 1838
+    },
+    {
+      "epoch": 0.21365088585535869,
+      "grad_norm": 0.38570842146873474,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2137670636073192,
+      "grad_norm": 0.4038733243942261,
+      "learning_rate": 0.0001,
+      "loss": 1.7587,
+      "step": 1840
+    },
+    {
+      "epoch": 0.2138832413592797,
+      "grad_norm": 0.3932957351207733,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 1841
+    },
+    {
+      "epoch": 0.2139994191112402,
+      "grad_norm": 0.3737775981426239,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 1842
+    },
+    {
+      "epoch": 0.2141155968632007,
+      "grad_norm": 0.39853671193122864,
+      "learning_rate": 0.0001,
+      "loss": 1.7764,
+      "step": 1843
+    },
+    {
+      "epoch": 0.21423177461516119,
+      "grad_norm": 0.37162113189697266,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2143479523671217,
+      "grad_norm": 0.38657307624816895,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 1845
+    },
+    {
+      "epoch": 0.2144641301190822,
+      "grad_norm": 0.40344858169555664,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 1846
+    },
+    {
+      "epoch": 0.2145803078710427,
+      "grad_norm": 0.34257638454437256,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 1847
+    },
+    {
+      "epoch": 0.2146964856230032,
+      "grad_norm": 0.3880448043346405,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 1848
+    },
+    {
+      "epoch": 0.21481266337496369,
+      "grad_norm": 0.3583984971046448,
+      "learning_rate": 0.0001,
+      "loss": 1.596,
+      "step": 1849
+    },
+    {
+      "epoch": 0.2149288411269242,
+      "grad_norm": 0.36240407824516296,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 1850
+    },
+    {
+      "epoch": 0.2150450188788847,
+      "grad_norm": 0.3811880648136139,
+      "learning_rate": 0.0001,
+      "loss": 1.7082,
+      "step": 1851
+    },
+    {
+      "epoch": 0.21516119663084518,
+      "grad_norm": 0.4095620810985565,
+      "learning_rate": 0.0001,
+      "loss": 1.8439,
+      "step": 1852
+    },
+    {
+      "epoch": 0.2152773743828057,
+      "grad_norm": 0.43665236234664917,
+      "learning_rate": 0.0001,
+      "loss": 1.8788,
+      "step": 1853
+    },
+    {
+      "epoch": 0.2153935521347662,
+      "grad_norm": 0.37620803713798523,
+      "learning_rate": 0.0001,
+      "loss": 1.7409,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2155097298867267,
+      "grad_norm": 0.3814584016799927,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 1855
+    },
+    {
+      "epoch": 0.2156259076386872,
+      "grad_norm": 0.349738210439682,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 1856
+    },
+    {
+      "epoch": 0.21574208539064768,
+      "grad_norm": 0.388741135597229,
+      "learning_rate": 0.0001,
+      "loss": 1.8032,
+      "step": 1857
+    },
+    {
+      "epoch": 0.2158582631426082,
+      "grad_norm": 0.37817415595054626,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 1858
+    },
+    {
+      "epoch": 0.2159744408945687,
+      "grad_norm": 0.3564184606075287,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2160906186465292,
+      "grad_norm": 0.3882390260696411,
+      "learning_rate": 0.0001,
+      "loss": 1.8215,
+      "step": 1860
+    },
+    {
+      "epoch": 0.2162067963984897,
+      "grad_norm": 0.3727805018424988,
+      "learning_rate": 0.0001,
+      "loss": 1.7673,
+      "step": 1861
+    },
+    {
+      "epoch": 0.21632297415045018,
+      "grad_norm": 0.3581012785434723,
+      "learning_rate": 0.0001,
+      "loss": 1.741,
+      "step": 1862
+    },
+    {
+      "epoch": 0.2164391519024107,
+      "grad_norm": 0.4040369391441345,
+      "learning_rate": 0.0001,
+      "loss": 1.8377,
+      "step": 1863
+    },
+    {
+      "epoch": 0.2165553296543712,
+      "grad_norm": 0.3850097060203552,
+      "learning_rate": 0.0001,
+      "loss": 1.7752,
+      "step": 1864
+    },
+    {
+      "epoch": 0.21667150740633168,
+      "grad_norm": 0.3745688796043396,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 1865
+    },
+    {
+      "epoch": 0.2167876851582922,
+      "grad_norm": 0.3721669018268585,
+      "learning_rate": 0.0001,
+      "loss": 1.7012,
+      "step": 1866
+    },
+    {
+      "epoch": 0.21690386291025268,
+      "grad_norm": 0.3597140610218048,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 1867
+    },
+    {
+      "epoch": 0.2170200406622132,
+      "grad_norm": 0.3266042470932007,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 1868
+    },
+    {
+      "epoch": 0.2171362184141737,
+      "grad_norm": 0.3769254982471466,
+      "learning_rate": 0.0001,
+      "loss": 1.9181,
+      "step": 1869
+    },
+    {
+      "epoch": 0.21725239616613418,
+      "grad_norm": 0.386996328830719,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 1870
+    },
+    {
+      "epoch": 0.2173685739180947,
+      "grad_norm": 0.3945530354976654,
+      "learning_rate": 0.0001,
+      "loss": 1.7364,
+      "step": 1871
+    },
+    {
+      "epoch": 0.21748475167005518,
+      "grad_norm": 0.39462509751319885,
+      "learning_rate": 0.0001,
+      "loss": 1.763,
+      "step": 1872
+    },
+    {
+      "epoch": 0.21760092942201567,
+      "grad_norm": 0.35269200801849365,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 1873
+    },
+    {
+      "epoch": 0.2177171071739762,
+      "grad_norm": 0.401889443397522,
+      "learning_rate": 0.0001,
+      "loss": 1.9107,
+      "step": 1874
+    },
+    {
+      "epoch": 0.21783328492593668,
+      "grad_norm": 0.3577077090740204,
+      "learning_rate": 0.0001,
+      "loss": 1.7043,
+      "step": 1875
+    },
+    {
+      "epoch": 0.2179494626778972,
+      "grad_norm": 0.3647770881652832,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 1876
+    },
+    {
+      "epoch": 0.21806564042985768,
+      "grad_norm": 0.37258070707321167,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 1877
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 0.38503706455230713,
+      "learning_rate": 0.0001,
+      "loss": 1.7757,
+      "step": 1878
+    },
+    {
+      "epoch": 0.2182979959337787,
+      "grad_norm": 0.3539654314517975,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 1879
+    },
+    {
+      "epoch": 0.21841417368573918,
+      "grad_norm": 0.3822212517261505,
+      "learning_rate": 0.0001,
+      "loss": 1.8839,
+      "step": 1880
+    },
+    {
+      "epoch": 0.2185303514376997,
+      "grad_norm": 0.34567397832870483,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 1881
+    },
+    {
+      "epoch": 0.21864652918966018,
+      "grad_norm": 0.359351247549057,
+      "learning_rate": 0.0001,
+      "loss": 1.6512,
+      "step": 1882
+    },
+    {
+      "epoch": 0.21876270694162067,
+      "grad_norm": 0.3686642348766327,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 1883
+    },
+    {
+      "epoch": 0.2188788846935812,
+      "grad_norm": 0.37803003191947937,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 1884
+    },
+    {
+      "epoch": 0.21899506244554168,
+      "grad_norm": 0.3631937503814697,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 1885
+    },
+    {
+      "epoch": 0.21911124019750217,
+      "grad_norm": 0.3699375092983246,
+      "learning_rate": 0.0001,
+      "loss": 1.7432,
+      "step": 1886
+    },
+    {
+      "epoch": 0.21922741794946268,
+      "grad_norm": 0.3758851885795593,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 1887
+    },
+    {
+      "epoch": 0.21934359570142317,
+      "grad_norm": 0.38159826397895813,
+      "learning_rate": 0.0001,
+      "loss": 1.6982,
+      "step": 1888
+    },
+    {
+      "epoch": 0.2194597734533837,
+      "grad_norm": 0.3970431983470917,
+      "learning_rate": 0.0001,
+      "loss": 1.7361,
+      "step": 1889
+    },
+    {
+      "epoch": 0.21957595120534418,
+      "grad_norm": 0.5288736820220947,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 1890
+    },
+    {
+      "epoch": 0.21969212895730467,
+      "grad_norm": 0.3814886212348938,
+      "learning_rate": 0.0001,
+      "loss": 1.7492,
+      "step": 1891
+    },
+    {
+      "epoch": 0.21980830670926518,
+      "grad_norm": 0.3462470471858978,
+      "learning_rate": 0.0001,
+      "loss": 1.5657,
+      "step": 1892
+    },
+    {
+      "epoch": 0.21992448446122567,
+      "grad_norm": 0.41686296463012695,
+      "learning_rate": 0.0001,
+      "loss": 1.7641,
+      "step": 1893
+    },
+    {
+      "epoch": 0.22004066221318616,
+      "grad_norm": 0.35188373923301697,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 1894
+    },
+    {
+      "epoch": 0.22015683996514668,
+      "grad_norm": 0.366418719291687,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 1895
+    },
+    {
+      "epoch": 0.22027301771710717,
+      "grad_norm": 0.3827008306980133,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 1896
+    },
+    {
+      "epoch": 0.22038919546906768,
+      "grad_norm": 0.38810524344444275,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 1897
+    },
+    {
+      "epoch": 0.22050537322102817,
+      "grad_norm": 0.3892178237438202,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 1898
+    },
+    {
+      "epoch": 0.22062155097298866,
+      "grad_norm": 0.34898847341537476,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 1899
+    },
+    {
+      "epoch": 0.22073772872494918,
+      "grad_norm": 0.38347798585891724,
+      "learning_rate": 0.0001,
+      "loss": 1.778,
+      "step": 1900
+    },
+    {
+      "epoch": 0.22085390647690967,
+      "grad_norm": 0.45999932289123535,
+      "learning_rate": 0.0001,
+      "loss": 1.8044,
+      "step": 1901
+    },
+    {
+      "epoch": 0.22097008422887018,
+      "grad_norm": 0.39396339654922485,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 1902
+    },
+    {
+      "epoch": 0.22108626198083067,
+      "grad_norm": 0.3777587413787842,
+      "learning_rate": 0.0001,
+      "loss": 1.885,
+      "step": 1903
+    },
+    {
+      "epoch": 0.22120243973279116,
+      "grad_norm": 0.38406991958618164,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 1904
+    },
+    {
+      "epoch": 0.22131861748475168,
+      "grad_norm": 0.4187374413013458,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 1905
+    },
+    {
+      "epoch": 0.22143479523671217,
+      "grad_norm": 0.35657206177711487,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 1906
+    },
+    {
+      "epoch": 0.22155097298867266,
+      "grad_norm": 0.3913048803806305,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 1907
+    },
+    {
+      "epoch": 0.22166715074063317,
+      "grad_norm": 0.38149452209472656,
+      "learning_rate": 0.0001,
+      "loss": 1.7873,
+      "step": 1908
+    },
+    {
+      "epoch": 0.22178332849259366,
+      "grad_norm": 0.3800641596317291,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 1909
+    },
+    {
+      "epoch": 0.22189950624455418,
+      "grad_norm": 0.3555182218551636,
+      "learning_rate": 0.0001,
+      "loss": 1.5213,
+      "step": 1910
+    },
+    {
+      "epoch": 0.22201568399651467,
+      "grad_norm": 0.3945232331752777,
+      "learning_rate": 0.0001,
+      "loss": 1.807,
+      "step": 1911
+    },
+    {
+      "epoch": 0.22213186174847516,
+      "grad_norm": 0.38593828678131104,
+      "learning_rate": 0.0001,
+      "loss": 1.8026,
+      "step": 1912
+    },
+    {
+      "epoch": 0.22224803950043567,
+      "grad_norm": 0.38784652948379517,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 1913
+    },
+    {
+      "epoch": 0.22236421725239616,
+      "grad_norm": 0.35396912693977356,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 1914
+    },
+    {
+      "epoch": 0.22248039500435668,
+      "grad_norm": 0.3825778365135193,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 1915
+    },
+    {
+      "epoch": 0.22259657275631717,
+      "grad_norm": 0.36573922634124756,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 1916
+    },
+    {
+      "epoch": 0.22271275050827766,
+      "grad_norm": 0.37928685545921326,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 1917
+    },
+    {
+      "epoch": 0.22282892826023817,
+      "grad_norm": 0.3511602580547333,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 1918
+    },
+    {
+      "epoch": 0.22294510601219866,
+      "grad_norm": 0.3809853494167328,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 1919
+    },
+    {
+      "epoch": 0.22306128376415915,
+      "grad_norm": 0.36619848012924194,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 1920
+    },
+    {
+      "epoch": 0.22317746151611967,
+      "grad_norm": 0.3719876706600189,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 1921
+    },
+    {
+      "epoch": 0.22329363926808016,
+      "grad_norm": 0.3887604773044586,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 1922
+    },
+    {
+      "epoch": 0.22340981702004067,
+      "grad_norm": 0.36279502511024475,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 1923
+    },
+    {
+      "epoch": 0.22352599477200116,
+      "grad_norm": 0.3789854645729065,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 1924
+    },
+    {
+      "epoch": 0.22364217252396165,
+      "grad_norm": 0.3797144293785095,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 1925
+    },
+    {
+      "epoch": 0.22375835027592217,
+      "grad_norm": 0.35952475666999817,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 1926
+    },
+    {
+      "epoch": 0.22387452802788266,
+      "grad_norm": 0.3811299800872803,
+      "learning_rate": 0.0001,
+      "loss": 1.9059,
+      "step": 1927
+    },
+    {
+      "epoch": 0.22399070577984315,
+      "grad_norm": 0.3837408721446991,
+      "learning_rate": 0.0001,
+      "loss": 1.7887,
+      "step": 1928
+    },
+    {
+      "epoch": 0.22410688353180366,
+      "grad_norm": 0.4026016891002655,
+      "learning_rate": 0.0001,
+      "loss": 1.8315,
+      "step": 1929
+    },
+    {
+      "epoch": 0.22422306128376415,
+      "grad_norm": 0.35901689529418945,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 1930
+    },
+    {
+      "epoch": 0.22433923903572467,
+      "grad_norm": 0.3816909193992615,
+      "learning_rate": 0.0001,
+      "loss": 1.7365,
+      "step": 1931
+    },
+    {
+      "epoch": 0.22445541678768516,
+      "grad_norm": 0.3439340889453888,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 1932
+    },
+    {
+      "epoch": 0.22457159453964565,
+      "grad_norm": 0.37772074341773987,
+      "learning_rate": 0.0001,
+      "loss": 1.7376,
+      "step": 1933
+    },
+    {
+      "epoch": 0.22468777229160616,
+      "grad_norm": 0.3728993833065033,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 1934
+    },
+    {
+      "epoch": 0.22480395004356665,
+      "grad_norm": 0.3640214502811432,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 1935
+    },
+    {
+      "epoch": 0.22492012779552717,
+      "grad_norm": 0.37045779824256897,
+      "learning_rate": 0.0001,
+      "loss": 1.8218,
+      "step": 1936
+    },
+    {
+      "epoch": 0.22503630554748766,
+      "grad_norm": 0.3745948374271393,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 1937
+    },
+    {
+      "epoch": 0.22515248329944815,
+      "grad_norm": 0.36098146438598633,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 1938
+    },
+    {
+      "epoch": 0.22526866105140866,
+      "grad_norm": 0.3912782073020935,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 1939
+    },
+    {
+      "epoch": 0.22538483880336915,
+      "grad_norm": 0.35821422934532166,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 1940
+    },
+    {
+      "epoch": 0.22550101655532964,
+      "grad_norm": 0.3956640958786011,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 1941
+    },
+    {
+      "epoch": 0.22561719430729016,
+      "grad_norm": 0.3759971559047699,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 1942
+    },
+    {
+      "epoch": 0.22573337205925065,
+      "grad_norm": 0.42271122336387634,
+      "learning_rate": 0.0001,
+      "loss": 1.8635,
+      "step": 1943
+    },
+    {
+      "epoch": 0.22584954981121116,
+      "grad_norm": 0.350124329328537,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 1944
+    },
+    {
+      "epoch": 0.22596572756317165,
+      "grad_norm": 0.35673707723617554,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 1945
+    },
+    {
+      "epoch": 0.22608190531513214,
+      "grad_norm": 0.3845730721950531,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 1946
+    },
+    {
+      "epoch": 0.22619808306709266,
+      "grad_norm": 0.3668268620967865,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 1947
+    },
+    {
+      "epoch": 0.22631426081905315,
+      "grad_norm": 0.36193057894706726,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 1948
+    },
+    {
+      "epoch": 0.22643043857101366,
+      "grad_norm": 0.3776628077030182,
+      "learning_rate": 0.0001,
+      "loss": 1.7175,
+      "step": 1949
+    },
+    {
+      "epoch": 0.22654661632297415,
+      "grad_norm": 0.35988494753837585,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 1950
+    },
+    {
+      "epoch": 0.22666279407493464,
+      "grad_norm": 0.3582817614078522,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 1951
+    },
+    {
+      "epoch": 0.22677897182689516,
+      "grad_norm": 0.34842002391815186,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 1952
+    },
+    {
+      "epoch": 0.22689514957885565,
+      "grad_norm": 0.3792864978313446,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 1953
+    },
+    {
+      "epoch": 0.22701132733081614,
+      "grad_norm": 0.3827436566352844,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 1954
+    },
+    {
+      "epoch": 0.22712750508277665,
+      "grad_norm": 0.3784099817276001,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 1955
+    },
+    {
+      "epoch": 0.22724368283473714,
+      "grad_norm": 0.3828442096710205,
+      "learning_rate": 0.0001,
+      "loss": 1.7672,
+      "step": 1956
+    },
+    {
+      "epoch": 0.22735986058669766,
+      "grad_norm": 0.3599461019039154,
+      "learning_rate": 0.0001,
+      "loss": 1.7539,
+      "step": 1957
+    },
+    {
+      "epoch": 0.22747603833865815,
+      "grad_norm": 0.3842355012893677,
+      "learning_rate": 0.0001,
+      "loss": 1.8452,
+      "step": 1958
+    },
+    {
+      "epoch": 0.22759221609061864,
+      "grad_norm": 0.375945508480072,
+      "learning_rate": 0.0001,
+      "loss": 1.8502,
+      "step": 1959
+    },
+    {
+      "epoch": 0.22770839384257915,
+      "grad_norm": 0.4141533374786377,
+      "learning_rate": 0.0001,
+      "loss": 1.7301,
+      "step": 1960
+    },
+    {
+      "epoch": 0.22782457159453964,
+      "grad_norm": 0.37964141368865967,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 1961
+    },
+    {
+      "epoch": 0.22794074934650013,
+      "grad_norm": 0.39130285382270813,
+      "learning_rate": 0.0001,
+      "loss": 1.8114,
+      "step": 1962
+    },
+    {
+      "epoch": 0.22805692709846065,
+      "grad_norm": 0.3492740988731384,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 1963
+    },
+    {
+      "epoch": 0.22817310485042114,
+      "grad_norm": 0.36972182989120483,
+      "learning_rate": 0.0001,
+      "loss": 1.6917,
+      "step": 1964
+    },
+    {
+      "epoch": 0.22828928260238165,
+      "grad_norm": 0.3815283179283142,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 1965
+    },
+    {
+      "epoch": 0.22840546035434214,
+      "grad_norm": 0.3695911169052124,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 1966
+    },
+    {
+      "epoch": 0.22852163810630263,
+      "grad_norm": 0.3688741624355316,
+      "learning_rate": 0.0001,
+      "loss": 1.8406,
+      "step": 1967
+    },
+    {
+      "epoch": 0.22863781585826315,
+      "grad_norm": 0.3708171248435974,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 1968
+    },
+    {
+      "epoch": 0.22875399361022364,
+      "grad_norm": 0.406654417514801,
+      "learning_rate": 0.0001,
+      "loss": 1.8354,
+      "step": 1969
+    },
+    {
+      "epoch": 0.22887017136218415,
+      "grad_norm": 0.3633515536785126,
+      "learning_rate": 0.0001,
+      "loss": 1.6008,
+      "step": 1970
+    },
+    {
+      "epoch": 0.22898634911414464,
+      "grad_norm": 0.37781691551208496,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 1971
+    },
+    {
+      "epoch": 0.22910252686610513,
+      "grad_norm": 0.36972054839134216,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 1972
+    },
+    {
+      "epoch": 0.22921870461806565,
+      "grad_norm": 0.40130361914634705,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 1973
+    },
+    {
+      "epoch": 0.22933488237002614,
+      "grad_norm": 0.38525599241256714,
+      "learning_rate": 0.0001,
+      "loss": 1.7806,
+      "step": 1974
+    },
+    {
+      "epoch": 0.22945106012198663,
+      "grad_norm": 0.38195157051086426,
+      "learning_rate": 0.0001,
+      "loss": 1.6819,
+      "step": 1975
+    },
+    {
+      "epoch": 0.22956723787394714,
+      "grad_norm": 0.36674100160598755,
+      "learning_rate": 0.0001,
+      "loss": 1.5677,
+      "step": 1976
+    },
+    {
+      "epoch": 0.22968341562590763,
+      "grad_norm": 0.37068042159080505,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 1977
+    },
+    {
+      "epoch": 0.22979959337786815,
+      "grad_norm": 0.35614022612571716,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 1978
+    },
+    {
+      "epoch": 0.22991577112982864,
+      "grad_norm": 0.37935879826545715,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 1979
+    },
+    {
+      "epoch": 0.23003194888178913,
+      "grad_norm": 0.3637298047542572,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 1980
+    },
+    {
+      "epoch": 0.23014812663374964,
+      "grad_norm": 0.40035533905029297,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 1981
+    },
+    {
+      "epoch": 0.23026430438571013,
+      "grad_norm": 0.40497007966041565,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 1982
+    },
+    {
+      "epoch": 0.23038048213767062,
+      "grad_norm": 0.3771011233329773,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 1983
+    },
+    {
+      "epoch": 0.23049665988963114,
+      "grad_norm": 0.3938141465187073,
+      "learning_rate": 0.0001,
+      "loss": 1.791,
+      "step": 1984
+    },
+    {
+      "epoch": 0.23061283764159163,
+      "grad_norm": 0.3939863443374634,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 1985
+    },
+    {
+      "epoch": 0.23072901539355214,
+      "grad_norm": 0.3473072052001953,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 1986
+    },
+    {
+      "epoch": 0.23084519314551263,
+      "grad_norm": 0.371954083442688,
+      "learning_rate": 0.0001,
+      "loss": 1.7404,
+      "step": 1987
+    },
+    {
+      "epoch": 0.23096137089747312,
+      "grad_norm": 0.38401806354522705,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 1988
+    },
+    {
+      "epoch": 0.23107754864943364,
+      "grad_norm": 0.36638298630714417,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 1989
+    },
+    {
+      "epoch": 0.23119372640139413,
+      "grad_norm": 0.365355521440506,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 1990
+    },
+    {
+      "epoch": 0.23130990415335465,
+      "grad_norm": 0.3541325330734253,
+      "learning_rate": 0.0001,
+      "loss": 1.6865,
+      "step": 1991
+    },
+    {
+      "epoch": 0.23142608190531513,
+      "grad_norm": 0.3768599033355713,
+      "learning_rate": 0.0001,
+      "loss": 1.8022,
+      "step": 1992
+    },
+    {
+      "epoch": 0.23154225965727562,
+      "grad_norm": 0.36274993419647217,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 1993
+    },
+    {
+      "epoch": 0.23165843740923614,
+      "grad_norm": 0.3988153636455536,
+      "learning_rate": 0.0001,
+      "loss": 1.8395,
+      "step": 1994
+    },
+    {
+      "epoch": 0.23177461516119663,
+      "grad_norm": 0.3925970494747162,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 1995
+    },
+    {
+      "epoch": 0.23189079291315712,
+      "grad_norm": 0.3741264045238495,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 1996
+    },
+    {
+      "epoch": 0.23200697066511763,
+      "grad_norm": 0.3511716425418854,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 1997
+    },
+    {
+      "epoch": 0.23212314841707812,
+      "grad_norm": 0.3565942645072937,
+      "learning_rate": 0.0001,
+      "loss": 1.7113,
+      "step": 1998
+    },
+    {
+      "epoch": 0.23223932616903864,
+      "grad_norm": 0.36350539326667786,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 1999
+    },
+    {
+      "epoch": 0.23235550392099913,
+      "grad_norm": 0.39565804600715637,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 2000
+    },
+    {
+      "epoch": 0.23247168167295962,
+      "grad_norm": 0.39192995429039,
+      "learning_rate": 0.0001,
+      "loss": 1.821,
+      "step": 2001
+    },
+    {
+      "epoch": 0.23258785942492013,
+      "grad_norm": 0.3590277433395386,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 2002
+    },
+    {
+      "epoch": 0.23270403717688062,
+      "grad_norm": 0.3515608310699463,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 2003
+    },
+    {
+      "epoch": 0.23282021492884114,
+      "grad_norm": 0.3733135759830475,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 2004
+    },
+    {
+      "epoch": 0.23293639268080163,
+      "grad_norm": 0.3820706903934479,
+      "learning_rate": 0.0001,
+      "loss": 1.625,
+      "step": 2005
+    },
+    {
+      "epoch": 0.23305257043276212,
+      "grad_norm": 0.38101106882095337,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 2006
+    },
+    {
+      "epoch": 0.23316874818472264,
+      "grad_norm": 0.3717200756072998,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 2007
+    },
+    {
+      "epoch": 0.23328492593668312,
+      "grad_norm": 0.3890402913093567,
+      "learning_rate": 0.0001,
+      "loss": 1.8481,
+      "step": 2008
+    },
+    {
+      "epoch": 0.2334011036886436,
+      "grad_norm": 0.374593585729599,
+      "learning_rate": 0.0001,
+      "loss": 1.6713,
+      "step": 2009
+    },
+    {
+      "epoch": 0.23351728144060413,
+      "grad_norm": 0.37107715010643005,
+      "learning_rate": 0.0001,
+      "loss": 1.6466,
+      "step": 2010
+    },
+    {
+      "epoch": 0.23363345919256462,
+      "grad_norm": 0.37940695881843567,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 2011
+    },
+    {
+      "epoch": 0.23374963694452514,
+      "grad_norm": 0.351849228143692,
+      "learning_rate": 0.0001,
+      "loss": 1.6345,
+      "step": 2012
+    },
+    {
+      "epoch": 0.23386581469648562,
+      "grad_norm": 0.35233619809150696,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2339819924484461,
+      "grad_norm": 0.39142873883247375,
+      "learning_rate": 0.0001,
+      "loss": 1.8342,
+      "step": 2014
+    },
+    {
+      "epoch": 0.23409817020040663,
+      "grad_norm": 0.36893418431282043,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 2015
+    },
+    {
+      "epoch": 0.23421434795236712,
+      "grad_norm": 0.4079015851020813,
+      "learning_rate": 0.0001,
+      "loss": 1.9148,
+      "step": 2016
+    },
+    {
+      "epoch": 0.2343305257043276,
+      "grad_norm": 0.35830172896385193,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 2017
+    },
+    {
+      "epoch": 0.23444670345628812,
+      "grad_norm": 0.35987070202827454,
+      "learning_rate": 0.0001,
+      "loss": 1.5225,
+      "step": 2018
+    },
+    {
+      "epoch": 0.2345628812082486,
+      "grad_norm": 0.3772944211959839,
+      "learning_rate": 0.0001,
+      "loss": 1.7637,
+      "step": 2019
+    },
+    {
+      "epoch": 0.23467905896020913,
+      "grad_norm": 0.36652514338493347,
+      "learning_rate": 0.0001,
+      "loss": 1.786,
+      "step": 2020
+    },
+    {
+      "epoch": 0.23479523671216962,
+      "grad_norm": 0.3576715886592865,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 2021
+    },
+    {
+      "epoch": 0.2349114144641301,
+      "grad_norm": 0.3853584825992584,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 2022
+    },
+    {
+      "epoch": 0.23502759221609063,
+      "grad_norm": 0.3581581115722656,
+      "learning_rate": 0.0001,
+      "loss": 1.7853,
+      "step": 2023
+    },
+    {
+      "epoch": 0.23514376996805111,
+      "grad_norm": 0.396012544631958,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 2024
+    },
+    {
+      "epoch": 0.23525994772001163,
+      "grad_norm": 0.37956202030181885,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 2025
+    },
+    {
+      "epoch": 0.23537612547197212,
+      "grad_norm": 0.377013623714447,
+      "learning_rate": 0.0001,
+      "loss": 1.7764,
+      "step": 2026
+    },
+    {
+      "epoch": 0.2354923032239326,
+      "grad_norm": 0.37704232335090637,
+      "learning_rate": 0.0001,
+      "loss": 1.8591,
+      "step": 2027
+    },
+    {
+      "epoch": 0.23560848097589313,
+      "grad_norm": 0.3992510437965393,
+      "learning_rate": 0.0001,
+      "loss": 2.0132,
+      "step": 2028
+    },
+    {
+      "epoch": 0.23572465872785361,
+      "grad_norm": 0.36997881531715393,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 2029
+    },
+    {
+      "epoch": 0.2358408364798141,
+      "grad_norm": 0.3613041043281555,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 2030
+    },
+    {
+      "epoch": 0.23595701423177462,
+      "grad_norm": 0.36582064628601074,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 2031
+    },
+    {
+      "epoch": 0.2360731919837351,
+      "grad_norm": 0.3395959436893463,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 2032
+    },
+    {
+      "epoch": 0.23618936973569563,
+      "grad_norm": 0.3602873384952545,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 2033
+    },
+    {
+      "epoch": 0.23630554748765611,
+      "grad_norm": 0.3407090902328491,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2364217252396166,
+      "grad_norm": 0.3737788200378418,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 2035
+    },
+    {
+      "epoch": 0.23653790299157712,
+      "grad_norm": 0.3733769953250885,
+      "learning_rate": 0.0001,
+      "loss": 1.7639,
+      "step": 2036
+    },
+    {
+      "epoch": 0.2366540807435376,
+      "grad_norm": 0.3718021512031555,
+      "learning_rate": 0.0001,
+      "loss": 1.8264,
+      "step": 2037
+    },
+    {
+      "epoch": 0.23677025849549813,
+      "grad_norm": 0.33534038066864014,
+      "learning_rate": 0.0001,
+      "loss": 1.4352,
+      "step": 2038
+    },
+    {
+      "epoch": 0.23688643624745861,
+      "grad_norm": 0.4014089107513428,
+      "learning_rate": 0.0001,
+      "loss": 1.8223,
+      "step": 2039
+    },
+    {
+      "epoch": 0.2370026139994191,
+      "grad_norm": 0.37306705117225647,
+      "learning_rate": 0.0001,
+      "loss": 1.8535,
+      "step": 2040
+    },
+    {
+      "epoch": 0.23711879175137962,
+      "grad_norm": 0.36598867177963257,
+      "learning_rate": 0.0001,
+      "loss": 1.7722,
+      "step": 2041
+    },
+    {
+      "epoch": 0.2372349695033401,
+      "grad_norm": 0.3649059236049652,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 2042
+    },
+    {
+      "epoch": 0.2373511472553006,
+      "grad_norm": 0.3907899558544159,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 2043
+    },
+    {
+      "epoch": 0.23746732500726112,
+      "grad_norm": 0.37756413221359253,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 2044
+    },
+    {
+      "epoch": 0.2375835027592216,
+      "grad_norm": 0.3651416599750519,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 2045
+    },
+    {
+      "epoch": 0.23769968051118212,
+      "grad_norm": 0.3615970015525818,
+      "learning_rate": 0.0001,
+      "loss": 1.7378,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2378158582631426,
+      "grad_norm": 0.3712145686149597,
+      "learning_rate": 0.0001,
+      "loss": 1.7052,
+      "step": 2047
+    },
+    {
+      "epoch": 0.2379320360151031,
+      "grad_norm": 0.39952540397644043,
+      "learning_rate": 0.0001,
+      "loss": 1.6936,
+      "step": 2048
+    },
+    {
+      "epoch": 0.23804821376706362,
+      "grad_norm": 0.37439173460006714,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 2049
+    },
+    {
+      "epoch": 0.2381643915190241,
+      "grad_norm": 0.37691599130630493,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 2050
+    },
+    {
+      "epoch": 0.2382805692709846,
+      "grad_norm": 0.3760530352592468,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 2051
+    },
+    {
+      "epoch": 0.2383967470229451,
+      "grad_norm": 0.3865487575531006,
+      "learning_rate": 0.0001,
+      "loss": 1.9395,
+      "step": 2052
+    },
+    {
+      "epoch": 0.2385129247749056,
+      "grad_norm": 0.3658308982849121,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 2053
+    },
+    {
+      "epoch": 0.23862910252686612,
+      "grad_norm": 0.3813599646091461,
+      "learning_rate": 0.0001,
+      "loss": 1.7707,
+      "step": 2054
+    },
+    {
+      "epoch": 0.2387452802788266,
+      "grad_norm": 0.3815939128398895,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 2055
+    },
+    {
+      "epoch": 0.2388614580307871,
+      "grad_norm": 0.3719700872898102,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 2056
+    },
+    {
+      "epoch": 0.2389776357827476,
+      "grad_norm": 0.3591654896736145,
+      "learning_rate": 0.0001,
+      "loss": 1.5548,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2390938135347081,
+      "grad_norm": 0.37177640199661255,
+      "learning_rate": 0.0001,
+      "loss": 1.8142,
+      "step": 2058
+    },
+    {
+      "epoch": 0.23920999128666862,
+      "grad_norm": 0.36674222350120544,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 2059
+    },
+    {
+      "epoch": 0.2393261690386291,
+      "grad_norm": 0.36910292506217957,
+      "learning_rate": 0.0001,
+      "loss": 1.8695,
+      "step": 2060
+    },
+    {
+      "epoch": 0.2394423467905896,
+      "grad_norm": 0.37248843908309937,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 2061
+    },
+    {
+      "epoch": 0.2395585245425501,
+      "grad_norm": 0.3622843027114868,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 2062
+    },
+    {
+      "epoch": 0.2396747022945106,
+      "grad_norm": 0.38794079422950745,
+      "learning_rate": 0.0001,
+      "loss": 1.7049,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2397908800464711,
+      "grad_norm": 0.36130279302597046,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 2064
+    },
+    {
+      "epoch": 0.2399070577984316,
+      "grad_norm": 0.3953266441822052,
+      "learning_rate": 0.0001,
+      "loss": 1.7695,
+      "step": 2065
+    },
+    {
+      "epoch": 0.2400232355503921,
+      "grad_norm": 0.3774740397930145,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 2066
+    },
+    {
+      "epoch": 0.2401394133023526,
+      "grad_norm": 0.380696564912796,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 2067
+    },
+    {
+      "epoch": 0.2402555910543131,
+      "grad_norm": 0.3862208127975464,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2403717688062736,
+      "grad_norm": 0.34933963418006897,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 2069
+    },
+    {
+      "epoch": 0.2404879465582341,
+      "grad_norm": 0.37984904646873474,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 2070
+    },
+    {
+      "epoch": 0.2406041243101946,
+      "grad_norm": 0.3804422616958618,
+      "learning_rate": 0.0001,
+      "loss": 1.7629,
+      "step": 2071
+    },
+    {
+      "epoch": 0.24072030206215508,
+      "grad_norm": 0.3919108808040619,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 2072
+    },
+    {
+      "epoch": 0.2408364798141156,
+      "grad_norm": 0.4050542712211609,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 2073
+    },
+    {
+      "epoch": 0.2409526575660761,
+      "grad_norm": 0.378836989402771,
+      "learning_rate": 0.0001,
+      "loss": 1.7873,
+      "step": 2074
+    },
+    {
+      "epoch": 0.2410688353180366,
+      "grad_norm": 0.39815083146095276,
+      "learning_rate": 0.0001,
+      "loss": 1.8006,
+      "step": 2075
+    },
+    {
+      "epoch": 0.2411850130699971,
+      "grad_norm": 0.39791086316108704,
+      "learning_rate": 0.0001,
+      "loss": 1.8538,
+      "step": 2076
+    },
+    {
+      "epoch": 0.24130119082195758,
+      "grad_norm": 0.3830665349960327,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 2077
+    },
+    {
+      "epoch": 0.2414173685739181,
+      "grad_norm": 0.38453802466392517,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 2078
+    },
+    {
+      "epoch": 0.2415335463258786,
+      "grad_norm": 0.40375974774360657,
+      "learning_rate": 0.0001,
+      "loss": 1.8335,
+      "step": 2079
+    },
+    {
+      "epoch": 0.2416497240778391,
+      "grad_norm": 0.38834068179130554,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2417659018297996,
+      "grad_norm": 0.38506248593330383,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 2081
+    },
+    {
+      "epoch": 0.24188207958176008,
+      "grad_norm": 0.3646622896194458,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 2082
+    },
+    {
+      "epoch": 0.2419982573337206,
+      "grad_norm": 0.37850677967071533,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 2083
+    },
+    {
+      "epoch": 0.2421144350856811,
+      "grad_norm": 0.386091411113739,
+      "learning_rate": 0.0001,
+      "loss": 1.8543,
+      "step": 2084
+    },
+    {
+      "epoch": 0.24223061283764158,
+      "grad_norm": 0.38737788796424866,
+      "learning_rate": 0.0001,
+      "loss": 1.7008,
+      "step": 2085
+    },
+    {
+      "epoch": 0.2423467905896021,
+      "grad_norm": 0.37103384733200073,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 2086
+    },
+    {
+      "epoch": 0.24246296834156258,
+      "grad_norm": 0.34921854734420776,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 2087
+    },
+    {
+      "epoch": 0.2425791460935231,
+      "grad_norm": 0.3560514748096466,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 2088
+    },
+    {
+      "epoch": 0.2426953238454836,
+      "grad_norm": 0.38028186559677124,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 2089
+    },
+    {
+      "epoch": 0.24281150159744408,
+      "grad_norm": 0.40691468119621277,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2429276793494046,
+      "grad_norm": 0.3687189221382141,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 2091
+    },
+    {
+      "epoch": 0.24304385710136509,
+      "grad_norm": 0.35396918654441833,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 2092
+    },
+    {
+      "epoch": 0.2431600348533256,
+      "grad_norm": 0.39168643951416016,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 2093
+    },
+    {
+      "epoch": 0.2432762126052861,
+      "grad_norm": 0.4195053279399872,
+      "learning_rate": 0.0001,
+      "loss": 1.9219,
+      "step": 2094
+    },
+    {
+      "epoch": 0.24339239035724658,
+      "grad_norm": 0.3593606948852539,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 2095
+    },
+    {
+      "epoch": 0.2435085681092071,
+      "grad_norm": 0.35588037967681885,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 2096
+    },
+    {
+      "epoch": 0.24362474586116759,
+      "grad_norm": 0.3821358382701874,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 2097
+    },
+    {
+      "epoch": 0.24374092361312807,
+      "grad_norm": 0.36826246976852417,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 2098
+    },
+    {
+      "epoch": 0.2438571013650886,
+      "grad_norm": 0.3902747929096222,
+      "learning_rate": 0.0001,
+      "loss": 1.8032,
+      "step": 2099
+    },
+    {
+      "epoch": 0.24397327911704908,
+      "grad_norm": 0.3716658651828766,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 2100
+    },
+    {
+      "epoch": 0.2440894568690096,
+      "grad_norm": 0.3505365550518036,
+      "learning_rate": 0.0001,
+      "loss": 1.6708,
+      "step": 2101
+    },
+    {
+      "epoch": 0.24420563462097009,
+      "grad_norm": 0.36862871050834656,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 2102
+    },
+    {
+      "epoch": 0.24432181237293057,
+      "grad_norm": 0.35868167877197266,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2444379901248911,
+      "grad_norm": 0.3973603844642639,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 2104
+    },
+    {
+      "epoch": 0.24455416787685158,
+      "grad_norm": 0.37595170736312866,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 2105
+    },
+    {
+      "epoch": 0.24467034562881207,
+      "grad_norm": 0.3656224012374878,
+      "learning_rate": 0.0001,
+      "loss": 1.6813,
+      "step": 2106
+    },
+    {
+      "epoch": 0.24478652338077259,
+      "grad_norm": 0.39221543073654175,
+      "learning_rate": 0.0001,
+      "loss": 1.7968,
+      "step": 2107
+    },
+    {
+      "epoch": 0.24490270113273307,
+      "grad_norm": 0.36675718426704407,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 2108
+    },
+    {
+      "epoch": 0.2450188788846936,
+      "grad_norm": 0.3545796871185303,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 2109
+    },
+    {
+      "epoch": 0.24513505663665408,
+      "grad_norm": 0.40240517258644104,
+      "learning_rate": 0.0001,
+      "loss": 1.9461,
+      "step": 2110
+    },
+    {
+      "epoch": 0.24525123438861457,
+      "grad_norm": 0.38813287019729614,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 2111
+    },
+    {
+      "epoch": 0.2453674121405751,
+      "grad_norm": 0.37340447306632996,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 2112
+    },
+    {
+      "epoch": 0.24548358989253558,
+      "grad_norm": 0.37866276502609253,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 2113
+    },
+    {
+      "epoch": 0.2455997676444961,
+      "grad_norm": 0.3932071328163147,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 2114
+    },
+    {
+      "epoch": 0.24571594539645658,
+      "grad_norm": 0.37961530685424805,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 2115
+    },
+    {
+      "epoch": 0.24583212314841707,
+      "grad_norm": 0.38449040055274963,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 2116
+    },
+    {
+      "epoch": 0.2459483009003776,
+      "grad_norm": 0.39510199427604675,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 2117
+    },
+    {
+      "epoch": 0.24606447865233808,
+      "grad_norm": 0.40252119302749634,
+      "learning_rate": 0.0001,
+      "loss": 1.7532,
+      "step": 2118
+    },
+    {
+      "epoch": 0.24618065640429856,
+      "grad_norm": 0.3880206346511841,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 2119
+    },
+    {
+      "epoch": 0.24629683415625908,
+      "grad_norm": 0.3541505038738251,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 2120
+    },
+    {
+      "epoch": 0.24641301190821957,
+      "grad_norm": 0.3985120952129364,
+      "learning_rate": 0.0001,
+      "loss": 1.7176,
+      "step": 2121
+    },
+    {
+      "epoch": 0.2465291896601801,
+      "grad_norm": 0.39708301424980164,
+      "learning_rate": 0.0001,
+      "loss": 1.7669,
+      "step": 2122
+    },
+    {
+      "epoch": 0.24664536741214058,
+      "grad_norm": 0.3605371415615082,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 2123
+    },
+    {
+      "epoch": 0.24676154516410106,
+      "grad_norm": 0.38313308358192444,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 2124
+    },
+    {
+      "epoch": 0.24687772291606158,
+      "grad_norm": 0.4088183641433716,
+      "learning_rate": 0.0001,
+      "loss": 1.7498,
+      "step": 2125
+    },
+    {
+      "epoch": 0.24699390066802207,
+      "grad_norm": 0.3966735899448395,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 2126
+    },
+    {
+      "epoch": 0.24711007841998256,
+      "grad_norm": 0.3689868748188019,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 2127
+    },
+    {
+      "epoch": 0.24722625617194308,
+      "grad_norm": 0.41865813732147217,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 2128
+    },
+    {
+      "epoch": 0.24734243392390357,
+      "grad_norm": 0.39082497358322144,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 2129
+    },
+    {
+      "epoch": 0.24745861167586408,
+      "grad_norm": 0.40093672275543213,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 2130
+    },
+    {
+      "epoch": 0.24757478942782457,
+      "grad_norm": 0.37469765543937683,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 2131
+    },
+    {
+      "epoch": 0.24769096717978506,
+      "grad_norm": 0.3949383795261383,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 2132
+    },
+    {
+      "epoch": 0.24780714493174558,
+      "grad_norm": 0.3878518342971802,
+      "learning_rate": 0.0001,
+      "loss": 1.9247,
+      "step": 2133
+    },
+    {
+      "epoch": 0.24792332268370607,
+      "grad_norm": 0.38046330213546753,
+      "learning_rate": 0.0001,
+      "loss": 1.7708,
+      "step": 2134
+    },
+    {
+      "epoch": 0.24803950043566658,
+      "grad_norm": 0.34667521715164185,
+      "learning_rate": 0.0001,
+      "loss": 1.7178,
+      "step": 2135
+    },
+    {
+      "epoch": 0.24815567818762707,
+      "grad_norm": 0.376510888338089,
+      "learning_rate": 0.0001,
+      "loss": 1.8054,
+      "step": 2136
+    },
+    {
+      "epoch": 0.24827185593958756,
+      "grad_norm": 0.371574729681015,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 2137
+    },
+    {
+      "epoch": 0.24838803369154808,
+      "grad_norm": 0.4060184955596924,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 2138
+    },
+    {
+      "epoch": 0.24850421144350857,
+      "grad_norm": 0.4174020290374756,
+      "learning_rate": 0.0001,
+      "loss": 1.8037,
+      "step": 2139
+    },
+    {
+      "epoch": 0.24862038919546905,
+      "grad_norm": 0.3601975440979004,
+      "learning_rate": 0.0001,
+      "loss": 1.3983,
+      "step": 2140
+    },
+    {
+      "epoch": 0.24873656694742957,
+      "grad_norm": 0.3981928825378418,
+      "learning_rate": 0.0001,
+      "loss": 1.7354,
+      "step": 2141
+    },
+    {
+      "epoch": 0.24885274469939006,
+      "grad_norm": 0.3899874687194824,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 2142
+    },
+    {
+      "epoch": 0.24896892245135058,
+      "grad_norm": 0.364611953496933,
+      "learning_rate": 0.0001,
+      "loss": 1.71,
+      "step": 2143
+    },
+    {
+      "epoch": 0.24908510020331107,
+      "grad_norm": 0.3837345540523529,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 2144
+    },
+    {
+      "epoch": 0.24920127795527156,
+      "grad_norm": 0.3657201826572418,
+      "learning_rate": 0.0001,
+      "loss": 1.7091,
+      "step": 2145
+    },
+    {
+      "epoch": 0.24931745570723207,
+      "grad_norm": 0.3665958642959595,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 2146
+    },
+    {
+      "epoch": 0.24943363345919256,
+      "grad_norm": 0.35987091064453125,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 2147
+    },
+    {
+      "epoch": 0.24954981121115308,
+      "grad_norm": 0.3838944733142853,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 2148
+    },
+    {
+      "epoch": 0.24966598896311357,
+      "grad_norm": 0.3626435101032257,
+      "learning_rate": 0.0001,
+      "loss": 1.5494,
+      "step": 2149
+    },
+    {
+      "epoch": 0.24978216671507406,
+      "grad_norm": 0.37998467683792114,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 2150
+    },
+    {
+      "epoch": 0.24989834446703457,
+      "grad_norm": 0.37865573167800903,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 2151
+    },
+    {
+      "epoch": 0.25001452221899506,
+      "grad_norm": 0.38255107402801514,
+      "learning_rate": 0.0001,
+      "loss": 1.7066,
+      "step": 2152
+    },
+    {
+      "epoch": 0.25013069997095555,
+      "grad_norm": 0.375410258769989,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 2153
+    },
+    {
+      "epoch": 0.25024687772291604,
+      "grad_norm": 0.3752192556858063,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 2154
+    },
+    {
+      "epoch": 0.2503630554748766,
+      "grad_norm": 0.3927246034145355,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 2155
+    },
+    {
+      "epoch": 0.2504792332268371,
+      "grad_norm": 0.38156911730766296,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 2156
+    },
+    {
+      "epoch": 0.25059541097879756,
+      "grad_norm": 0.3630543351173401,
+      "learning_rate": 0.0001,
+      "loss": 1.6824,
+      "step": 2157
+    },
+    {
+      "epoch": 0.25071158873075805,
+      "grad_norm": 0.38114479184150696,
+      "learning_rate": 0.0001,
+      "loss": 1.7344,
+      "step": 2158
+    },
+    {
+      "epoch": 0.25082776648271854,
+      "grad_norm": 0.3578778803348541,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 2159
+    },
+    {
+      "epoch": 0.2509439442346791,
+      "grad_norm": 0.3649848997592926,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 2160
+    },
+    {
+      "epoch": 0.2510601219866396,
+      "grad_norm": 0.39916592836380005,
+      "learning_rate": 0.0001,
+      "loss": 1.7755,
+      "step": 2161
+    },
+    {
+      "epoch": 0.25117629973860006,
+      "grad_norm": 0.34567490220069885,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 2162
+    },
+    {
+      "epoch": 0.25129247749056055,
+      "grad_norm": 0.3676474690437317,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 2163
+    },
+    {
+      "epoch": 0.25140865524252104,
+      "grad_norm": 0.3737073838710785,
+      "learning_rate": 0.0001,
+      "loss": 1.6569,
+      "step": 2164
+    },
+    {
+      "epoch": 0.2515248329944816,
+      "grad_norm": 0.3624831438064575,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 2165
+    },
+    {
+      "epoch": 0.2516410107464421,
+      "grad_norm": 0.4120437800884247,
+      "learning_rate": 0.0001,
+      "loss": 1.9489,
+      "step": 2166
+    },
+    {
+      "epoch": 0.25175718849840256,
+      "grad_norm": 0.3748963177204132,
+      "learning_rate": 0.0001,
+      "loss": 1.613,
+      "step": 2167
+    },
+    {
+      "epoch": 0.25187336625036305,
+      "grad_norm": 0.38849174976348877,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 2168
+    },
+    {
+      "epoch": 0.25198954400232354,
+      "grad_norm": 0.3984411358833313,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 2169
+    },
+    {
+      "epoch": 0.25210572175428403,
+      "grad_norm": 0.40575408935546875,
+      "learning_rate": 0.0001,
+      "loss": 1.8617,
+      "step": 2170
+    },
+    {
+      "epoch": 0.2522218995062446,
+      "grad_norm": 0.3502597212791443,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 2171
+    },
+    {
+      "epoch": 0.25233807725820506,
+      "grad_norm": 0.36587563157081604,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 2172
+    },
+    {
+      "epoch": 0.25245425501016555,
+      "grad_norm": 0.395804226398468,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 2173
+    },
+    {
+      "epoch": 0.25257043276212604,
+      "grad_norm": 0.3610069751739502,
+      "learning_rate": 0.0001,
+      "loss": 1.6106,
+      "step": 2174
+    },
+    {
+      "epoch": 0.25268661051408653,
+      "grad_norm": 0.38481786847114563,
+      "learning_rate": 0.0001,
+      "loss": 1.7379,
+      "step": 2175
+    },
+    {
+      "epoch": 0.2528027882660471,
+      "grad_norm": 0.4121454656124115,
+      "learning_rate": 0.0001,
+      "loss": 1.7449,
+      "step": 2176
+    },
+    {
+      "epoch": 0.25291896601800756,
+      "grad_norm": 0.40824562311172485,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 2177
+    },
+    {
+      "epoch": 0.25303514376996805,
+      "grad_norm": 0.38010165095329285,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 2178
+    },
+    {
+      "epoch": 0.25315132152192854,
+      "grad_norm": 0.3835202753543854,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 2179
+    },
+    {
+      "epoch": 0.25326749927388903,
+      "grad_norm": 0.4025690257549286,
+      "learning_rate": 0.0001,
+      "loss": 1.7204,
+      "step": 2180
+    },
+    {
+      "epoch": 0.2533836770258496,
+      "grad_norm": 0.3862835466861725,
+      "learning_rate": 0.0001,
+      "loss": 1.8057,
+      "step": 2181
+    },
+    {
+      "epoch": 0.25349985477781006,
+      "grad_norm": 0.3741121292114258,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 2182
+    },
+    {
+      "epoch": 0.25361603252977055,
+      "grad_norm": 0.38978350162506104,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 2183
+    },
+    {
+      "epoch": 0.25373221028173104,
+      "grad_norm": 0.3566723167896271,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 2184
+    },
+    {
+      "epoch": 0.25384838803369153,
+      "grad_norm": 0.4036724269390106,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 2185
+    },
+    {
+      "epoch": 0.2539645657856521,
+      "grad_norm": 0.3604584336280823,
+      "learning_rate": 0.0001,
+      "loss": 1.6988,
+      "step": 2186
+    },
+    {
+      "epoch": 0.25408074353761256,
+      "grad_norm": 0.3696625232696533,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 2187
+    },
+    {
+      "epoch": 0.25419692128957305,
+      "grad_norm": 0.38797250390052795,
+      "learning_rate": 0.0001,
+      "loss": 1.6658,
+      "step": 2188
+    },
+    {
+      "epoch": 0.25431309904153354,
+      "grad_norm": 0.3951641321182251,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 2189
+    },
+    {
+      "epoch": 0.25442927679349403,
+      "grad_norm": 0.37504735589027405,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 2190
+    },
+    {
+      "epoch": 0.2545454545454545,
+      "grad_norm": 0.38757580518722534,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 2191
+    },
+    {
+      "epoch": 0.25466163229741506,
+      "grad_norm": 0.41315603256225586,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 2192
+    },
+    {
+      "epoch": 0.25477781004937555,
+      "grad_norm": 0.36828580498695374,
+      "learning_rate": 0.0001,
+      "loss": 1.7711,
+      "step": 2193
+    },
+    {
+      "epoch": 0.25489398780133604,
+      "grad_norm": 0.39044126868247986,
+      "learning_rate": 0.0001,
+      "loss": 1.7516,
+      "step": 2194
+    },
+    {
+      "epoch": 0.25501016555329653,
+      "grad_norm": 0.3624008893966675,
+      "learning_rate": 0.0001,
+      "loss": 1.6642,
+      "step": 2195
+    },
+    {
+      "epoch": 0.255126343305257,
+      "grad_norm": 0.35610583424568176,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 2196
+    },
+    {
+      "epoch": 0.25524252105721756,
+      "grad_norm": 0.3624838888645172,
+      "learning_rate": 0.0001,
+      "loss": 1.6916,
+      "step": 2197
+    },
+    {
+      "epoch": 0.25535869880917805,
+      "grad_norm": 0.3851155638694763,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 2198
+    },
+    {
+      "epoch": 0.25547487656113854,
+      "grad_norm": 0.3829360604286194,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 2199
+    },
+    {
+      "epoch": 0.25559105431309903,
+      "grad_norm": 0.3810238242149353,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 2200
+    },
+    {
+      "epoch": 0.2557072320650595,
+      "grad_norm": 0.384115993976593,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 2201
+    },
+    {
+      "epoch": 0.25582340981702006,
+      "grad_norm": 0.3733699321746826,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 2202
+    },
+    {
+      "epoch": 0.25593958756898055,
+      "grad_norm": 0.37444037199020386,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 2203
+    },
+    {
+      "epoch": 0.25605576532094104,
+      "grad_norm": 0.3865280747413635,
+      "learning_rate": 0.0001,
+      "loss": 1.7525,
+      "step": 2204
+    },
+    {
+      "epoch": 0.25617194307290153,
+      "grad_norm": 0.3753882944583893,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 2205
+    },
+    {
+      "epoch": 0.256288120824862,
+      "grad_norm": 0.4033168852329254,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 2206
+    },
+    {
+      "epoch": 0.25640429857682256,
+      "grad_norm": 0.4054439961910248,
+      "learning_rate": 0.0001,
+      "loss": 1.76,
+      "step": 2207
+    },
+    {
+      "epoch": 0.25652047632878305,
+      "grad_norm": 0.3870498239994049,
+      "learning_rate": 0.0001,
+      "loss": 1.8224,
+      "step": 2208
+    },
+    {
+      "epoch": 0.25663665408074354,
+      "grad_norm": 0.3696228861808777,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 2209
+    },
+    {
+      "epoch": 0.25675283183270403,
+      "grad_norm": 0.36032870411872864,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 2210
+    },
+    {
+      "epoch": 0.2568690095846645,
+      "grad_norm": 0.3826732635498047,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 2211
+    },
+    {
+      "epoch": 0.256985187336625,
+      "grad_norm": 0.3756740391254425,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 2212
+    },
+    {
+      "epoch": 0.25710136508858555,
+      "grad_norm": 0.39152175188064575,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 2213
+    },
+    {
+      "epoch": 0.25721754284054604,
+      "grad_norm": 0.3959135413169861,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 2214
+    },
+    {
+      "epoch": 0.25733372059250653,
+      "grad_norm": 0.3905220627784729,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 2215
+    },
+    {
+      "epoch": 0.257449898344467,
+      "grad_norm": 0.3713771402835846,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 2216
+    },
+    {
+      "epoch": 0.2575660760964275,
+      "grad_norm": 0.39879244565963745,
+      "learning_rate": 0.0001,
+      "loss": 1.7314,
+      "step": 2217
+    },
+    {
+      "epoch": 0.25768225384838805,
+      "grad_norm": 0.3479825556278229,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 2218
+    },
+    {
+      "epoch": 0.25779843160034854,
+      "grad_norm": 0.38801833987236023,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 2219
+    },
+    {
+      "epoch": 0.25791460935230903,
+      "grad_norm": 0.37792959809303284,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 2220
+    },
+    {
+      "epoch": 0.2580307871042695,
+      "grad_norm": 0.3557312786579132,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 2221
+    },
+    {
+      "epoch": 0.25814696485623,
+      "grad_norm": 0.37623703479766846,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 2222
+    },
+    {
+      "epoch": 0.25826314260819055,
+      "grad_norm": 0.35475483536720276,
+      "learning_rate": 0.0001,
+      "loss": 1.5467,
+      "step": 2223
+    },
+    {
+      "epoch": 0.25837932036015104,
+      "grad_norm": 0.379874587059021,
+      "learning_rate": 0.0001,
+      "loss": 1.6733,
+      "step": 2224
+    },
+    {
+      "epoch": 0.25849549811211153,
+      "grad_norm": 0.3651340901851654,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 2225
+    },
+    {
+      "epoch": 0.258611675864072,
+      "grad_norm": 0.37844133377075195,
+      "learning_rate": 0.0001,
+      "loss": 1.7023,
+      "step": 2226
+    },
+    {
+      "epoch": 0.2587278536160325,
+      "grad_norm": 0.36854520440101624,
+      "learning_rate": 0.0001,
+      "loss": 1.799,
+      "step": 2227
+    },
+    {
+      "epoch": 0.25884403136799305,
+      "grad_norm": 0.4082055389881134,
+      "learning_rate": 0.0001,
+      "loss": 1.8261,
+      "step": 2228
+    },
+    {
+      "epoch": 0.25896020911995354,
+      "grad_norm": 0.3727450668811798,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 2229
+    },
+    {
+      "epoch": 0.25907638687191403,
+      "grad_norm": 0.38609281182289124,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 2230
+    },
+    {
+      "epoch": 0.2591925646238745,
+      "grad_norm": 0.4126780331134796,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 2231
+    },
+    {
+      "epoch": 0.259308742375835,
+      "grad_norm": 0.3675900399684906,
+      "learning_rate": 0.0001,
+      "loss": 1.7965,
+      "step": 2232
+    },
+    {
+      "epoch": 0.2594249201277955,
+      "grad_norm": 0.3680029511451721,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 2233
+    },
+    {
+      "epoch": 0.25954109787975604,
+      "grad_norm": 0.3985327482223511,
+      "learning_rate": 0.0001,
+      "loss": 1.7455,
+      "step": 2234
+    },
+    {
+      "epoch": 0.25965727563171653,
+      "grad_norm": 0.4041634798049927,
+      "learning_rate": 0.0001,
+      "loss": 1.9009,
+      "step": 2235
+    },
+    {
+      "epoch": 0.259773453383677,
+      "grad_norm": 0.3741016685962677,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 2236
+    },
+    {
+      "epoch": 0.2598896311356375,
+      "grad_norm": 0.38083234429359436,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 2237
+    },
+    {
+      "epoch": 0.260005808887598,
+      "grad_norm": 0.40651369094848633,
+      "learning_rate": 0.0001,
+      "loss": 1.9219,
+      "step": 2238
+    },
+    {
+      "epoch": 0.26012198663955854,
+      "grad_norm": 0.3514697849750519,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 2239
+    },
+    {
+      "epoch": 0.26023816439151903,
+      "grad_norm": 0.402826189994812,
+      "learning_rate": 0.0001,
+      "loss": 1.769,
+      "step": 2240
+    },
+    {
+      "epoch": 0.2603543421434795,
+      "grad_norm": 0.3742333948612213,
+      "learning_rate": 0.0001,
+      "loss": 1.784,
+      "step": 2241
+    },
+    {
+      "epoch": 0.26047051989544,
+      "grad_norm": 0.3999182879924774,
+      "learning_rate": 0.0001,
+      "loss": 1.9366,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2605866976474005,
+      "grad_norm": 0.3678368031978607,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 2243
+    },
+    {
+      "epoch": 0.26070287539936104,
+      "grad_norm": 0.36702072620391846,
+      "learning_rate": 0.0001,
+      "loss": 1.4785,
+      "step": 2244
+    },
+    {
+      "epoch": 0.26081905315132153,
+      "grad_norm": 0.3562362790107727,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 2245
+    },
+    {
+      "epoch": 0.260935230903282,
+      "grad_norm": 0.40297916531562805,
+      "learning_rate": 0.0001,
+      "loss": 1.7167,
+      "step": 2246
+    },
+    {
+      "epoch": 0.2610514086552425,
+      "grad_norm": 0.37424853444099426,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 2247
+    },
+    {
+      "epoch": 0.261167586407203,
+      "grad_norm": 0.36831429600715637,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 2248
+    },
+    {
+      "epoch": 0.26128376415916355,
+      "grad_norm": 0.38578173518180847,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 2249
+    },
+    {
+      "epoch": 0.26139994191112403,
+      "grad_norm": 0.40575075149536133,
+      "learning_rate": 0.0001,
+      "loss": 1.732,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2615161196630845,
+      "grad_norm": 0.3883078396320343,
+      "learning_rate": 0.0001,
+      "loss": 1.6865,
+      "step": 2251
+    },
+    {
+      "epoch": 0.261632297415045,
+      "grad_norm": 0.3869827687740326,
+      "learning_rate": 0.0001,
+      "loss": 1.7362,
+      "step": 2252
+    },
+    {
+      "epoch": 0.2617484751670055,
+      "grad_norm": 0.38627856969833374,
+      "learning_rate": 0.0001,
+      "loss": 1.751,
+      "step": 2253
+    },
+    {
+      "epoch": 0.26186465291896605,
+      "grad_norm": 0.3833690285682678,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 2254
+    },
+    {
+      "epoch": 0.26198083067092653,
+      "grad_norm": 0.39400362968444824,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 2255
+    },
+    {
+      "epoch": 0.262097008422887,
+      "grad_norm": 0.3659832775592804,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2622131861748475,
+      "grad_norm": 0.3808686137199402,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 2257
+    },
+    {
+      "epoch": 0.262329363926808,
+      "grad_norm": 0.36852577328681946,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 2258
+    },
+    {
+      "epoch": 0.2624455416787685,
+      "grad_norm": 0.37290260195732117,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 2259
+    },
+    {
+      "epoch": 0.26256171943072903,
+      "grad_norm": 0.4122512936592102,
+      "learning_rate": 0.0001,
+      "loss": 1.7609,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2626778971826895,
+      "grad_norm": 0.3920998275279999,
+      "learning_rate": 0.0001,
+      "loss": 1.7818,
+      "step": 2261
+    },
+    {
+      "epoch": 0.26279407493465,
+      "grad_norm": 0.40477001667022705,
+      "learning_rate": 0.0001,
+      "loss": 1.8105,
+      "step": 2262
+    },
+    {
+      "epoch": 0.2629102526866105,
+      "grad_norm": 0.3666556179523468,
+      "learning_rate": 0.0001,
+      "loss": 1.5604,
+      "step": 2263
+    },
+    {
+      "epoch": 0.263026430438571,
+      "grad_norm": 0.3544174134731293,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 2264
+    },
+    {
+      "epoch": 0.26314260819053154,
+      "grad_norm": 0.3537236154079437,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 2265
+    },
+    {
+      "epoch": 0.263258785942492,
+      "grad_norm": 0.3871021866798401,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 2266
+    },
+    {
+      "epoch": 0.2633749636944525,
+      "grad_norm": 0.3693544566631317,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 2267
+    },
+    {
+      "epoch": 0.263491141446413,
+      "grad_norm": 0.3744482696056366,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 2268
+    },
+    {
+      "epoch": 0.2636073191983735,
+      "grad_norm": 0.38229045271873474,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 2269
+    },
+    {
+      "epoch": 0.26372349695033404,
+      "grad_norm": 0.38412609696388245,
+      "learning_rate": 0.0001,
+      "loss": 1.806,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2638396747022945,
+      "grad_norm": 0.3687461018562317,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 2271
+    },
+    {
+      "epoch": 0.263955852454255,
+      "grad_norm": 0.40106597542762756,
+      "learning_rate": 0.0001,
+      "loss": 1.7895,
+      "step": 2272
+    },
+    {
+      "epoch": 0.2640720302062155,
+      "grad_norm": 0.3707484304904938,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 2273
+    },
+    {
+      "epoch": 0.264188207958176,
+      "grad_norm": 0.3728141188621521,
+      "learning_rate": 0.0001,
+      "loss": 1.7414,
+      "step": 2274
+    },
+    {
+      "epoch": 0.26430438571013654,
+      "grad_norm": 0.3807673752307892,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 2275
+    },
+    {
+      "epoch": 0.264420563462097,
+      "grad_norm": 0.39177513122558594,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 2276
+    },
+    {
+      "epoch": 0.2645367412140575,
+      "grad_norm": 0.3902243971824646,
+      "learning_rate": 0.0001,
+      "loss": 1.7135,
+      "step": 2277
+    },
+    {
+      "epoch": 0.264652918966018,
+      "grad_norm": 0.36920416355133057,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 2278
+    },
+    {
+      "epoch": 0.2647690967179785,
+      "grad_norm": 0.37540876865386963,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 2279
+    },
+    {
+      "epoch": 0.264885274469939,
+      "grad_norm": 0.3670298159122467,
+      "learning_rate": 0.0001,
+      "loss": 1.8018,
+      "step": 2280
+    },
+    {
+      "epoch": 0.2650014522218995,
+      "grad_norm": 0.344098836183548,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 2281
+    },
+    {
+      "epoch": 0.26511762997386,
+      "grad_norm": 0.36059024930000305,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 2282
+    },
+    {
+      "epoch": 0.2652338077258205,
+      "grad_norm": 0.37826424837112427,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 2283
+    },
+    {
+      "epoch": 0.265349985477781,
+      "grad_norm": 0.40158766508102417,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2654661632297415,
+      "grad_norm": 0.37319740653038025,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 2285
+    },
+    {
+      "epoch": 0.265582340981702,
+      "grad_norm": 0.3656269311904907,
+      "learning_rate": 0.0001,
+      "loss": 1.6144,
+      "step": 2286
+    },
+    {
+      "epoch": 0.2656985187336625,
+      "grad_norm": 0.3666156232357025,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 2287
+    },
+    {
+      "epoch": 0.265814696485623,
+      "grad_norm": 0.3868063688278198,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 2288
+    },
+    {
+      "epoch": 0.2659308742375835,
+      "grad_norm": 0.36704185605049133,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 2289
+    },
+    {
+      "epoch": 0.266047051989544,
+      "grad_norm": 0.37085020542144775,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2661632297415045,
+      "grad_norm": 0.3715968430042267,
+      "learning_rate": 0.0001,
+      "loss": 1.8348,
+      "step": 2291
+    },
+    {
+      "epoch": 0.266279407493465,
+      "grad_norm": 0.3815777599811554,
+      "learning_rate": 0.0001,
+      "loss": 1.7571,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2663955852454255,
+      "grad_norm": 0.3628351092338562,
+      "learning_rate": 0.0001,
+      "loss": 1.8001,
+      "step": 2293
+    },
+    {
+      "epoch": 0.266511762997386,
+      "grad_norm": 0.39187735319137573,
+      "learning_rate": 0.0001,
+      "loss": 1.7246,
+      "step": 2294
+    },
+    {
+      "epoch": 0.2666279407493465,
+      "grad_norm": 0.4006575644016266,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 2295
+    },
+    {
+      "epoch": 0.266744118501307,
+      "grad_norm": 0.3559897243976593,
+      "learning_rate": 0.0001,
+      "loss": 1.6132,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2668602962532675,
+      "grad_norm": 0.3798125088214874,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 2297
+    },
+    {
+      "epoch": 0.266976474005228,
+      "grad_norm": 0.40822911262512207,
+      "learning_rate": 0.0001,
+      "loss": 1.7918,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2670926517571885,
+      "grad_norm": 0.36172357201576233,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 2299
+    },
+    {
+      "epoch": 0.267208829509149,
+      "grad_norm": 0.3883090913295746,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 2300
+    },
+    {
+      "epoch": 0.26732500726110947,
+      "grad_norm": 0.3930845856666565,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 2301
+    },
+    {
+      "epoch": 0.26744118501307,
+      "grad_norm": 0.34050822257995605,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 2302
+    },
+    {
+      "epoch": 0.2675573627650305,
+      "grad_norm": 0.37255436182022095,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 2303
+    },
+    {
+      "epoch": 0.267673540516991,
+      "grad_norm": 0.383784681558609,
+      "learning_rate": 0.0001,
+      "loss": 1.7766,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2677897182689515,
+      "grad_norm": 0.3867911100387573,
+      "learning_rate": 0.0001,
+      "loss": 1.855,
+      "step": 2305
+    },
+    {
+      "epoch": 0.26790589602091197,
+      "grad_norm": 0.40486153960227966,
+      "learning_rate": 0.0001,
+      "loss": 1.6078,
+      "step": 2306
+    },
+    {
+      "epoch": 0.2680220737728725,
+      "grad_norm": 0.42723652720451355,
+      "learning_rate": 0.0001,
+      "loss": 1.909,
+      "step": 2307
+    },
+    {
+      "epoch": 0.268138251524833,
+      "grad_norm": 0.3602335751056671,
+      "learning_rate": 0.0001,
+      "loss": 1.6768,
+      "step": 2308
+    },
+    {
+      "epoch": 0.2682544292767935,
+      "grad_norm": 0.3916855454444885,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 2309
+    },
+    {
+      "epoch": 0.268370607028754,
+      "grad_norm": 0.3967686891555786,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 2310
+    },
+    {
+      "epoch": 0.26848678478071447,
+      "grad_norm": 0.37065213918685913,
+      "learning_rate": 0.0001,
+      "loss": 1.7548,
+      "step": 2311
+    },
+    {
+      "epoch": 0.268602962532675,
+      "grad_norm": 0.3448849618434906,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 2312
+    },
+    {
+      "epoch": 0.2687191402846355,
+      "grad_norm": 0.36089155077934265,
+      "learning_rate": 0.0001,
+      "loss": 1.845,
+      "step": 2313
+    },
+    {
+      "epoch": 0.268835318036596,
+      "grad_norm": 0.3895665109157562,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2689514957885565,
+      "grad_norm": 0.3936954736709595,
+      "learning_rate": 0.0001,
+      "loss": 1.6873,
+      "step": 2315
+    },
+    {
+      "epoch": 0.26906767354051697,
+      "grad_norm": 0.38149169087409973,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 2316
+    },
+    {
+      "epoch": 0.2691838512924775,
+      "grad_norm": 0.34843528270721436,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 2317
+    },
+    {
+      "epoch": 0.269300029044438,
+      "grad_norm": 0.3937288522720337,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 2318
+    },
+    {
+      "epoch": 0.2694162067963985,
+      "grad_norm": 0.431801438331604,
+      "learning_rate": 0.0001,
+      "loss": 1.7015,
+      "step": 2319
+    },
+    {
+      "epoch": 0.269532384548359,
+      "grad_norm": 0.3963301181793213,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 2320
+    },
+    {
+      "epoch": 0.26964856230031947,
+      "grad_norm": 0.39253467321395874,
+      "learning_rate": 0.0001,
+      "loss": 1.8633,
+      "step": 2321
+    },
+    {
+      "epoch": 0.26976474005227996,
+      "grad_norm": 0.3785027861595154,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 2322
+    },
+    {
+      "epoch": 0.2698809178042405,
+      "grad_norm": 0.4014844596385956,
+      "learning_rate": 0.0001,
+      "loss": 1.8149,
+      "step": 2323
+    },
+    {
+      "epoch": 0.269997095556201,
+      "grad_norm": 0.3676471710205078,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2701132733081615,
+      "grad_norm": 0.3606375455856323,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 2325
+    },
+    {
+      "epoch": 0.270229451060122,
+      "grad_norm": 0.3681615889072418,
+      "learning_rate": 0.0001,
+      "loss": 1.747,
+      "step": 2326
+    },
+    {
+      "epoch": 0.27034562881208246,
+      "grad_norm": 0.3740110397338867,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 2327
+    },
+    {
+      "epoch": 0.270461806564043,
+      "grad_norm": 0.39461883902549744,
+      "learning_rate": 0.0001,
+      "loss": 1.7891,
+      "step": 2328
+    },
+    {
+      "epoch": 0.2705779843160035,
+      "grad_norm": 0.3810250759124756,
+      "learning_rate": 0.0001,
+      "loss": 1.7613,
+      "step": 2329
+    },
+    {
+      "epoch": 0.270694162067964,
+      "grad_norm": 0.39059239625930786,
+      "learning_rate": 0.0001,
+      "loss": 1.7721,
+      "step": 2330
+    },
+    {
+      "epoch": 0.2708103398199245,
+      "grad_norm": 0.40021392703056335,
+      "learning_rate": 0.0001,
+      "loss": 1.8333,
+      "step": 2331
+    },
+    {
+      "epoch": 0.27092651757188496,
+      "grad_norm": 0.42178237438201904,
+      "learning_rate": 0.0001,
+      "loss": 1.819,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2710426953238455,
+      "grad_norm": 0.3722633123397827,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 2333
+    },
+    {
+      "epoch": 0.271158873075806,
+      "grad_norm": 0.3866879940032959,
+      "learning_rate": 0.0001,
+      "loss": 1.8033,
+      "step": 2334
+    },
+    {
+      "epoch": 0.2712750508277665,
+      "grad_norm": 0.38638031482696533,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 2335
+    },
+    {
+      "epoch": 0.271391228579727,
+      "grad_norm": 0.37124332785606384,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 2336
+    },
+    {
+      "epoch": 0.27150740633168746,
+      "grad_norm": 0.39737075567245483,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 2337
+    },
+    {
+      "epoch": 0.271623584083648,
+      "grad_norm": 0.41190680861473083,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2717397618356085,
+      "grad_norm": 0.36484986543655396,
+      "learning_rate": 0.0001,
+      "loss": 1.7156,
+      "step": 2339
+    },
+    {
+      "epoch": 0.271855939587569,
+      "grad_norm": 0.38202008605003357,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 2340
+    },
+    {
+      "epoch": 0.2719721173395295,
+      "grad_norm": 0.3791213035583496,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 2341
+    },
+    {
+      "epoch": 0.27208829509148996,
+      "grad_norm": 0.3870936930179596,
+      "learning_rate": 0.0001,
+      "loss": 1.6692,
+      "step": 2342
+    },
+    {
+      "epoch": 0.2722044728434505,
+      "grad_norm": 0.3662189543247223,
+      "learning_rate": 0.0001,
+      "loss": 1.7003,
+      "step": 2343
+    },
+    {
+      "epoch": 0.272320650595411,
+      "grad_norm": 0.3746212422847748,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2724368283473715,
+      "grad_norm": 0.3724815845489502,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 2345
+    },
+    {
+      "epoch": 0.272553006099332,
+      "grad_norm": 0.390982985496521,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 2346
+    },
+    {
+      "epoch": 0.27266918385129246,
+      "grad_norm": 0.38862210512161255,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 2347
+    },
+    {
+      "epoch": 0.27278536160325295,
+      "grad_norm": 0.3926958441734314,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2729015393552135,
+      "grad_norm": 0.3797786831855774,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 2349
+    },
+    {
+      "epoch": 0.273017717107174,
+      "grad_norm": 0.40166667103767395,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2731338948591345,
+      "grad_norm": 0.37161746621131897,
+      "learning_rate": 0.0001,
+      "loss": 1.7761,
+      "step": 2351
+    },
+    {
+      "epoch": 0.27325007261109496,
+      "grad_norm": 0.3659614324569702,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 2352
+    },
+    {
+      "epoch": 0.27336625036305545,
+      "grad_norm": 0.41991570591926575,
+      "learning_rate": 0.0001,
+      "loss": 1.7652,
+      "step": 2353
+    },
+    {
+      "epoch": 0.273482428115016,
+      "grad_norm": 0.41460558772087097,
+      "learning_rate": 0.0001,
+      "loss": 1.7779,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2735986058669765,
+      "grad_norm": 0.3913847506046295,
+      "learning_rate": 0.0001,
+      "loss": 1.6966,
+      "step": 2355
+    },
+    {
+      "epoch": 0.273714783618937,
+      "grad_norm": 0.3612228035926819,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 2356
+    },
+    {
+      "epoch": 0.27383096137089746,
+      "grad_norm": 0.3820975422859192,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 2357
+    },
+    {
+      "epoch": 0.27394713912285795,
+      "grad_norm": 0.37913262844085693,
+      "learning_rate": 0.0001,
+      "loss": 1.7815,
+      "step": 2358
+    },
+    {
+      "epoch": 0.2740633168748185,
+      "grad_norm": 0.3759218454360962,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 2359
+    },
+    {
+      "epoch": 0.274179494626779,
+      "grad_norm": 0.3867158889770508,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2742956723787395,
+      "grad_norm": 0.36591243743896484,
+      "learning_rate": 0.0001,
+      "loss": 1.7097,
+      "step": 2361
+    },
+    {
+      "epoch": 0.27441185013069996,
+      "grad_norm": 0.37191227078437805,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 2362
+    },
+    {
+      "epoch": 0.27452802788266045,
+      "grad_norm": 0.40537264943122864,
+      "learning_rate": 0.0001,
+      "loss": 1.7831,
+      "step": 2363
+    },
+    {
+      "epoch": 0.274644205634621,
+      "grad_norm": 0.3994043171405792,
+      "learning_rate": 0.0001,
+      "loss": 1.7722,
+      "step": 2364
+    },
+    {
+      "epoch": 0.2747603833865815,
+      "grad_norm": 0.3657456040382385,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 2365
+    },
+    {
+      "epoch": 0.274876561138542,
+      "grad_norm": 0.3876878321170807,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 2366
+    },
+    {
+      "epoch": 0.27499273889050246,
+      "grad_norm": 0.3732263445854187,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 2367
+    },
+    {
+      "epoch": 0.27510891664246295,
+      "grad_norm": 0.3888086974620819,
+      "learning_rate": 0.0001,
+      "loss": 1.8368,
+      "step": 2368
+    },
+    {
+      "epoch": 0.27522509439442344,
+      "grad_norm": 0.3704921007156372,
+      "learning_rate": 0.0001,
+      "loss": 1.7033,
+      "step": 2369
+    },
+    {
+      "epoch": 0.275341272146384,
+      "grad_norm": 0.4058105945587158,
+      "learning_rate": 0.0001,
+      "loss": 1.907,
+      "step": 2370
+    },
+    {
+      "epoch": 0.2754574498983445,
+      "grad_norm": 0.35868632793426514,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 2371
+    },
+    {
+      "epoch": 0.27557362765030496,
+      "grad_norm": 0.3760453462600708,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 2372
+    },
+    {
+      "epoch": 0.27568980540226545,
+      "grad_norm": 0.43682530522346497,
+      "learning_rate": 0.0001,
+      "loss": 1.8795,
+      "step": 2373
+    },
+    {
+      "epoch": 0.27580598315422594,
+      "grad_norm": 0.4160868525505066,
+      "learning_rate": 0.0001,
+      "loss": 1.7954,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2759221609061865,
+      "grad_norm": 0.3560098707675934,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 2375
+    },
+    {
+      "epoch": 0.276038338658147,
+      "grad_norm": 0.37041234970092773,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 2376
+    },
+    {
+      "epoch": 0.27615451641010746,
+      "grad_norm": 0.3486596345901489,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 2377
+    },
+    {
+      "epoch": 0.27627069416206795,
+      "grad_norm": 0.368912935256958,
+      "learning_rate": 0.0001,
+      "loss": 1.7511,
+      "step": 2378
+    },
+    {
+      "epoch": 0.27638687191402844,
+      "grad_norm": 0.41473448276519775,
+      "learning_rate": 0.0001,
+      "loss": 1.8201,
+      "step": 2379
+    },
+    {
+      "epoch": 0.276503049665989,
+      "grad_norm": 0.39010536670684814,
+      "learning_rate": 0.0001,
+      "loss": 1.8835,
+      "step": 2380
+    },
+    {
+      "epoch": 0.2766192274179495,
+      "grad_norm": 0.44715237617492676,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 2381
+    },
+    {
+      "epoch": 0.27673540516990996,
+      "grad_norm": 0.38639402389526367,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 2382
+    },
+    {
+      "epoch": 0.27685158292187045,
+      "grad_norm": 0.3690156638622284,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 2383
+    },
+    {
+      "epoch": 0.27696776067383094,
+      "grad_norm": 0.380719393491745,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2770839384257915,
+      "grad_norm": 0.37003418803215027,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 2385
+    },
+    {
+      "epoch": 0.277200116177752,
+      "grad_norm": 0.3689243197441101,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 2386
+    },
+    {
+      "epoch": 0.27731629392971247,
+      "grad_norm": 0.3837697505950928,
+      "learning_rate": 0.0001,
+      "loss": 1.6301,
+      "step": 2387
+    },
+    {
+      "epoch": 0.27743247168167295,
+      "grad_norm": 0.39931994676589966,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 2388
+    },
+    {
+      "epoch": 0.27754864943363344,
+      "grad_norm": 0.3684816062450409,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 2389
+    },
+    {
+      "epoch": 0.27766482718559393,
+      "grad_norm": 0.37523365020751953,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2777810049375545,
+      "grad_norm": 0.38715073466300964,
+      "learning_rate": 0.0001,
+      "loss": 1.8479,
+      "step": 2391
+    },
+    {
+      "epoch": 0.27789718268951497,
+      "grad_norm": 0.3796006441116333,
+      "learning_rate": 0.0001,
+      "loss": 1.6852,
+      "step": 2392
+    },
+    {
+      "epoch": 0.27801336044147545,
+      "grad_norm": 0.3867599070072174,
+      "learning_rate": 0.0001,
+      "loss": 1.6786,
+      "step": 2393
+    },
+    {
+      "epoch": 0.27812953819343594,
+      "grad_norm": 0.36347195506095886,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 2394
+    },
+    {
+      "epoch": 0.27824571594539643,
+      "grad_norm": 0.40452706813812256,
+      "learning_rate": 0.0001,
+      "loss": 1.7461,
+      "step": 2395
+    },
+    {
+      "epoch": 0.278361893697357,
+      "grad_norm": 0.37691494822502136,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 2396
+    },
+    {
+      "epoch": 0.27847807144931747,
+      "grad_norm": 0.3693540394306183,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 2397
+    },
+    {
+      "epoch": 0.27859424920127795,
+      "grad_norm": 0.3859393894672394,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 2398
+    },
+    {
+      "epoch": 0.27871042695323844,
+      "grad_norm": 0.36615633964538574,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 2399
+    },
+    {
+      "epoch": 0.27882660470519893,
+      "grad_norm": 0.4025718867778778,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 2400
+    },
+    {
+      "epoch": 0.2789427824571595,
+      "grad_norm": 0.35570523142814636,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 2401
+    },
+    {
+      "epoch": 0.27905896020911997,
+      "grad_norm": 0.4003136456012726,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 2402
+    },
+    {
+      "epoch": 0.27917513796108046,
+      "grad_norm": 0.38738471269607544,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 2403
+    },
+    {
+      "epoch": 0.27929131571304094,
+      "grad_norm": 0.38564881682395935,
+      "learning_rate": 0.0001,
+      "loss": 1.8121,
+      "step": 2404
+    },
+    {
+      "epoch": 0.27940749346500143,
+      "grad_norm": 0.4316978454589844,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 2405
+    },
+    {
+      "epoch": 0.279523671216962,
+      "grad_norm": 0.3853331208229065,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 2406
+    },
+    {
+      "epoch": 0.27963984896892247,
+      "grad_norm": 0.3825344443321228,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 2407
+    },
+    {
+      "epoch": 0.27975602672088296,
+      "grad_norm": 0.3726043999195099,
+      "learning_rate": 0.0001,
+      "loss": 1.7153,
+      "step": 2408
+    },
+    {
+      "epoch": 0.27987220447284344,
+      "grad_norm": 0.36002033948898315,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 2409
+    },
+    {
+      "epoch": 0.27998838222480393,
+      "grad_norm": 0.388629674911499,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2801045599767644,
+      "grad_norm": 0.3558464050292969,
+      "learning_rate": 0.0001,
+      "loss": 1.3596,
+      "step": 2411
+    },
+    {
+      "epoch": 0.28022073772872497,
+      "grad_norm": 0.36634495854377747,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 2412
+    },
+    {
+      "epoch": 0.28033691548068546,
+      "grad_norm": 0.36792120337486267,
+      "learning_rate": 0.0001,
+      "loss": 1.775,
+      "step": 2413
+    },
+    {
+      "epoch": 0.28045309323264594,
+      "grad_norm": 0.35753536224365234,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 2414
+    },
+    {
+      "epoch": 0.28056927098460643,
+      "grad_norm": 0.3687507212162018,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2806854487365669,
+      "grad_norm": 0.35439378023147583,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 2416
+    },
+    {
+      "epoch": 0.28080162648852747,
+      "grad_norm": 0.3909814655780792,
+      "learning_rate": 0.0001,
+      "loss": 1.843,
+      "step": 2417
+    },
+    {
+      "epoch": 0.28091780424048796,
+      "grad_norm": 0.4006040096282959,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 2418
+    },
+    {
+      "epoch": 0.28103398199244845,
+      "grad_norm": 0.3820417821407318,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 2419
+    },
+    {
+      "epoch": 0.28115015974440893,
+      "grad_norm": 0.41113728284835815,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 2420
+    },
+    {
+      "epoch": 0.2812663374963694,
+      "grad_norm": 0.37190499901771545,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 2421
+    },
+    {
+      "epoch": 0.28138251524832997,
+      "grad_norm": 0.3800449073314667,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 2422
+    },
+    {
+      "epoch": 0.28149869300029046,
+      "grad_norm": 0.3811475932598114,
+      "learning_rate": 0.0001,
+      "loss": 1.7483,
+      "step": 2423
+    },
+    {
+      "epoch": 0.28161487075225095,
+      "grad_norm": 0.3777383863925934,
+      "learning_rate": 0.0001,
+      "loss": 1.6842,
+      "step": 2424
+    },
+    {
+      "epoch": 0.28173104850421143,
+      "grad_norm": 0.3724311590194702,
+      "learning_rate": 0.0001,
+      "loss": 1.7839,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2818472262561719,
+      "grad_norm": 0.38829484581947327,
+      "learning_rate": 0.0001,
+      "loss": 1.8328,
+      "step": 2426
+    },
+    {
+      "epoch": 0.28196340400813247,
+      "grad_norm": 0.4069208800792694,
+      "learning_rate": 0.0001,
+      "loss": 1.873,
+      "step": 2427
+    },
+    {
+      "epoch": 0.28207958176009296,
+      "grad_norm": 0.3797883093357086,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 2428
+    },
+    {
+      "epoch": 0.28219575951205345,
+      "grad_norm": 0.3935191333293915,
+      "learning_rate": 0.0001,
+      "loss": 1.8087,
+      "step": 2429
+    },
+    {
+      "epoch": 0.28231193726401393,
+      "grad_norm": 0.37056753039360046,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 2430
+    },
+    {
+      "epoch": 0.2824281150159744,
+      "grad_norm": 0.3883243799209595,
+      "learning_rate": 0.0001,
+      "loss": 1.7572,
+      "step": 2431
+    },
+    {
+      "epoch": 0.28254429276793497,
+      "grad_norm": 0.38896408677101135,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 2432
+    },
+    {
+      "epoch": 0.28266047051989546,
+      "grad_norm": 0.3767715096473694,
+      "learning_rate": 0.0001,
+      "loss": 1.7514,
+      "step": 2433
+    },
+    {
+      "epoch": 0.28277664827185595,
+      "grad_norm": 0.36977216601371765,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 2434
+    },
+    {
+      "epoch": 0.28289282602381643,
+      "grad_norm": 0.38601481914520264,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 2435
+    },
+    {
+      "epoch": 0.2830090037757769,
+      "grad_norm": 0.36568722128868103,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 2436
+    },
+    {
+      "epoch": 0.2831251815277374,
+      "grad_norm": 0.3678687512874603,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 2437
+    },
+    {
+      "epoch": 0.28324135927969796,
+      "grad_norm": 0.4139416813850403,
+      "learning_rate": 0.0001,
+      "loss": 1.7692,
+      "step": 2438
+    },
+    {
+      "epoch": 0.28335753703165845,
+      "grad_norm": 0.3862224519252777,
+      "learning_rate": 0.0001,
+      "loss": 1.7017,
+      "step": 2439
+    },
+    {
+      "epoch": 0.28347371478361894,
+      "grad_norm": 0.3898821771144867,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2835898925355794,
+      "grad_norm": 0.35055509209632874,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2837060702875399,
+      "grad_norm": 0.3779016137123108,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 2442
+    },
+    {
+      "epoch": 0.28382224803950046,
+      "grad_norm": 0.38517260551452637,
+      "learning_rate": 0.0001,
+      "loss": 1.6497,
+      "step": 2443
+    },
+    {
+      "epoch": 0.28393842579146095,
+      "grad_norm": 0.3903183043003082,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 2444
+    },
+    {
+      "epoch": 0.28405460354342144,
+      "grad_norm": 0.41626691818237305,
+      "learning_rate": 0.0001,
+      "loss": 1.8199,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2841707812953819,
+      "grad_norm": 0.37726128101348877,
+      "learning_rate": 0.0001,
+      "loss": 1.7504,
+      "step": 2446
+    },
+    {
+      "epoch": 0.2842869590473424,
+      "grad_norm": 0.3679426908493042,
+      "learning_rate": 0.0001,
+      "loss": 1.6864,
+      "step": 2447
+    },
+    {
+      "epoch": 0.28440313679930296,
+      "grad_norm": 0.3660835921764374,
+      "learning_rate": 0.0001,
+      "loss": 1.7442,
+      "step": 2448
+    },
+    {
+      "epoch": 0.28451931455126345,
+      "grad_norm": 0.39392563700675964,
+      "learning_rate": 0.0001,
+      "loss": 1.7503,
+      "step": 2449
+    },
+    {
+      "epoch": 0.28463549230322394,
+      "grad_norm": 0.38993775844573975,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2847516700551844,
+      "grad_norm": 0.3619860112667084,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2848678478071449,
+      "grad_norm": 0.3857851028442383,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 2452
+    },
+    {
+      "epoch": 0.28498402555910546,
+      "grad_norm": 0.3716614544391632,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 2453
+    },
+    {
+      "epoch": 0.28510020331106595,
+      "grad_norm": 0.4253447949886322,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 2454
+    },
+    {
+      "epoch": 0.28521638106302644,
+      "grad_norm": 0.39747342467308044,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2853325588149869,
+      "grad_norm": 0.3874737322330475,
+      "learning_rate": 0.0001,
+      "loss": 1.763,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2854487365669474,
+      "grad_norm": 0.37462908029556274,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 2457
+    },
+    {
+      "epoch": 0.2855649143189079,
+      "grad_norm": 0.4207324981689453,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 2458
+    },
+    {
+      "epoch": 0.28568109207086845,
+      "grad_norm": 0.3763648569583893,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 2459
+    },
+    {
+      "epoch": 0.28579726982282894,
+      "grad_norm": 0.40208232402801514,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2859134475747894,
+      "grad_norm": 0.3689945638179779,
+      "learning_rate": 0.0001,
+      "loss": 1.7245,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2860296253267499,
+      "grad_norm": 0.3570806086063385,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2861458030787104,
+      "grad_norm": 0.39865806698799133,
+      "learning_rate": 0.0001,
+      "loss": 1.888,
+      "step": 2463
+    },
+    {
+      "epoch": 0.28626198083067095,
+      "grad_norm": 0.3934881389141083,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 2464
+    },
+    {
+      "epoch": 0.28637815858263144,
+      "grad_norm": 0.38317766785621643,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2864943363345919,
+      "grad_norm": 0.3968998193740845,
+      "learning_rate": 0.0001,
+      "loss": 1.7444,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2866105140865524,
+      "grad_norm": 0.3885013461112976,
+      "learning_rate": 0.0001,
+      "loss": 1.7969,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2867266918385129,
+      "grad_norm": 0.3675188720226288,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 2468
+    },
+    {
+      "epoch": 0.28684286959047345,
+      "grad_norm": 0.39261195063591003,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 2469
+    },
+    {
+      "epoch": 0.28695904734243394,
+      "grad_norm": 0.3737514615058899,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 2470
+    },
+    {
+      "epoch": 0.2870752250943944,
+      "grad_norm": 0.3958921730518341,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2871914028463549,
+      "grad_norm": 0.37590277194976807,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2873075805983154,
+      "grad_norm": 0.39498046040534973,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 2473
+    },
+    {
+      "epoch": 0.28742375835027595,
+      "grad_norm": 0.35599952936172485,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 2474
+    },
+    {
+      "epoch": 0.28753993610223644,
+      "grad_norm": 0.3650219440460205,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2876561138541969,
+      "grad_norm": 0.3618330955505371,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2877722916061574,
+      "grad_norm": 0.35137006640434265,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 2477
+    },
+    {
+      "epoch": 0.2878884693581179,
+      "grad_norm": 0.3683086931705475,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2880046471100784,
+      "grad_norm": 0.4060046374797821,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 2479
+    },
+    {
+      "epoch": 0.28812082486203894,
+      "grad_norm": 0.39657512307167053,
+      "learning_rate": 0.0001,
+      "loss": 1.7616,
+      "step": 2480
+    },
+    {
+      "epoch": 0.2882370026139994,
+      "grad_norm": 0.38853272795677185,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 2481
+    },
+    {
+      "epoch": 0.2883531803659599,
+      "grad_norm": 0.39028510451316833,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 2482
+    },
+    {
+      "epoch": 0.2884693581179204,
+      "grad_norm": 0.3892669379711151,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2885855358698809,
+      "grad_norm": 0.3654754161834717,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 2484
+    },
+    {
+      "epoch": 0.28870171362184144,
+      "grad_norm": 0.3528408110141754,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 2485
+    },
+    {
+      "epoch": 0.2888178913738019,
+      "grad_norm": 0.4061609208583832,
+      "learning_rate": 0.0001,
+      "loss": 1.9495,
+      "step": 2486
+    },
+    {
+      "epoch": 0.2889340691257624,
+      "grad_norm": 0.4418545663356781,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2890502468777229,
+      "grad_norm": 0.3858497142791748,
+      "learning_rate": 0.0001,
+      "loss": 1.6543,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2891664246296834,
+      "grad_norm": 0.43995893001556396,
+      "learning_rate": 0.0001,
+      "loss": 1.7826,
+      "step": 2489
+    },
+    {
+      "epoch": 0.28928260238164394,
+      "grad_norm": 0.3793260157108307,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2893987801336044,
+      "grad_norm": 0.3741684556007385,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2895149578855649,
+      "grad_norm": 0.3923763632774353,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 2492
+    },
+    {
+      "epoch": 0.2896311356375254,
+      "grad_norm": 0.3590010702610016,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2897473133894859,
+      "grad_norm": 0.37135443091392517,
+      "learning_rate": 0.0001,
+      "loss": 1.6868,
+      "step": 2494
+    },
+    {
+      "epoch": 0.28986349114144644,
+      "grad_norm": 0.36070337891578674,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2899796688934069,
+      "grad_norm": 0.3781169652938843,
+      "learning_rate": 0.0001,
+      "loss": 1.5814,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2900958466453674,
+      "grad_norm": 0.3880577087402344,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 2497
+    },
+    {
+      "epoch": 0.2902120243973279,
+      "grad_norm": 0.38813501596450806,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 2498
+    },
+    {
+      "epoch": 0.2903282021492884,
+      "grad_norm": 0.4048740267753601,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2904443799012489,
+      "grad_norm": 0.35958972573280334,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 2500
+    },
+    {
+      "epoch": 0.29056055765320943,
+      "grad_norm": 0.38591668009757996,
+      "learning_rate": 0.0001,
+      "loss": 1.7163,
+      "step": 2501
+    },
+    {
+      "epoch": 0.2906767354051699,
+      "grad_norm": 0.37877535820007324,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2907929131571304,
+      "grad_norm": 0.39159923791885376,
+      "learning_rate": 0.0001,
+      "loss": 1.8899,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2909090909090909,
+      "grad_norm": 0.4107246994972229,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 2504
+    },
+    {
+      "epoch": 0.2910252686610514,
+      "grad_norm": 0.39638829231262207,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 2505
+    },
+    {
+      "epoch": 0.29114144641301193,
+      "grad_norm": 0.3886748254299164,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2912576241649724,
+      "grad_norm": 0.34036189317703247,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2913738019169329,
+      "grad_norm": 0.4072575867176056,
+      "learning_rate": 0.0001,
+      "loss": 1.7982,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2914899796688934,
+      "grad_norm": 0.38684821128845215,
+      "learning_rate": 0.0001,
+      "loss": 1.6675,
+      "step": 2509
+    },
+    {
+      "epoch": 0.2916061574208539,
+      "grad_norm": 0.3691238462924957,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 2510
+    },
+    {
+      "epoch": 0.29172233517281443,
+      "grad_norm": 0.39206942915916443,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2918385129247749,
+      "grad_norm": 0.3901553452014923,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 2512
+    },
+    {
+      "epoch": 0.2919546906767354,
+      "grad_norm": 0.37073972821235657,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 2513
+    },
+    {
+      "epoch": 0.2920708684286959,
+      "grad_norm": 0.3903072476387024,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2921870461806564,
+      "grad_norm": 0.4193265438079834,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 2515
+    },
+    {
+      "epoch": 0.29230322393261693,
+      "grad_norm": 0.4069857895374298,
+      "learning_rate": 0.0001,
+      "loss": 1.7801,
+      "step": 2516
+    },
+    {
+      "epoch": 0.2924194016845774,
+      "grad_norm": 0.40134382247924805,
+      "learning_rate": 0.0001,
+      "loss": 1.8263,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2925355794365379,
+      "grad_norm": 0.39630135893821716,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2926517571884984,
+      "grad_norm": 0.37301602959632874,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 2519
+    },
+    {
+      "epoch": 0.2927679349404589,
+      "grad_norm": 0.3810461759567261,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 2520
+    },
+    {
+      "epoch": 0.29288411269241943,
+      "grad_norm": 0.36505237221717834,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2930002904443799,
+      "grad_norm": 0.3974515199661255,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2931164681963404,
+      "grad_norm": 0.3713068664073944,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2932326459483009,
+      "grad_norm": 0.4444441497325897,
+      "learning_rate": 0.0001,
+      "loss": 1.9543,
+      "step": 2524
+    },
+    {
+      "epoch": 0.2933488237002614,
+      "grad_norm": 0.37668412923812866,
+      "learning_rate": 0.0001,
+      "loss": 1.7126,
+      "step": 2525
+    },
+    {
+      "epoch": 0.2934650014522219,
+      "grad_norm": 0.40829259157180786,
+      "learning_rate": 0.0001,
+      "loss": 1.8668,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2935811792041824,
+      "grad_norm": 0.3872903287410736,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2936973569561429,
+      "grad_norm": 0.4001368284225464,
+      "learning_rate": 0.0001,
+      "loss": 1.7448,
+      "step": 2528
+    },
+    {
+      "epoch": 0.2938135347081034,
+      "grad_norm": 0.36894676089286804,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2939297124600639,
+      "grad_norm": 0.4085611402988434,
+      "learning_rate": 0.0001,
+      "loss": 1.8068,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2940458902120244,
+      "grad_norm": 0.3942314386367798,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 2531
+    },
+    {
+      "epoch": 0.2941620679639849,
+      "grad_norm": 0.3652056157588959,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 2532
+    },
+    {
+      "epoch": 0.2942782457159454,
+      "grad_norm": 0.39049550890922546,
+      "learning_rate": 0.0001,
+      "loss": 1.7542,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2943944234679059,
+      "grad_norm": 0.38500627875328064,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2945106012198664,
+      "grad_norm": 0.372928649187088,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2946267789718269,
+      "grad_norm": 0.4132663607597351,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 2536
+    },
+    {
+      "epoch": 0.2947429567237874,
+      "grad_norm": 0.41574445366859436,
+      "learning_rate": 0.0001,
+      "loss": 1.7713,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2948591344757479,
+      "grad_norm": 0.3948863744735718,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2949753122277084,
+      "grad_norm": 0.3811478018760681,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2950914899796689,
+      "grad_norm": 0.3667415678501129,
+      "learning_rate": 0.0001,
+      "loss": 1.7235,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2952076677316294,
+      "grad_norm": 0.41427987813949585,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2953238454835899,
+      "grad_norm": 0.4144713282585144,
+      "learning_rate": 0.0001,
+      "loss": 1.7562,
+      "step": 2542
+    },
+    {
+      "epoch": 0.2954400232355504,
+      "grad_norm": 0.3656145930290222,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 2543
+    },
+    {
+      "epoch": 0.2955562009875109,
+      "grad_norm": 0.3743351697921753,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2956723787394714,
+      "grad_norm": 0.4163815379142761,
+      "learning_rate": 0.0001,
+      "loss": 1.7352,
+      "step": 2545
+    },
+    {
+      "epoch": 0.2957885564914319,
+      "grad_norm": 0.3973884582519531,
+      "learning_rate": 0.0001,
+      "loss": 1.8309,
+      "step": 2546
+    },
+    {
+      "epoch": 0.29590473424339236,
+      "grad_norm": 0.4104848802089691,
+      "learning_rate": 0.0001,
+      "loss": 1.7723,
+      "step": 2547
+    },
+    {
+      "epoch": 0.2960209119953529,
+      "grad_norm": 0.38057518005371094,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 2548
+    },
+    {
+      "epoch": 0.2961370897473134,
+      "grad_norm": 0.3594435751438141,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2962532674992739,
+      "grad_norm": 0.3822750747203827,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2963694452512344,
+      "grad_norm": 0.4234026372432709,
+      "learning_rate": 0.0001,
+      "loss": 1.7732,
+      "step": 2551
+    },
+    {
+      "epoch": 0.29648562300319486,
+      "grad_norm": 0.3969663083553314,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2966018007551554,
+      "grad_norm": 0.36822569370269775,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 2553
+    },
+    {
+      "epoch": 0.2967179785071159,
+      "grad_norm": 0.4209086298942566,
+      "learning_rate": 0.0001,
+      "loss": 1.9934,
+      "step": 2554
+    },
+    {
+      "epoch": 0.2968341562590764,
+      "grad_norm": 0.42897096276283264,
+      "learning_rate": 0.0001,
+      "loss": 1.733,
+      "step": 2555
+    },
+    {
+      "epoch": 0.2969503340110369,
+      "grad_norm": 0.36142057180404663,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 2556
+    },
+    {
+      "epoch": 0.29706651176299737,
+      "grad_norm": 0.38266682624816895,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2971826895149579,
+      "grad_norm": 0.3493916094303131,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 2558
+    },
+    {
+      "epoch": 0.2972988672669184,
+      "grad_norm": 0.3982014060020447,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 2559
+    },
+    {
+      "epoch": 0.2974150450188789,
+      "grad_norm": 0.3709663450717926,
+      "learning_rate": 0.0001,
+      "loss": 1.7749,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2975312227708394,
+      "grad_norm": 0.3823295831680298,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 2561
+    },
+    {
+      "epoch": 0.29764740052279987,
+      "grad_norm": 0.36725711822509766,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2977635782747604,
+      "grad_norm": 0.407084584236145,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 2563
+    },
+    {
+      "epoch": 0.2978797560267209,
+      "grad_norm": 0.3887426555156708,
+      "learning_rate": 0.0001,
+      "loss": 1.7267,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2979959337786814,
+      "grad_norm": 0.3916597068309784,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 2565
+    },
+    {
+      "epoch": 0.2981121115306419,
+      "grad_norm": 0.3767833113670349,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 2566
+    },
+    {
+      "epoch": 0.29822828928260237,
+      "grad_norm": 0.3760228753089905,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 2567
+    },
+    {
+      "epoch": 0.29834446703456285,
+      "grad_norm": 0.3806806802749634,
+      "learning_rate": 0.0001,
+      "loss": 1.5708,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2984606447865234,
+      "grad_norm": 0.38834068179130554,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 2569
+    },
+    {
+      "epoch": 0.2985768225384839,
+      "grad_norm": 0.4086969196796417,
+      "learning_rate": 0.0001,
+      "loss": 1.7382,
+      "step": 2570
+    },
+    {
+      "epoch": 0.2986930002904444,
+      "grad_norm": 0.3873027265071869,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 2571
+    },
+    {
+      "epoch": 0.29880917804240487,
+      "grad_norm": 0.38422465324401855,
+      "learning_rate": 0.0001,
+      "loss": 1.7938,
+      "step": 2572
+    },
+    {
+      "epoch": 0.29892535579436535,
+      "grad_norm": 0.3803096115589142,
+      "learning_rate": 0.0001,
+      "loss": 1.7387,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2990415335463259,
+      "grad_norm": 0.41523873805999756,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 2574
+    },
+    {
+      "epoch": 0.2991577112982864,
+      "grad_norm": 0.4138016104698181,
+      "learning_rate": 0.0001,
+      "loss": 1.9156,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2992738890502469,
+      "grad_norm": 0.3932843506336212,
+      "learning_rate": 0.0001,
+      "loss": 1.8139,
+      "step": 2576
+    },
+    {
+      "epoch": 0.29939006680220737,
+      "grad_norm": 0.37751099467277527,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 2577
+    },
+    {
+      "epoch": 0.29950624455416786,
+      "grad_norm": 0.37788084149360657,
+      "learning_rate": 0.0001,
+      "loss": 1.6342,
+      "step": 2578
+    },
+    {
+      "epoch": 0.2996224223061284,
+      "grad_norm": 0.381740927696228,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 2579
+    },
+    {
+      "epoch": 0.2997386000580889,
+      "grad_norm": 0.38412848114967346,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2998547778100494,
+      "grad_norm": 0.3746092617511749,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 2581
+    },
+    {
+      "epoch": 0.29997095556200987,
+      "grad_norm": 0.3701321482658386,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 2582
+    },
+    {
+      "epoch": 0.30008713331397036,
+      "grad_norm": 0.3820204436779022,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 2583
+    },
+    {
+      "epoch": 0.3002033110659309,
+      "grad_norm": 0.3748900592327118,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 2584
+    },
+    {
+      "epoch": 0.3003194888178914,
+      "grad_norm": 0.3851977288722992,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 2585
+    },
+    {
+      "epoch": 0.3004356665698519,
+      "grad_norm": 0.3978486955165863,
+      "learning_rate": 0.0001,
+      "loss": 1.8482,
+      "step": 2586
+    },
+    {
+      "epoch": 0.30055184432181237,
+      "grad_norm": 0.402482807636261,
+      "learning_rate": 0.0001,
+      "loss": 1.8079,
+      "step": 2587
+    },
+    {
+      "epoch": 0.30066802207377286,
+      "grad_norm": 0.41250428557395935,
+      "learning_rate": 0.0001,
+      "loss": 1.7092,
+      "step": 2588
+    },
+    {
+      "epoch": 0.30078419982573334,
+      "grad_norm": 0.4030408561229706,
+      "learning_rate": 0.0001,
+      "loss": 1.8394,
+      "step": 2589
+    },
+    {
+      "epoch": 0.3009003775776939,
+      "grad_norm": 0.3963468670845032,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 2590
+    },
+    {
+      "epoch": 0.3010165553296544,
+      "grad_norm": 0.37878894805908203,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 2591
+    },
+    {
+      "epoch": 0.30113273308161487,
+      "grad_norm": 0.3783760666847229,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 2592
+    },
+    {
+      "epoch": 0.30124891083357536,
+      "grad_norm": 0.3853001892566681,
+      "learning_rate": 0.0001,
+      "loss": 1.8408,
+      "step": 2593
+    },
+    {
+      "epoch": 0.30136508858553585,
+      "grad_norm": 0.37395623326301575,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 2594
+    },
+    {
+      "epoch": 0.3014812663374964,
+      "grad_norm": 0.3772829473018646,
+      "learning_rate": 0.0001,
+      "loss": 1.6901,
+      "step": 2595
+    },
+    {
+      "epoch": 0.3015974440894569,
+      "grad_norm": 0.3889354467391968,
+      "learning_rate": 0.0001,
+      "loss": 1.7094,
+      "step": 2596
+    },
+    {
+      "epoch": 0.30171362184141737,
+      "grad_norm": 0.4137793481349945,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 2597
+    },
+    {
+      "epoch": 0.30182979959337786,
+      "grad_norm": 0.37697646021842957,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 2598
+    },
+    {
+      "epoch": 0.30194597734533835,
+      "grad_norm": 0.3857576847076416,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 2599
+    },
+    {
+      "epoch": 0.3020621550972989,
+      "grad_norm": 0.41573137044906616,
+      "learning_rate": 0.0001,
+      "loss": 1.8468,
+      "step": 2600
+    },
+    {
+      "epoch": 0.3021783328492594,
+      "grad_norm": 0.3891277313232422,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 2601
+    },
+    {
+      "epoch": 0.30229451060121987,
+      "grad_norm": 0.3497909605503082,
+      "learning_rate": 0.0001,
+      "loss": 1.398,
+      "step": 2602
+    },
+    {
+      "epoch": 0.30241068835318036,
+      "grad_norm": 0.3606802523136139,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 2603
+    },
+    {
+      "epoch": 0.30252686610514085,
+      "grad_norm": 0.4066804349422455,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 2604
+    },
+    {
+      "epoch": 0.3026430438571014,
+      "grad_norm": 0.4184049367904663,
+      "learning_rate": 0.0001,
+      "loss": 1.7302,
+      "step": 2605
+    },
+    {
+      "epoch": 0.3027592216090619,
+      "grad_norm": 0.3844599723815918,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 2606
+    },
+    {
+      "epoch": 0.30287539936102237,
+      "grad_norm": 0.4283202290534973,
+      "learning_rate": 0.0001,
+      "loss": 1.8115,
+      "step": 2607
+    },
+    {
+      "epoch": 0.30299157711298286,
+      "grad_norm": 0.4154917895793915,
+      "learning_rate": 0.0001,
+      "loss": 1.7869,
+      "step": 2608
+    },
+    {
+      "epoch": 0.30310775486494335,
+      "grad_norm": 0.4059176445007324,
+      "learning_rate": 0.0001,
+      "loss": 1.8573,
+      "step": 2609
+    },
+    {
+      "epoch": 0.3032239326169039,
+      "grad_norm": 0.38833245635032654,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 2610
+    },
+    {
+      "epoch": 0.3033401103688644,
+      "grad_norm": 0.3859376013278961,
+      "learning_rate": 0.0001,
+      "loss": 1.7628,
+      "step": 2611
+    },
+    {
+      "epoch": 0.30345628812082487,
+      "grad_norm": 0.3899228572845459,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 2612
+    },
+    {
+      "epoch": 0.30357246587278536,
+      "grad_norm": 0.38399478793144226,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 2613
+    },
+    {
+      "epoch": 0.30368864362474585,
+      "grad_norm": 0.440913587808609,
+      "learning_rate": 0.0001,
+      "loss": 1.8794,
+      "step": 2614
+    },
+    {
+      "epoch": 0.30380482137670634,
+      "grad_norm": 0.3857106864452362,
+      "learning_rate": 0.0001,
+      "loss": 1.7513,
+      "step": 2615
+    },
+    {
+      "epoch": 0.3039209991286669,
+      "grad_norm": 0.3604831099510193,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 2616
+    },
+    {
+      "epoch": 0.30403717688062737,
+      "grad_norm": 0.3680822551250458,
+      "learning_rate": 0.0001,
+      "loss": 1.6749,
+      "step": 2617
+    },
+    {
+      "epoch": 0.30415335463258786,
+      "grad_norm": 0.3898998498916626,
+      "learning_rate": 0.0001,
+      "loss": 1.7768,
+      "step": 2618
+    },
+    {
+      "epoch": 0.30426953238454835,
+      "grad_norm": 0.3705196678638458,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 2619
+    },
+    {
+      "epoch": 0.30438571013650884,
+      "grad_norm": 0.3996119797229767,
+      "learning_rate": 0.0001,
+      "loss": 1.7945,
+      "step": 2620
+    },
+    {
+      "epoch": 0.3045018878884694,
+      "grad_norm": 0.391072541475296,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 2621
+    },
+    {
+      "epoch": 0.30461806564042987,
+      "grad_norm": 0.5729893445968628,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 2622
+    },
+    {
+      "epoch": 0.30473424339239036,
+      "grad_norm": 0.38251304626464844,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 2623
+    },
+    {
+      "epoch": 0.30485042114435085,
+      "grad_norm": 0.3925994634628296,
+      "learning_rate": 0.0001,
+      "loss": 1.7669,
+      "step": 2624
+    },
+    {
+      "epoch": 0.30496659889631134,
+      "grad_norm": 0.3850267827510834,
+      "learning_rate": 0.0001,
+      "loss": 1.7181,
+      "step": 2625
+    },
+    {
+      "epoch": 0.3050827766482719,
+      "grad_norm": 0.37147024273872375,
+      "learning_rate": 0.0001,
+      "loss": 1.6673,
+      "step": 2626
+    },
+    {
+      "epoch": 0.30519895440023237,
+      "grad_norm": 0.41201236844062805,
+      "learning_rate": 0.0001,
+      "loss": 1.7335,
+      "step": 2627
+    },
+    {
+      "epoch": 0.30531513215219286,
+      "grad_norm": 0.4249459505081177,
+      "learning_rate": 0.0001,
+      "loss": 1.8026,
+      "step": 2628
+    },
+    {
+      "epoch": 0.30543130990415335,
+      "grad_norm": 0.37525978684425354,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 2629
+    },
+    {
+      "epoch": 0.30554748765611384,
+      "grad_norm": 0.3670780658721924,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 2630
+    },
+    {
+      "epoch": 0.3056636654080744,
+      "grad_norm": 0.37397563457489014,
+      "learning_rate": 0.0001,
+      "loss": 1.7848,
+      "step": 2631
+    },
+    {
+      "epoch": 0.30577984316003487,
+      "grad_norm": 0.35987600684165955,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 2632
+    },
+    {
+      "epoch": 0.30589602091199536,
+      "grad_norm": 0.3674522340297699,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 2633
+    },
+    {
+      "epoch": 0.30601219866395585,
+      "grad_norm": 0.3551689684391022,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 2634
+    },
+    {
+      "epoch": 0.30612837641591634,
+      "grad_norm": 0.39974987506866455,
+      "learning_rate": 0.0001,
+      "loss": 1.7497,
+      "step": 2635
+    },
+    {
+      "epoch": 0.3062445541678768,
+      "grad_norm": 0.37489983439445496,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 2636
+    },
+    {
+      "epoch": 0.30636073191983737,
+      "grad_norm": 0.38141462206840515,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 2637
+    },
+    {
+      "epoch": 0.30647690967179786,
+      "grad_norm": 0.38110142946243286,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 2638
+    },
+    {
+      "epoch": 0.30659308742375835,
+      "grad_norm": 0.3580850064754486,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 2639
+    },
+    {
+      "epoch": 0.30670926517571884,
+      "grad_norm": 0.4292284846305847,
+      "learning_rate": 0.0001,
+      "loss": 1.7091,
+      "step": 2640
+    },
+    {
+      "epoch": 0.3068254429276793,
+      "grad_norm": 0.37014874815940857,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 2641
+    },
+    {
+      "epoch": 0.30694162067963987,
+      "grad_norm": 0.36894306540489197,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 2642
+    },
+    {
+      "epoch": 0.30705779843160036,
+      "grad_norm": 0.37392929196357727,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 2643
+    },
+    {
+      "epoch": 0.30717397618356085,
+      "grad_norm": 0.3983522653579712,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 2644
+    },
+    {
+      "epoch": 0.30729015393552134,
+      "grad_norm": 0.3379395604133606,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3074063316874818,
+      "grad_norm": 0.35767263174057007,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 2646
+    },
+    {
+      "epoch": 0.30752250943944237,
+      "grad_norm": 0.3817872405052185,
+      "learning_rate": 0.0001,
+      "loss": 1.7924,
+      "step": 2647
+    },
+    {
+      "epoch": 0.30763868719140286,
+      "grad_norm": 0.3904268741607666,
+      "learning_rate": 0.0001,
+      "loss": 1.6342,
+      "step": 2648
+    },
+    {
+      "epoch": 0.30775486494336335,
+      "grad_norm": 0.37173137068748474,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 2649
+    },
+    {
+      "epoch": 0.30787104269532384,
+      "grad_norm": 0.3717529773712158,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 2650
+    },
+    {
+      "epoch": 0.3079872204472843,
+      "grad_norm": 0.3810470998287201,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 2651
+    },
+    {
+      "epoch": 0.30810339819924487,
+      "grad_norm": 0.3781997263431549,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 2652
+    },
+    {
+      "epoch": 0.30821957595120536,
+      "grad_norm": 0.4403388798236847,
+      "learning_rate": 0.0001,
+      "loss": 1.7052,
+      "step": 2653
+    },
+    {
+      "epoch": 0.30833575370316585,
+      "grad_norm": 0.3832712769508362,
+      "learning_rate": 0.0001,
+      "loss": 1.7543,
+      "step": 2654
+    },
+    {
+      "epoch": 0.30845193145512634,
+      "grad_norm": 0.3549423813819885,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 2655
+    },
+    {
+      "epoch": 0.3085681092070868,
+      "grad_norm": 0.37209948897361755,
+      "learning_rate": 0.0001,
+      "loss": 1.6505,
+      "step": 2656
+    },
+    {
+      "epoch": 0.3086842869590473,
+      "grad_norm": 0.39987820386886597,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 2657
+    },
+    {
+      "epoch": 0.30880046471100786,
+      "grad_norm": 0.40743595361709595,
+      "learning_rate": 0.0001,
+      "loss": 1.8868,
+      "step": 2658
+    },
+    {
+      "epoch": 0.30891664246296835,
+      "grad_norm": 0.41529178619384766,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 2659
+    },
+    {
+      "epoch": 0.30903282021492884,
+      "grad_norm": 0.420540988445282,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 2660
+    },
+    {
+      "epoch": 0.3091489979668893,
+      "grad_norm": 0.38582152128219604,
+      "learning_rate": 0.0001,
+      "loss": 1.8365,
+      "step": 2661
+    },
+    {
+      "epoch": 0.3092651757188498,
+      "grad_norm": 0.382856547832489,
+      "learning_rate": 0.0001,
+      "loss": 1.7678,
+      "step": 2662
+    },
+    {
+      "epoch": 0.30938135347081036,
+      "grad_norm": 0.3981390595436096,
+      "learning_rate": 0.0001,
+      "loss": 1.7592,
+      "step": 2663
+    },
+    {
+      "epoch": 0.30949753122277085,
+      "grad_norm": 0.4064244031906128,
+      "learning_rate": 0.0001,
+      "loss": 1.783,
+      "step": 2664
+    },
+    {
+      "epoch": 0.30961370897473134,
+      "grad_norm": 0.3797922730445862,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 2665
+    },
+    {
+      "epoch": 0.3097298867266918,
+      "grad_norm": 0.366719514131546,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 2666
+    },
+    {
+      "epoch": 0.3098460644786523,
+      "grad_norm": 0.3816395103931427,
+      "learning_rate": 0.0001,
+      "loss": 1.7305,
+      "step": 2667
+    },
+    {
+      "epoch": 0.30996224223061286,
+      "grad_norm": 0.3834119141101837,
+      "learning_rate": 0.0001,
+      "loss": 1.6341,
+      "step": 2668
+    },
+    {
+      "epoch": 0.31007841998257335,
+      "grad_norm": 0.3797139525413513,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 2669
+    },
+    {
+      "epoch": 0.31019459773453384,
+      "grad_norm": 0.3854397237300873,
+      "learning_rate": 0.0001,
+      "loss": 1.7163,
+      "step": 2670
+    },
+    {
+      "epoch": 0.31031077548649433,
+      "grad_norm": 0.3634074330329895,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 2671
+    },
+    {
+      "epoch": 0.3104269532384548,
+      "grad_norm": 0.37497079372406006,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 2672
+    },
+    {
+      "epoch": 0.31054313099041536,
+      "grad_norm": 0.39349761605262756,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 2673
+    },
+    {
+      "epoch": 0.31065930874237585,
+      "grad_norm": 0.4020683765411377,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 2674
+    },
+    {
+      "epoch": 0.31077548649433634,
+      "grad_norm": 0.4030044376850128,
+      "learning_rate": 0.0001,
+      "loss": 1.7939,
+      "step": 2675
+    },
+    {
+      "epoch": 0.31089166424629683,
+      "grad_norm": 0.3698487877845764,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 2676
+    },
+    {
+      "epoch": 0.3110078419982573,
+      "grad_norm": 0.39510759711265564,
+      "learning_rate": 0.0001,
+      "loss": 1.4885,
+      "step": 2677
+    },
+    {
+      "epoch": 0.3111240197502178,
+      "grad_norm": 0.3986801207065582,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 2678
+    },
+    {
+      "epoch": 0.31124019750217835,
+      "grad_norm": 0.39601030945777893,
+      "learning_rate": 0.0001,
+      "loss": 1.7573,
+      "step": 2679
+    },
+    {
+      "epoch": 0.31135637525413884,
+      "grad_norm": 0.42814406752586365,
+      "learning_rate": 0.0001,
+      "loss": 1.7842,
+      "step": 2680
+    },
+    {
+      "epoch": 0.31147255300609933,
+      "grad_norm": 0.39685603976249695,
+      "learning_rate": 0.0001,
+      "loss": 1.886,
+      "step": 2681
+    },
+    {
+      "epoch": 0.3115887307580598,
+      "grad_norm": 0.4117811322212219,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 2682
+    },
+    {
+      "epoch": 0.3117049085100203,
+      "grad_norm": 0.40006887912750244,
+      "learning_rate": 0.0001,
+      "loss": 1.7719,
+      "step": 2683
+    },
+    {
+      "epoch": 0.31182108626198085,
+      "grad_norm": 0.4050239026546478,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 2684
+    },
+    {
+      "epoch": 0.31193726401394134,
+      "grad_norm": 0.38189923763275146,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 2685
+    },
+    {
+      "epoch": 0.31205344176590183,
+      "grad_norm": 0.4016052484512329,
+      "learning_rate": 0.0001,
+      "loss": 1.6297,
+      "step": 2686
+    },
+    {
+      "epoch": 0.3121696195178623,
+      "grad_norm": 0.37489327788352966,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 2687
+    },
+    {
+      "epoch": 0.3122857972698228,
+      "grad_norm": 0.3983065187931061,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 2688
+    },
+    {
+      "epoch": 0.31240197502178335,
+      "grad_norm": 0.40546756982803345,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 2689
+    },
+    {
+      "epoch": 0.31251815277374384,
+      "grad_norm": 0.37291330099105835,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 2690
+    },
+    {
+      "epoch": 0.31263433052570433,
+      "grad_norm": 0.3653269410133362,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 2691
+    },
+    {
+      "epoch": 0.3127505082776648,
+      "grad_norm": 0.3700310289859772,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 2692
+    },
+    {
+      "epoch": 0.3128666860296253,
+      "grad_norm": 0.38666659593582153,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 2693
+    },
+    {
+      "epoch": 0.31298286378158585,
+      "grad_norm": 0.3867506980895996,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 2694
+    },
+    {
+      "epoch": 0.31309904153354634,
+      "grad_norm": 0.38117802143096924,
+      "learning_rate": 0.0001,
+      "loss": 1.7902,
+      "step": 2695
+    },
+    {
+      "epoch": 0.31321521928550683,
+      "grad_norm": 0.403726726770401,
+      "learning_rate": 0.0001,
+      "loss": 1.8469,
+      "step": 2696
+    },
+    {
+      "epoch": 0.3133313970374673,
+      "grad_norm": 0.3811923861503601,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 2697
+    },
+    {
+      "epoch": 0.3134475747894278,
+      "grad_norm": 0.3940551280975342,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 2698
+    },
+    {
+      "epoch": 0.3135637525413883,
+      "grad_norm": 0.3885653614997864,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 2699
+    },
+    {
+      "epoch": 0.31367993029334884,
+      "grad_norm": 0.36595413088798523,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 2700
+    },
+    {
+      "epoch": 0.31379610804530933,
+      "grad_norm": 0.38449594378471375,
+      "learning_rate": 0.0001,
+      "loss": 1.7474,
+      "step": 2701
+    },
+    {
+      "epoch": 0.3139122857972698,
+      "grad_norm": 0.3768537640571594,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 2702
+    },
+    {
+      "epoch": 0.3140284635492303,
+      "grad_norm": 0.3723905384540558,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 2703
+    },
+    {
+      "epoch": 0.3141446413011908,
+      "grad_norm": 0.3800404667854309,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 2704
+    },
+    {
+      "epoch": 0.31426081905315134,
+      "grad_norm": 0.38999584317207336,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 2705
+    },
+    {
+      "epoch": 0.31437699680511183,
+      "grad_norm": 0.40568429231643677,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 2706
+    },
+    {
+      "epoch": 0.3144931745570723,
+      "grad_norm": 0.3945614993572235,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 2707
+    },
+    {
+      "epoch": 0.3146093523090328,
+      "grad_norm": 0.36772727966308594,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 2708
+    },
+    {
+      "epoch": 0.3147255300609933,
+      "grad_norm": 0.38640451431274414,
+      "learning_rate": 0.0001,
+      "loss": 1.677,
+      "step": 2709
+    },
+    {
+      "epoch": 0.31484170781295384,
+      "grad_norm": 0.3605159819126129,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 2710
+    },
+    {
+      "epoch": 0.31495788556491433,
+      "grad_norm": 0.38972237706184387,
+      "learning_rate": 0.0001,
+      "loss": 1.801,
+      "step": 2711
+    },
+    {
+      "epoch": 0.3150740633168748,
+      "grad_norm": 0.44629716873168945,
+      "learning_rate": 0.0001,
+      "loss": 1.7858,
+      "step": 2712
+    },
+    {
+      "epoch": 0.3151902410688353,
+      "grad_norm": 0.39029544591903687,
+      "learning_rate": 0.0001,
+      "loss": 1.7399,
+      "step": 2713
+    },
+    {
+      "epoch": 0.3153064188207958,
+      "grad_norm": 0.37057486176490784,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 2714
+    },
+    {
+      "epoch": 0.31542259657275634,
+      "grad_norm": 0.38961511850357056,
+      "learning_rate": 0.0001,
+      "loss": 1.7939,
+      "step": 2715
+    },
+    {
+      "epoch": 0.31553877432471683,
+      "grad_norm": 0.3788926899433136,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 2716
+    },
+    {
+      "epoch": 0.3156549520766773,
+      "grad_norm": 0.3727104961872101,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 2717
+    },
+    {
+      "epoch": 0.3157711298286378,
+      "grad_norm": 0.38792872428894043,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 2718
+    },
+    {
+      "epoch": 0.3158873075805983,
+      "grad_norm": 0.39093390107154846,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 2719
+    },
+    {
+      "epoch": 0.31600348533255884,
+      "grad_norm": 0.39481261372566223,
+      "learning_rate": 0.0001,
+      "loss": 1.7409,
+      "step": 2720
+    },
+    {
+      "epoch": 0.31611966308451933,
+      "grad_norm": 0.39427947998046875,
+      "learning_rate": 0.0001,
+      "loss": 1.7673,
+      "step": 2721
+    },
+    {
+      "epoch": 0.3162358408364798,
+      "grad_norm": 0.3655182719230652,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 2722
+    },
+    {
+      "epoch": 0.3163520185884403,
+      "grad_norm": 0.4002794921398163,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 2723
+    },
+    {
+      "epoch": 0.3164681963404008,
+      "grad_norm": 0.4090864360332489,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 2724
+    },
+    {
+      "epoch": 0.3165843740923613,
+      "grad_norm": 0.3690468370914459,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 2725
+    },
+    {
+      "epoch": 0.31670055184432183,
+      "grad_norm": 0.3945186734199524,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 2726
+    },
+    {
+      "epoch": 0.3168167295962823,
+      "grad_norm": 0.3898504972457886,
+      "learning_rate": 0.0001,
+      "loss": 1.6971,
+      "step": 2727
+    },
+    {
+      "epoch": 0.3169329073482428,
+      "grad_norm": 0.38657867908477783,
+      "learning_rate": 0.0001,
+      "loss": 1.8112,
+      "step": 2728
+    },
+    {
+      "epoch": 0.3170490851002033,
+      "grad_norm": 0.39374595880508423,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 2729
+    },
+    {
+      "epoch": 0.3171652628521638,
+      "grad_norm": 0.3979948163032532,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 2730
+    },
+    {
+      "epoch": 0.31728144060412433,
+      "grad_norm": 0.3945339620113373,
+      "learning_rate": 0.0001,
+      "loss": 1.7315,
+      "step": 2731
+    },
+    {
+      "epoch": 0.3173976183560848,
+      "grad_norm": 0.4169275760650635,
+      "learning_rate": 0.0001,
+      "loss": 1.7789,
+      "step": 2732
+    },
+    {
+      "epoch": 0.3175137961080453,
+      "grad_norm": 0.37716567516326904,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 2733
+    },
+    {
+      "epoch": 0.3176299738600058,
+      "grad_norm": 0.38198474049568176,
+      "learning_rate": 0.0001,
+      "loss": 1.7881,
+      "step": 2734
+    },
+    {
+      "epoch": 0.3177461516119663,
+      "grad_norm": 0.3758987486362457,
+      "learning_rate": 0.0001,
+      "loss": 1.5993,
+      "step": 2735
+    },
+    {
+      "epoch": 0.31786232936392683,
+      "grad_norm": 0.4033893644809723,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 2736
+    },
+    {
+      "epoch": 0.3179785071158873,
+      "grad_norm": 0.3776089549064636,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 2737
+    },
+    {
+      "epoch": 0.3180946848678478,
+      "grad_norm": 0.38801753520965576,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 2738
+    },
+    {
+      "epoch": 0.3182108626198083,
+      "grad_norm": 0.3814786970615387,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 2739
+    },
+    {
+      "epoch": 0.3183270403717688,
+      "grad_norm": 0.35588279366493225,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 2740
+    },
+    {
+      "epoch": 0.31844321812372933,
+      "grad_norm": 0.378668874502182,
+      "learning_rate": 0.0001,
+      "loss": 1.6447,
+      "step": 2741
+    },
+    {
+      "epoch": 0.3185593958756898,
+      "grad_norm": 0.4211976230144501,
+      "learning_rate": 0.0001,
+      "loss": 1.882,
+      "step": 2742
+    },
+    {
+      "epoch": 0.3186755736276503,
+      "grad_norm": 0.36686578392982483,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 2743
+    },
+    {
+      "epoch": 0.3187917513796108,
+      "grad_norm": 0.37785688042640686,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 2744
+    },
+    {
+      "epoch": 0.3189079291315713,
+      "grad_norm": 0.3932788372039795,
+      "learning_rate": 0.0001,
+      "loss": 1.7125,
+      "step": 2745
+    },
+    {
+      "epoch": 0.3190241068835318,
+      "grad_norm": 0.3749261200428009,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 2746
+    },
+    {
+      "epoch": 0.3191402846354923,
+      "grad_norm": 0.36720603704452515,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 2747
+    },
+    {
+      "epoch": 0.3192564623874528,
+      "grad_norm": 0.38770416378974915,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 2748
+    },
+    {
+      "epoch": 0.3193726401394133,
+      "grad_norm": 0.42165276408195496,
+      "learning_rate": 0.0001,
+      "loss": 1.8394,
+      "step": 2749
+    },
+    {
+      "epoch": 0.3194888178913738,
+      "grad_norm": 0.3547378480434418,
+      "learning_rate": 0.0001,
+      "loss": 1.7143,
+      "step": 2750
+    },
+    {
+      "epoch": 0.3196049956433343,
+      "grad_norm": 0.3825453519821167,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 2751
+    },
+    {
+      "epoch": 0.3197211733952948,
+      "grad_norm": 0.42200520634651184,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 2752
+    },
+    {
+      "epoch": 0.3198373511472553,
+      "grad_norm": 0.3694903552532196,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 2753
+    },
+    {
+      "epoch": 0.3199535288992158,
+      "grad_norm": 0.38652074337005615,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 2754
+    },
+    {
+      "epoch": 0.3200697066511763,
+      "grad_norm": 0.4044737219810486,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 2755
+    },
+    {
+      "epoch": 0.3201858844031368,
+      "grad_norm": 0.3883349597454071,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 2756
+    },
+    {
+      "epoch": 0.3203020621550973,
+      "grad_norm": 0.3901846408843994,
+      "learning_rate": 0.0001,
+      "loss": 1.8362,
+      "step": 2757
+    },
+    {
+      "epoch": 0.3204182399070578,
+      "grad_norm": 0.37919488549232483,
+      "learning_rate": 0.0001,
+      "loss": 1.6951,
+      "step": 2758
+    },
+    {
+      "epoch": 0.3205344176590183,
+      "grad_norm": 0.36535388231277466,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 2759
+    },
+    {
+      "epoch": 0.3206505954109788,
+      "grad_norm": 0.3907409608364105,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 2760
+    },
+    {
+      "epoch": 0.3207667731629393,
+      "grad_norm": 0.39692288637161255,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 2761
+    },
+    {
+      "epoch": 0.3208829509148998,
+      "grad_norm": 0.431797593832016,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 2762
+    },
+    {
+      "epoch": 0.3209991286668603,
+      "grad_norm": 0.38208791613578796,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 2763
+    },
+    {
+      "epoch": 0.3211153064188208,
+      "grad_norm": 0.40455830097198486,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 2764
+    },
+    {
+      "epoch": 0.3212314841707813,
+      "grad_norm": 0.3595045506954193,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 2765
+    },
+    {
+      "epoch": 0.3213476619227418,
+      "grad_norm": 0.42558008432388306,
+      "learning_rate": 0.0001,
+      "loss": 1.7101,
+      "step": 2766
+    },
+    {
+      "epoch": 0.32146383967470227,
+      "grad_norm": 0.39539337158203125,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 2767
+    },
+    {
+      "epoch": 0.3215800174266628,
+      "grad_norm": 0.3601813316345215,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 2768
+    },
+    {
+      "epoch": 0.3216961951786233,
+      "grad_norm": 0.3683129847049713,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 2769
+    },
+    {
+      "epoch": 0.3218123729305838,
+      "grad_norm": 0.4206804633140564,
+      "learning_rate": 0.0001,
+      "loss": 1.7475,
+      "step": 2770
+    },
+    {
+      "epoch": 0.3219285506825443,
+      "grad_norm": 0.3908296525478363,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 2771
+    },
+    {
+      "epoch": 0.32204472843450477,
+      "grad_norm": 0.38299795985221863,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 2772
+    },
+    {
+      "epoch": 0.3221609061864653,
+      "grad_norm": 0.41059038043022156,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 2773
+    },
+    {
+      "epoch": 0.3222770839384258,
+      "grad_norm": 0.3887813687324524,
+      "learning_rate": 0.0001,
+      "loss": 1.8025,
+      "step": 2774
+    },
+    {
+      "epoch": 0.3223932616903863,
+      "grad_norm": 0.36034008860588074,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 2775
+    },
+    {
+      "epoch": 0.3225094394423468,
+      "grad_norm": 0.3732610046863556,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 2776
+    },
+    {
+      "epoch": 0.32262561719430727,
+      "grad_norm": 0.40316662192344666,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 2777
+    },
+    {
+      "epoch": 0.3227417949462678,
+      "grad_norm": 0.40975120663642883,
+      "learning_rate": 0.0001,
+      "loss": 1.5921,
+      "step": 2778
+    },
+    {
+      "epoch": 0.3228579726982283,
+      "grad_norm": 0.38018473982810974,
+      "learning_rate": 0.0001,
+      "loss": 1.711,
+      "step": 2779
+    },
+    {
+      "epoch": 0.3229741504501888,
+      "grad_norm": 0.3749701678752899,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 2780
+    },
+    {
+      "epoch": 0.3230903282021493,
+      "grad_norm": 0.37560439109802246,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 2781
+    },
+    {
+      "epoch": 0.32320650595410977,
+      "grad_norm": 0.3560850918292999,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 2782
+    },
+    {
+      "epoch": 0.3233226837060703,
+      "grad_norm": 0.41308125853538513,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 2783
+    },
+    {
+      "epoch": 0.3234388614580308,
+      "grad_norm": 0.35252368450164795,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 2784
+    },
+    {
+      "epoch": 0.3235550392099913,
+      "grad_norm": 0.3724053204059601,
+      "learning_rate": 0.0001,
+      "loss": 1.739,
+      "step": 2785
+    },
+    {
+      "epoch": 0.3236712169619518,
+      "grad_norm": 0.4030148684978485,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 2786
+    },
+    {
+      "epoch": 0.32378739471391227,
+      "grad_norm": 0.3721768260002136,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 2787
+    },
+    {
+      "epoch": 0.32390357246587276,
+      "grad_norm": 0.34921830892562866,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 2788
+    },
+    {
+      "epoch": 0.3240197502178333,
+      "grad_norm": 0.39488157629966736,
+      "learning_rate": 0.0001,
+      "loss": 1.7164,
+      "step": 2789
+    },
+    {
+      "epoch": 0.3241359279697938,
+      "grad_norm": 0.3864257037639618,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 2790
+    },
+    {
+      "epoch": 0.3242521057217543,
+      "grad_norm": 0.41633668541908264,
+      "learning_rate": 0.0001,
+      "loss": 1.7591,
+      "step": 2791
+    },
+    {
+      "epoch": 0.32436828347371477,
+      "grad_norm": 0.38308605551719666,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 2792
+    },
+    {
+      "epoch": 0.32448446122567526,
+      "grad_norm": 0.37461498379707336,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 2793
+    },
+    {
+      "epoch": 0.3246006389776358,
+      "grad_norm": 0.4059557020664215,
+      "learning_rate": 0.0001,
+      "loss": 1.8018,
+      "step": 2794
+    },
+    {
+      "epoch": 0.3247168167295963,
+      "grad_norm": 0.37268683314323425,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 2795
+    },
+    {
+      "epoch": 0.3248329944815568,
+      "grad_norm": 0.3826591372489929,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 2796
+    },
+    {
+      "epoch": 0.32494917223351727,
+      "grad_norm": 0.38750702142715454,
+      "learning_rate": 0.0001,
+      "loss": 1.7622,
+      "step": 2797
+    },
+    {
+      "epoch": 0.32506534998547776,
+      "grad_norm": 0.4062412679195404,
+      "learning_rate": 0.0001,
+      "loss": 1.7934,
+      "step": 2798
+    },
+    {
+      "epoch": 0.3251815277374383,
+      "grad_norm": 0.3924409747123718,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 2799
+    },
+    {
+      "epoch": 0.3252977054893988,
+      "grad_norm": 0.3657173216342926,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 2800
+    },
+    {
+      "epoch": 0.3254138832413593,
+      "grad_norm": 0.37579992413520813,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 2801
+    },
+    {
+      "epoch": 0.32553006099331977,
+      "grad_norm": 0.3894909620285034,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 2802
+    },
+    {
+      "epoch": 0.32564623874528026,
+      "grad_norm": 0.3836139440536499,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 2803
+    },
+    {
+      "epoch": 0.3257624164972408,
+      "grad_norm": 0.40788355469703674,
+      "learning_rate": 0.0001,
+      "loss": 1.7918,
+      "step": 2804
+    },
+    {
+      "epoch": 0.3258785942492013,
+      "grad_norm": 0.4235386848449707,
+      "learning_rate": 0.0001,
+      "loss": 1.8649,
+      "step": 2805
+    },
+    {
+      "epoch": 0.3259947720011618,
+      "grad_norm": 0.39928755164146423,
+      "learning_rate": 0.0001,
+      "loss": 1.7444,
+      "step": 2806
+    },
+    {
+      "epoch": 0.32611094975312227,
+      "grad_norm": 0.40659260749816895,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 2807
+    },
+    {
+      "epoch": 0.32622712750508276,
+      "grad_norm": 0.38179177045822144,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 2808
+    },
+    {
+      "epoch": 0.3263433052570433,
+      "grad_norm": 0.38274791836738586,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 2809
+    },
+    {
+      "epoch": 0.3264594830090038,
+      "grad_norm": 0.3890819549560547,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 2810
+    },
+    {
+      "epoch": 0.3265756607609643,
+      "grad_norm": 0.37747570872306824,
+      "learning_rate": 0.0001,
+      "loss": 1.6623,
+      "step": 2811
+    },
+    {
+      "epoch": 0.32669183851292477,
+      "grad_norm": 0.38250067830085754,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 2812
+    },
+    {
+      "epoch": 0.32680801626488526,
+      "grad_norm": 0.3761623799800873,
+      "learning_rate": 0.0001,
+      "loss": 1.6855,
+      "step": 2813
+    },
+    {
+      "epoch": 0.32692419401684575,
+      "grad_norm": 0.3789752721786499,
+      "learning_rate": 0.0001,
+      "loss": 1.718,
+      "step": 2814
+    },
+    {
+      "epoch": 0.3270403717688063,
+      "grad_norm": 0.3886089622974396,
+      "learning_rate": 0.0001,
+      "loss": 1.7111,
+      "step": 2815
+    },
+    {
+      "epoch": 0.3271565495207668,
+      "grad_norm": 0.3989008665084839,
+      "learning_rate": 0.0001,
+      "loss": 1.8128,
+      "step": 2816
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 0.4016900360584259,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 2817
+    },
+    {
+      "epoch": 0.32738890502468776,
+      "grad_norm": 0.41296061873435974,
+      "learning_rate": 0.0001,
+      "loss": 1.8053,
+      "step": 2818
+    },
+    {
+      "epoch": 0.32750508277664825,
+      "grad_norm": 0.3775239586830139,
+      "learning_rate": 0.0001,
+      "loss": 1.6908,
+      "step": 2819
+    },
+    {
+      "epoch": 0.3276212605286088,
+      "grad_norm": 0.4092549681663513,
+      "learning_rate": 0.0001,
+      "loss": 1.7844,
+      "step": 2820
+    },
+    {
+      "epoch": 0.3277374382805693,
+      "grad_norm": 0.3923521041870117,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 2821
+    },
+    {
+      "epoch": 0.32785361603252977,
+      "grad_norm": 0.3734949231147766,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 2822
+    },
+    {
+      "epoch": 0.32796979378449026,
+      "grad_norm": 0.3700851798057556,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 2823
+    },
+    {
+      "epoch": 0.32808597153645075,
+      "grad_norm": 0.3860493302345276,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 2824
+    },
+    {
+      "epoch": 0.3282021492884113,
+      "grad_norm": 0.4142455756664276,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 2825
+    },
+    {
+      "epoch": 0.3283183270403718,
+      "grad_norm": 0.3860050141811371,
+      "learning_rate": 0.0001,
+      "loss": 1.7274,
+      "step": 2826
+    },
+    {
+      "epoch": 0.32843450479233227,
+      "grad_norm": 0.39390870928764343,
+      "learning_rate": 0.0001,
+      "loss": 1.7529,
+      "step": 2827
+    },
+    {
+      "epoch": 0.32855068254429276,
+      "grad_norm": 0.3824588656425476,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 2828
+    },
+    {
+      "epoch": 0.32866686029625325,
+      "grad_norm": 0.3569580018520355,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 2829
+    },
+    {
+      "epoch": 0.3287830380482138,
+      "grad_norm": 0.3893423080444336,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 2830
+    },
+    {
+      "epoch": 0.3288992158001743,
+      "grad_norm": 0.3976670503616333,
+      "learning_rate": 0.0001,
+      "loss": 1.7884,
+      "step": 2831
+    },
+    {
+      "epoch": 0.32901539355213477,
+      "grad_norm": 0.38111555576324463,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 2832
+    },
+    {
+      "epoch": 0.32913157130409526,
+      "grad_norm": 0.38371211290359497,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 2833
+    },
+    {
+      "epoch": 0.32924774905605575,
+      "grad_norm": 0.4201413691043854,
+      "learning_rate": 0.0001,
+      "loss": 1.7543,
+      "step": 2834
+    },
+    {
+      "epoch": 0.32936392680801624,
+      "grad_norm": 0.386960506439209,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 2835
+    },
+    {
+      "epoch": 0.3294801045599768,
+      "grad_norm": 0.41088366508483887,
+      "learning_rate": 0.0001,
+      "loss": 1.9636,
+      "step": 2836
+    },
+    {
+      "epoch": 0.32959628231193727,
+      "grad_norm": 0.39831310510635376,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 2837
+    },
+    {
+      "epoch": 0.32971246006389776,
+      "grad_norm": 0.3723565638065338,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 2838
+    },
+    {
+      "epoch": 0.32982863781585825,
+      "grad_norm": 0.3974156975746155,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 2839
+    },
+    {
+      "epoch": 0.32994481556781874,
+      "grad_norm": 0.3748154640197754,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 2840
+    },
+    {
+      "epoch": 0.3300609933197793,
+      "grad_norm": 0.375637024641037,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 2841
+    },
+    {
+      "epoch": 0.3301771710717398,
+      "grad_norm": 0.397079199552536,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 2842
+    },
+    {
+      "epoch": 0.33029334882370026,
+      "grad_norm": 0.4001121520996094,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 2843
+    },
+    {
+      "epoch": 0.33040952657566075,
+      "grad_norm": 0.3832455277442932,
+      "learning_rate": 0.0001,
+      "loss": 1.7496,
+      "step": 2844
+    },
+    {
+      "epoch": 0.33052570432762124,
+      "grad_norm": 0.3800029456615448,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 2845
+    },
+    {
+      "epoch": 0.3306418820795818,
+      "grad_norm": 0.37468817830085754,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 2846
+    },
+    {
+      "epoch": 0.3307580598315423,
+      "grad_norm": 0.3698212802410126,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 2847
+    },
+    {
+      "epoch": 0.33087423758350276,
+      "grad_norm": 0.39800024032592773,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 2848
+    },
+    {
+      "epoch": 0.33099041533546325,
+      "grad_norm": 0.3913809657096863,
+      "learning_rate": 0.0001,
+      "loss": 1.7207,
+      "step": 2849
+    },
+    {
+      "epoch": 0.33110659308742374,
+      "grad_norm": 0.3854704201221466,
+      "learning_rate": 0.0001,
+      "loss": 1.696,
+      "step": 2850
+    },
+    {
+      "epoch": 0.3312227708393843,
+      "grad_norm": 0.36068427562713623,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 2851
+    },
+    {
+      "epoch": 0.3313389485913448,
+      "grad_norm": 0.4268893003463745,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 2852
+    },
+    {
+      "epoch": 0.33145512634330526,
+      "grad_norm": 0.3733718991279602,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 2853
+    },
+    {
+      "epoch": 0.33157130409526575,
+      "grad_norm": 0.36404022574424744,
+      "learning_rate": 0.0001,
+      "loss": 1.4211,
+      "step": 2854
+    },
+    {
+      "epoch": 0.33168748184722624,
+      "grad_norm": 0.4233626425266266,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 2855
+    },
+    {
+      "epoch": 0.33180365959918673,
+      "grad_norm": 0.48992058634757996,
+      "learning_rate": 0.0001,
+      "loss": 1.8077,
+      "step": 2856
+    },
+    {
+      "epoch": 0.3319198373511473,
+      "grad_norm": 0.4136912226676941,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 2857
+    },
+    {
+      "epoch": 0.33203601510310776,
+      "grad_norm": 0.36938685178756714,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 2858
+    },
+    {
+      "epoch": 0.33215219285506825,
+      "grad_norm": 0.36779820919036865,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 2859
+    },
+    {
+      "epoch": 0.33226837060702874,
+      "grad_norm": 0.36435508728027344,
+      "learning_rate": 0.0001,
+      "loss": 1.508,
+      "step": 2860
+    },
+    {
+      "epoch": 0.33238454835898923,
+      "grad_norm": 0.38518238067626953,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3325007261109498,
+      "grad_norm": 0.419127881526947,
+      "learning_rate": 0.0001,
+      "loss": 1.8029,
+      "step": 2862
+    },
+    {
+      "epoch": 0.33261690386291026,
+      "grad_norm": 0.3840404152870178,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 2863
+    },
+    {
+      "epoch": 0.33273308161487075,
+      "grad_norm": 0.4244234263896942,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 2864
+    },
+    {
+      "epoch": 0.33284925936683124,
+      "grad_norm": 0.40115001797676086,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 2865
+    },
+    {
+      "epoch": 0.33296543711879173,
+      "grad_norm": 0.4338228702545166,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3330816148707523,
+      "grad_norm": 0.40613770484924316,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 2867
+    },
+    {
+      "epoch": 0.33319779262271276,
+      "grad_norm": 0.4123576283454895,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 2868
+    },
+    {
+      "epoch": 0.33331397037467325,
+      "grad_norm": 0.3951095938682556,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 2869
+    },
+    {
+      "epoch": 0.33343014812663374,
+      "grad_norm": 0.410762220621109,
+      "learning_rate": 0.0001,
+      "loss": 1.82,
+      "step": 2870
+    },
+    {
+      "epoch": 0.33354632587859423,
+      "grad_norm": 0.4156797528266907,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 2871
+    },
+    {
+      "epoch": 0.3336625036305548,
+      "grad_norm": 0.3592885434627533,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 2872
+    },
+    {
+      "epoch": 0.33377868138251526,
+      "grad_norm": 0.42720362544059753,
+      "learning_rate": 0.0001,
+      "loss": 1.918,
+      "step": 2873
+    },
+    {
+      "epoch": 0.33389485913447575,
+      "grad_norm": 0.40637707710266113,
+      "learning_rate": 0.0001,
+      "loss": 1.8286,
+      "step": 2874
+    },
+    {
+      "epoch": 0.33401103688643624,
+      "grad_norm": 0.39554738998413086,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 2875
+    },
+    {
+      "epoch": 0.33412721463839673,
+      "grad_norm": 0.37411895394325256,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 2876
+    },
+    {
+      "epoch": 0.3342433923903572,
+      "grad_norm": 0.38688522577285767,
+      "learning_rate": 0.0001,
+      "loss": 1.5817,
+      "step": 2877
+    },
+    {
+      "epoch": 0.33435957014231776,
+      "grad_norm": 0.39456331729888916,
+      "learning_rate": 0.0001,
+      "loss": 1.7141,
+      "step": 2878
+    },
+    {
+      "epoch": 0.33447574789427825,
+      "grad_norm": 0.37233471870422363,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 2879
+    },
+    {
+      "epoch": 0.33459192564623874,
+      "grad_norm": 0.3917320668697357,
+      "learning_rate": 0.0001,
+      "loss": 1.7551,
+      "step": 2880
+    },
+    {
+      "epoch": 0.33470810339819923,
+      "grad_norm": 0.3961198627948761,
+      "learning_rate": 0.0001,
+      "loss": 1.7296,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3348242811501597,
+      "grad_norm": 0.4228503108024597,
+      "learning_rate": 0.0001,
+      "loss": 1.8795,
+      "step": 2882
+    },
+    {
+      "epoch": 0.33494045890212026,
+      "grad_norm": 0.3861311376094818,
+      "learning_rate": 0.0001,
+      "loss": 1.8777,
+      "step": 2883
+    },
+    {
+      "epoch": 0.33505663665408075,
+      "grad_norm": 0.3791095018386841,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 2884
+    },
+    {
+      "epoch": 0.33517281440604124,
+      "grad_norm": 0.45170632004737854,
+      "learning_rate": 0.0001,
+      "loss": 1.8281,
+      "step": 2885
+    },
+    {
+      "epoch": 0.33528899215800173,
+      "grad_norm": 0.4086534082889557,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 2886
+    },
+    {
+      "epoch": 0.3354051699099622,
+      "grad_norm": 0.4174976348876953,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 2887
+    },
+    {
+      "epoch": 0.33552134766192276,
+      "grad_norm": 0.39714473485946655,
+      "learning_rate": 0.0001,
+      "loss": 1.683,
+      "step": 2888
+    },
+    {
+      "epoch": 0.33563752541388325,
+      "grad_norm": 0.39770638942718506,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 2889
+    },
+    {
+      "epoch": 0.33575370316584374,
+      "grad_norm": 0.4018489122390747,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 2890
+    },
+    {
+      "epoch": 0.33586988091780423,
+      "grad_norm": 0.3446825444698334,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 2891
+    },
+    {
+      "epoch": 0.3359860586697647,
+      "grad_norm": 0.41117849946022034,
+      "learning_rate": 0.0001,
+      "loss": 1.7171,
+      "step": 2892
+    },
+    {
+      "epoch": 0.33610223642172526,
+      "grad_norm": 0.4037335216999054,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 2893
+    },
+    {
+      "epoch": 0.33621841417368575,
+      "grad_norm": 0.3712430000305176,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 2894
+    },
+    {
+      "epoch": 0.33633459192564624,
+      "grad_norm": 0.4003915786743164,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 2895
+    },
+    {
+      "epoch": 0.33645076967760673,
+      "grad_norm": 0.43774011731147766,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3365669474295672,
+      "grad_norm": 0.43321821093559265,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 2897
+    },
+    {
+      "epoch": 0.33668312518152776,
+      "grad_norm": 0.37499547004699707,
+      "learning_rate": 0.0001,
+      "loss": 1.5873,
+      "step": 2898
+    },
+    {
+      "epoch": 0.33679930293348825,
+      "grad_norm": 0.37291446328163147,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 2899
+    },
+    {
+      "epoch": 0.33691548068544874,
+      "grad_norm": 0.44143107533454895,
+      "learning_rate": 0.0001,
+      "loss": 1.8552,
+      "step": 2900
+    },
+    {
+      "epoch": 0.33703165843740923,
+      "grad_norm": 0.3805929720401764,
+      "learning_rate": 0.0001,
+      "loss": 1.7206,
+      "step": 2901
+    },
+    {
+      "epoch": 0.3371478361893697,
+      "grad_norm": 0.3873468339443207,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 2902
+    },
+    {
+      "epoch": 0.3372640139413302,
+      "grad_norm": 0.3733051121234894,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 2903
+    },
+    {
+      "epoch": 0.33738019169329075,
+      "grad_norm": 0.40816688537597656,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 2904
+    },
+    {
+      "epoch": 0.33749636944525124,
+      "grad_norm": 0.3908594250679016,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 2905
+    },
+    {
+      "epoch": 0.33761254719721173,
+      "grad_norm": 0.40407541394233704,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 2906
+    },
+    {
+      "epoch": 0.3377287249491722,
+      "grad_norm": 0.4066741168498993,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 2907
+    },
+    {
+      "epoch": 0.3378449027011327,
+      "grad_norm": 0.3964504599571228,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 2908
+    },
+    {
+      "epoch": 0.33796108045309325,
+      "grad_norm": 0.37795713543891907,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 2909
+    },
+    {
+      "epoch": 0.33807725820505374,
+      "grad_norm": 0.37210172414779663,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 2910
+    },
+    {
+      "epoch": 0.33819343595701423,
+      "grad_norm": 0.3974775969982147,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3383096137089747,
+      "grad_norm": 0.37600451707839966,
+      "learning_rate": 0.0001,
+      "loss": 1.8123,
+      "step": 2912
+    },
+    {
+      "epoch": 0.3384257914609352,
+      "grad_norm": 0.3853381276130676,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 2913
+    },
+    {
+      "epoch": 0.33854196921289575,
+      "grad_norm": 0.38297778367996216,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 2914
+    },
+    {
+      "epoch": 0.33865814696485624,
+      "grad_norm": 0.392566978931427,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 2915
+    },
+    {
+      "epoch": 0.33877432471681673,
+      "grad_norm": 0.3868968188762665,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 2916
+    },
+    {
+      "epoch": 0.3388905024687772,
+      "grad_norm": 0.39296168088912964,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 2917
+    },
+    {
+      "epoch": 0.3390066802207377,
+      "grad_norm": 0.3615785837173462,
+      "learning_rate": 0.0001,
+      "loss": 1.5509,
+      "step": 2918
+    },
+    {
+      "epoch": 0.33912285797269826,
+      "grad_norm": 0.3897397220134735,
+      "learning_rate": 0.0001,
+      "loss": 1.7675,
+      "step": 2919
+    },
+    {
+      "epoch": 0.33923903572465874,
+      "grad_norm": 0.3959611654281616,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 2920
+    },
+    {
+      "epoch": 0.33935521347661923,
+      "grad_norm": 0.39301490783691406,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 2921
+    },
+    {
+      "epoch": 0.3394713912285797,
+      "grad_norm": 0.3942652940750122,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3395875689805402,
+      "grad_norm": 0.43564826250076294,
+      "learning_rate": 0.0001,
+      "loss": 1.8724,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3397037467325007,
+      "grad_norm": 0.39441993832588196,
+      "learning_rate": 0.0001,
+      "loss": 1.7982,
+      "step": 2924
+    },
+    {
+      "epoch": 0.33981992448446124,
+      "grad_norm": 0.3842059373855591,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 2925
+    },
+    {
+      "epoch": 0.33993610223642173,
+      "grad_norm": 0.40640154480934143,
+      "learning_rate": 0.0001,
+      "loss": 1.8546,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3400522799883822,
+      "grad_norm": 0.41045522689819336,
+      "learning_rate": 0.0001,
+      "loss": 1.8838,
+      "step": 2927
+    },
+    {
+      "epoch": 0.3401684577403427,
+      "grad_norm": 0.37923598289489746,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 2928
+    },
+    {
+      "epoch": 0.3402846354923032,
+      "grad_norm": 0.3806171715259552,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 2929
+    },
+    {
+      "epoch": 0.34040081324426374,
+      "grad_norm": 0.3735258877277374,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 2930
+    },
+    {
+      "epoch": 0.34051699099622423,
+      "grad_norm": 0.3817983865737915,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 2931
+    },
+    {
+      "epoch": 0.3406331687481847,
+      "grad_norm": 0.4227299690246582,
+      "learning_rate": 0.0001,
+      "loss": 1.78,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3407493465001452,
+      "grad_norm": 0.4059399664402008,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 2933
+    },
+    {
+      "epoch": 0.3408655242521057,
+      "grad_norm": 0.3733903169631958,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 2934
+    },
+    {
+      "epoch": 0.34098170200406624,
+      "grad_norm": 0.3693414330482483,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 2935
+    },
+    {
+      "epoch": 0.34109787975602673,
+      "grad_norm": 0.3974681794643402,
+      "learning_rate": 0.0001,
+      "loss": 1.7623,
+      "step": 2936
+    },
+    {
+      "epoch": 0.3412140575079872,
+      "grad_norm": 0.40453028678894043,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3413302352599477,
+      "grad_norm": 0.3921975791454315,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 2938
+    },
+    {
+      "epoch": 0.3414464130119082,
+      "grad_norm": 0.39022400975227356,
+      "learning_rate": 0.0001,
+      "loss": 1.6862,
+      "step": 2939
+    },
+    {
+      "epoch": 0.34156259076386875,
+      "grad_norm": 0.39960476756095886,
+      "learning_rate": 0.0001,
+      "loss": 1.7099,
+      "step": 2940
+    },
+    {
+      "epoch": 0.34167876851582923,
+      "grad_norm": 0.40952742099761963,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 2941
+    },
+    {
+      "epoch": 0.3417949462677897,
+      "grad_norm": 0.3948822617530823,
+      "learning_rate": 0.0001,
+      "loss": 1.7191,
+      "step": 2942
+    },
+    {
+      "epoch": 0.3419111240197502,
+      "grad_norm": 0.4061760902404785,
+      "learning_rate": 0.0001,
+      "loss": 1.7105,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3420273017717107,
+      "grad_norm": 0.3945620357990265,
+      "learning_rate": 0.0001,
+      "loss": 1.7056,
+      "step": 2944
+    },
+    {
+      "epoch": 0.3421434795236712,
+      "grad_norm": 0.4100196659564972,
+      "learning_rate": 0.0001,
+      "loss": 1.7106,
+      "step": 2945
+    },
+    {
+      "epoch": 0.34225965727563173,
+      "grad_norm": 0.3933731019496918,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3423758350275922,
+      "grad_norm": 0.3884199559688568,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 2947
+    },
+    {
+      "epoch": 0.3424920127795527,
+      "grad_norm": 0.386350154876709,
+      "learning_rate": 0.0001,
+      "loss": 1.7033,
+      "step": 2948
+    },
+    {
+      "epoch": 0.3426081905315132,
+      "grad_norm": 0.40476351976394653,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3427243682834737,
+      "grad_norm": 0.407321572303772,
+      "learning_rate": 0.0001,
+      "loss": 1.7674,
+      "step": 2950
+    },
+    {
+      "epoch": 0.34284054603543423,
+      "grad_norm": 0.40488916635513306,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 2951
+    },
+    {
+      "epoch": 0.3429567237873947,
+      "grad_norm": 0.4045671224594116,
+      "learning_rate": 0.0001,
+      "loss": 1.8208,
+      "step": 2952
+    },
+    {
+      "epoch": 0.3430729015393552,
+      "grad_norm": 0.4002356231212616,
+      "learning_rate": 0.0001,
+      "loss": 1.6873,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3431890792913157,
+      "grad_norm": 0.3586932122707367,
+      "learning_rate": 0.0001,
+      "loss": 1.4746,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3433052570432762,
+      "grad_norm": 0.370332807302475,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 2955
+    },
+    {
+      "epoch": 0.34342143479523674,
+      "grad_norm": 0.39355921745300293,
+      "learning_rate": 0.0001,
+      "loss": 1.8061,
+      "step": 2956
+    },
+    {
+      "epoch": 0.3435376125471972,
+      "grad_norm": 0.39952757954597473,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 2957
+    },
+    {
+      "epoch": 0.3436537902991577,
+      "grad_norm": 0.3770568370819092,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 2958
+    },
+    {
+      "epoch": 0.3437699680511182,
+      "grad_norm": 0.4441458582878113,
+      "learning_rate": 0.0001,
+      "loss": 1.9858,
+      "step": 2959
+    },
+    {
+      "epoch": 0.3438861458030787,
+      "grad_norm": 0.38884633779525757,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 2960
+    },
+    {
+      "epoch": 0.34400232355503924,
+      "grad_norm": 0.41551780700683594,
+      "learning_rate": 0.0001,
+      "loss": 1.8076,
+      "step": 2961
+    },
+    {
+      "epoch": 0.3441185013069997,
+      "grad_norm": 0.3980303108692169,
+      "learning_rate": 0.0001,
+      "loss": 1.613,
+      "step": 2962
+    },
+    {
+      "epoch": 0.3442346790589602,
+      "grad_norm": 0.3553750813007355,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 2963
+    },
+    {
+      "epoch": 0.3443508568109207,
+      "grad_norm": 0.38143518567085266,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 2964
+    },
+    {
+      "epoch": 0.3444670345628812,
+      "grad_norm": 0.40971437096595764,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3445832123148417,
+      "grad_norm": 0.35537075996398926,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 2966
+    },
+    {
+      "epoch": 0.3446993900668022,
+      "grad_norm": 0.38093459606170654,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 2967
+    },
+    {
+      "epoch": 0.3448155678187627,
+      "grad_norm": 0.34431418776512146,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 2968
+    },
+    {
+      "epoch": 0.3449317455707232,
+      "grad_norm": 0.427884042263031,
+      "learning_rate": 0.0001,
+      "loss": 1.7658,
+      "step": 2969
+    },
+    {
+      "epoch": 0.3450479233226837,
+      "grad_norm": 0.4321180582046509,
+      "learning_rate": 0.0001,
+      "loss": 1.8182,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3451641010746442,
+      "grad_norm": 0.3925512731075287,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 2971
+    },
+    {
+      "epoch": 0.3452802788266047,
+      "grad_norm": 0.3812701106071472,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3453964565785652,
+      "grad_norm": 0.4410078227519989,
+      "learning_rate": 0.0001,
+      "loss": 1.8109,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3455126343305257,
+      "grad_norm": 0.38449546694755554,
+      "learning_rate": 0.0001,
+      "loss": 1.7467,
+      "step": 2974
+    },
+    {
+      "epoch": 0.3456288120824862,
+      "grad_norm": 0.3961304724216461,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 2975
+    },
+    {
+      "epoch": 0.3457449898344467,
+      "grad_norm": 0.3921011686325073,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3458611675864072,
+      "grad_norm": 0.39690279960632324,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3459773453383677,
+      "grad_norm": 0.4186713397502899,
+      "learning_rate": 0.0001,
+      "loss": 1.7538,
+      "step": 2978
+    },
+    {
+      "epoch": 0.3460935230903282,
+      "grad_norm": 0.39231374859809875,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 2979
+    },
+    {
+      "epoch": 0.3462097008422887,
+      "grad_norm": 0.3903275430202484,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 2980
+    },
+    {
+      "epoch": 0.3463258785942492,
+      "grad_norm": 0.4018222689628601,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3464420563462097,
+      "grad_norm": 0.41252797842025757,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3465582340981702,
+      "grad_norm": 0.38088223338127136,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 2983
+    },
+    {
+      "epoch": 0.3466744118501307,
+      "grad_norm": 0.3844488859176636,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 2984
+    },
+    {
+      "epoch": 0.3467905896020912,
+      "grad_norm": 0.3818972408771515,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3469067673540517,
+      "grad_norm": 0.4173097014427185,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 2986
+    },
+    {
+      "epoch": 0.3470229451060122,
+      "grad_norm": 0.428475946187973,
+      "learning_rate": 0.0001,
+      "loss": 1.8248,
+      "step": 2987
+    },
+    {
+      "epoch": 0.3471391228579727,
+      "grad_norm": 0.40576884150505066,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3472553006099332,
+      "grad_norm": 0.4049152433872223,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 2989
+    },
+    {
+      "epoch": 0.3473714783618937,
+      "grad_norm": 0.3948923647403717,
+      "learning_rate": 0.0001,
+      "loss": 1.6983,
+      "step": 2990
+    },
+    {
+      "epoch": 0.3474876561138542,
+      "grad_norm": 0.3774530291557312,
+      "learning_rate": 0.0001,
+      "loss": 1.7484,
+      "step": 2991
+    },
+    {
+      "epoch": 0.34760383386581467,
+      "grad_norm": 0.38777539134025574,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3477200116177752,
+      "grad_norm": 0.38459205627441406,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 2993
+    },
+    {
+      "epoch": 0.3478361893697357,
+      "grad_norm": 0.44415146112442017,
+      "learning_rate": 0.0001,
+      "loss": 1.8133,
+      "step": 2994
+    },
+    {
+      "epoch": 0.3479523671216962,
+      "grad_norm": 0.3697360157966614,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 2995
+    },
+    {
+      "epoch": 0.3480685448736567,
+      "grad_norm": 0.4112284779548645,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 2996
+    },
+    {
+      "epoch": 0.34818472262561717,
+      "grad_norm": 0.42278242111206055,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3483009003775777,
+      "grad_norm": 0.3720592260360718,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 2998
+    },
+    {
+      "epoch": 0.3484170781295382,
+      "grad_norm": 0.37524881958961487,
+      "learning_rate": 0.0001,
+      "loss": 1.6982,
+      "step": 2999
+    },
+    {
+      "epoch": 0.3485332558814987,
+      "grad_norm": 0.38815271854400635,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3486494336334592,
+      "grad_norm": 0.4013916552066803,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 3001
+    },
+    {
+      "epoch": 0.34876561138541967,
+      "grad_norm": 0.3819652497768402,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 3002
+    },
+    {
+      "epoch": 0.3488817891373802,
+      "grad_norm": 0.37497755885124207,
+      "learning_rate": 0.0001,
+      "loss": 1.5937,
+      "step": 3003
+    },
+    {
+      "epoch": 0.3489979668893407,
+      "grad_norm": 0.3700786232948303,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 3004
+    },
+    {
+      "epoch": 0.3491141446413012,
+      "grad_norm": 0.40467193722724915,
+      "learning_rate": 0.0001,
+      "loss": 1.7022,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3492303223932617,
+      "grad_norm": 0.3688233494758606,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 3006
+    },
+    {
+      "epoch": 0.34934650014522217,
+      "grad_norm": 0.39277154207229614,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 3007
+    },
+    {
+      "epoch": 0.3494626778971827,
+      "grad_norm": 0.3854866921901703,
+      "learning_rate": 0.0001,
+      "loss": 1.66,
+      "step": 3008
+    },
+    {
+      "epoch": 0.3495788556491432,
+      "grad_norm": 0.4092784821987152,
+      "learning_rate": 0.0001,
+      "loss": 1.7302,
+      "step": 3009
+    },
+    {
+      "epoch": 0.3496950334011037,
+      "grad_norm": 0.3558938205242157,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 3010
+    },
+    {
+      "epoch": 0.3498112111530642,
+      "grad_norm": 0.4179665148258209,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 3011
+    },
+    {
+      "epoch": 0.34992738890502467,
+      "grad_norm": 0.38339027762413025,
+      "learning_rate": 0.0001,
+      "loss": 1.4907,
+      "step": 3012
+    },
+    {
+      "epoch": 0.35004356665698516,
+      "grad_norm": 0.36649930477142334,
+      "learning_rate": 0.0001,
+      "loss": 1.6285,
+      "step": 3013
+    },
+    {
+      "epoch": 0.3501597444089457,
+      "grad_norm": 0.4202296733856201,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 3014
+    },
+    {
+      "epoch": 0.3502759221609062,
+      "grad_norm": 0.4036094844341278,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3503920999128667,
+      "grad_norm": 0.3799827992916107,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 3016
+    },
+    {
+      "epoch": 0.3505082776648272,
+      "grad_norm": 0.40348902344703674,
+      "learning_rate": 0.0001,
+      "loss": 1.7164,
+      "step": 3017
+    },
+    {
+      "epoch": 0.35062445541678766,
+      "grad_norm": 0.38102301955223083,
+      "learning_rate": 0.0001,
+      "loss": 1.7244,
+      "step": 3018
+    },
+    {
+      "epoch": 0.3507406331687482,
+      "grad_norm": 0.3788668215274811,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 3019
+    },
+    {
+      "epoch": 0.3508568109207087,
+      "grad_norm": 0.38864848017692566,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 3020
+    },
+    {
+      "epoch": 0.3509729886726692,
+      "grad_norm": 0.37422239780426025,
+      "learning_rate": 0.0001,
+      "loss": 1.7779,
+      "step": 3021
+    },
+    {
+      "epoch": 0.3510891664246297,
+      "grad_norm": 0.37829017639160156,
+      "learning_rate": 0.0001,
+      "loss": 1.7872,
+      "step": 3022
+    },
+    {
+      "epoch": 0.35120534417659016,
+      "grad_norm": 0.39269503951072693,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 3023
+    },
+    {
+      "epoch": 0.3513215219285507,
+      "grad_norm": 0.381878137588501,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 3024
+    },
+    {
+      "epoch": 0.3514376996805112,
+      "grad_norm": 0.39613422751426697,
+      "learning_rate": 0.0001,
+      "loss": 1.8325,
+      "step": 3025
+    },
+    {
+      "epoch": 0.3515538774324717,
+      "grad_norm": 0.42367449402809143,
+      "learning_rate": 0.0001,
+      "loss": 1.7125,
+      "step": 3026
+    },
+    {
+      "epoch": 0.3516700551844322,
+      "grad_norm": 0.3661247789859772,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 3027
+    },
+    {
+      "epoch": 0.35178623293639266,
+      "grad_norm": 0.36878344416618347,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 3028
+    },
+    {
+      "epoch": 0.3519024106883532,
+      "grad_norm": 0.38732531666755676,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 3029
+    },
+    {
+      "epoch": 0.3520185884403137,
+      "grad_norm": 0.4050348103046417,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3521347661922742,
+      "grad_norm": 0.40039730072021484,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 3031
+    },
+    {
+      "epoch": 0.3522509439442347,
+      "grad_norm": 0.40154829621315,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 3032
+    },
+    {
+      "epoch": 0.35236712169619516,
+      "grad_norm": 0.37400686740875244,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 3033
+    },
+    {
+      "epoch": 0.35248329944815565,
+      "grad_norm": 0.3930714726448059,
+      "learning_rate": 0.0001,
+      "loss": 1.874,
+      "step": 3034
+    },
+    {
+      "epoch": 0.3525994772001162,
+      "grad_norm": 0.4327416718006134,
+      "learning_rate": 0.0001,
+      "loss": 1.8191,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3527156549520767,
+      "grad_norm": 0.4135274887084961,
+      "learning_rate": 0.0001,
+      "loss": 1.9436,
+      "step": 3036
+    },
+    {
+      "epoch": 0.3528318327040372,
+      "grad_norm": 0.36766317486763,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 3037
+    },
+    {
+      "epoch": 0.35294801045599766,
+      "grad_norm": 0.3964691460132599,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 3038
+    },
+    {
+      "epoch": 0.35306418820795815,
+      "grad_norm": 0.4026515781879425,
+      "learning_rate": 0.0001,
+      "loss": 1.8446,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3531803659599187,
+      "grad_norm": 0.38630566000938416,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 3040
+    },
+    {
+      "epoch": 0.3532965437118792,
+      "grad_norm": 0.38038820028305054,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 3041
+    },
+    {
+      "epoch": 0.3534127214638397,
+      "grad_norm": 0.379242867231369,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 3042
+    },
+    {
+      "epoch": 0.35352889921580016,
+      "grad_norm": 0.3936004340648651,
+      "learning_rate": 0.0001,
+      "loss": 1.8992,
+      "step": 3043
+    },
+    {
+      "epoch": 0.35364507696776065,
+      "grad_norm": 0.3659883737564087,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 3044
+    },
+    {
+      "epoch": 0.3537612547197212,
+      "grad_norm": 0.4306775629520416,
+      "learning_rate": 0.0001,
+      "loss": 1.8698,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3538774324716817,
+      "grad_norm": 0.40540704131126404,
+      "learning_rate": 0.0001,
+      "loss": 1.7321,
+      "step": 3046
+    },
+    {
+      "epoch": 0.3539936102236422,
+      "grad_norm": 0.3898630440235138,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 3047
+    },
+    {
+      "epoch": 0.35410978797560266,
+      "grad_norm": 0.40471145510673523,
+      "learning_rate": 0.0001,
+      "loss": 1.7801,
+      "step": 3048
+    },
+    {
+      "epoch": 0.35422596572756315,
+      "grad_norm": 0.40043899416923523,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 3049
+    },
+    {
+      "epoch": 0.3543421434795237,
+      "grad_norm": 0.3994670808315277,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3544583212314842,
+      "grad_norm": 0.4123898446559906,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 3051
+    },
+    {
+      "epoch": 0.3545744989834447,
+      "grad_norm": 0.3823109567165375,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 3052
+    },
+    {
+      "epoch": 0.35469067673540516,
+      "grad_norm": 0.41799673438072205,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 3053
+    },
+    {
+      "epoch": 0.35480685448736565,
+      "grad_norm": 0.36214837431907654,
+      "learning_rate": 0.0001,
+      "loss": 1.5653,
+      "step": 3054
+    },
+    {
+      "epoch": 0.35492303223932614,
+      "grad_norm": 0.4229331910610199,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 3055
+    },
+    {
+      "epoch": 0.3550392099912867,
+      "grad_norm": 0.38769447803497314,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 3056
+    },
+    {
+      "epoch": 0.3551553877432472,
+      "grad_norm": 0.39255595207214355,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 3057
+    },
+    {
+      "epoch": 0.35527156549520766,
+      "grad_norm": 0.4204464852809906,
+      "learning_rate": 0.0001,
+      "loss": 1.7607,
+      "step": 3058
+    },
+    {
+      "epoch": 0.35538774324716815,
+      "grad_norm": 0.3666527271270752,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 3059
+    },
+    {
+      "epoch": 0.35550392099912864,
+      "grad_norm": 0.384753942489624,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3556200987510892,
+      "grad_norm": 0.3891599178314209,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 3061
+    },
+    {
+      "epoch": 0.3557362765030497,
+      "grad_norm": 0.37767064571380615,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 3062
+    },
+    {
+      "epoch": 0.35585245425501016,
+      "grad_norm": 0.40282127261161804,
+      "learning_rate": 0.0001,
+      "loss": 1.8185,
+      "step": 3063
+    },
+    {
+      "epoch": 0.35596863200697065,
+      "grad_norm": 0.3778342008590698,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 3064
+    },
+    {
+      "epoch": 0.35608480975893114,
+      "grad_norm": 0.37607836723327637,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3562009875108917,
+      "grad_norm": 0.3772094249725342,
+      "learning_rate": 0.0001,
+      "loss": 1.7671,
+      "step": 3066
+    },
+    {
+      "epoch": 0.3563171652628522,
+      "grad_norm": 0.4059067666530609,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 3067
+    },
+    {
+      "epoch": 0.35643334301481266,
+      "grad_norm": 0.39582398533821106,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 3068
+    },
+    {
+      "epoch": 0.35654952076677315,
+      "grad_norm": 0.3916482627391815,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 3069
+    },
+    {
+      "epoch": 0.35666569851873364,
+      "grad_norm": 0.41407474875450134,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 3070
+    },
+    {
+      "epoch": 0.3567818762706942,
+      "grad_norm": 0.41029298305511475,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 3071
+    },
+    {
+      "epoch": 0.3568980540226547,
+      "grad_norm": 0.4083024859428406,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 3072
+    },
+    {
+      "epoch": 0.35701423177461517,
+      "grad_norm": 0.44555673003196716,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 3073
+    },
+    {
+      "epoch": 0.35713040952657565,
+      "grad_norm": 0.37947797775268555,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 3074
+    },
+    {
+      "epoch": 0.35724658727853614,
+      "grad_norm": 0.3823384940624237,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 3075
+    },
+    {
+      "epoch": 0.3573627650304967,
+      "grad_norm": 0.3980938494205475,
+      "learning_rate": 0.0001,
+      "loss": 1.7366,
+      "step": 3076
+    },
+    {
+      "epoch": 0.3574789427824572,
+      "grad_norm": 0.39685919880867004,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 3077
+    },
+    {
+      "epoch": 0.35759512053441767,
+      "grad_norm": 0.4064877927303314,
+      "learning_rate": 0.0001,
+      "loss": 1.7869,
+      "step": 3078
+    },
+    {
+      "epoch": 0.35771129828637815,
+      "grad_norm": 0.41149288415908813,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 3079
+    },
+    {
+      "epoch": 0.35782747603833864,
+      "grad_norm": 0.36738306283950806,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 3080
+    },
+    {
+      "epoch": 0.35794365379029913,
+      "grad_norm": 0.393264502286911,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 3081
+    },
+    {
+      "epoch": 0.3580598315422597,
+      "grad_norm": 0.3866167366504669,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 3082
+    },
+    {
+      "epoch": 0.35817600929422017,
+      "grad_norm": 0.38712388277053833,
+      "learning_rate": 0.0001,
+      "loss": 1.6682,
+      "step": 3083
+    },
+    {
+      "epoch": 0.35829218704618065,
+      "grad_norm": 0.401869535446167,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 3084
+    },
+    {
+      "epoch": 0.35840836479814114,
+      "grad_norm": 0.3952792286872864,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 3085
+    },
+    {
+      "epoch": 0.35852454255010163,
+      "grad_norm": 0.36368271708488464,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 3086
+    },
+    {
+      "epoch": 0.3586407203020622,
+      "grad_norm": 0.39687228202819824,
+      "learning_rate": 0.0001,
+      "loss": 1.6623,
+      "step": 3087
+    },
+    {
+      "epoch": 0.35875689805402267,
+      "grad_norm": 0.41528600454330444,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 3088
+    },
+    {
+      "epoch": 0.35887307580598315,
+      "grad_norm": 0.420449435710907,
+      "learning_rate": 0.0001,
+      "loss": 1.7244,
+      "step": 3089
+    },
+    {
+      "epoch": 0.35898925355794364,
+      "grad_norm": 0.43798938393592834,
+      "learning_rate": 0.0001,
+      "loss": 1.8541,
+      "step": 3090
+    },
+    {
+      "epoch": 0.35910543130990413,
+      "grad_norm": 0.37088409066200256,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 3091
+    },
+    {
+      "epoch": 0.3592216090618647,
+      "grad_norm": 0.41478070616722107,
+      "learning_rate": 0.0001,
+      "loss": 1.7114,
+      "step": 3092
+    },
+    {
+      "epoch": 0.35933778681382517,
+      "grad_norm": 0.4017499089241028,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 3093
+    },
+    {
+      "epoch": 0.35945396456578566,
+      "grad_norm": 0.4266115128993988,
+      "learning_rate": 0.0001,
+      "loss": 1.7716,
+      "step": 3094
+    },
+    {
+      "epoch": 0.35957014231774614,
+      "grad_norm": 0.39603403210639954,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 3095
+    },
+    {
+      "epoch": 0.35968632006970663,
+      "grad_norm": 0.4058244824409485,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3598024978216672,
+      "grad_norm": 0.3646165728569031,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 3097
+    },
+    {
+      "epoch": 0.35991867557362767,
+      "grad_norm": 0.3995044529438019,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 3098
+    },
+    {
+      "epoch": 0.36003485332558816,
+      "grad_norm": 0.3803756535053253,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 3099
+    },
+    {
+      "epoch": 0.36015103107754864,
+      "grad_norm": 0.41200825572013855,
+      "learning_rate": 0.0001,
+      "loss": 1.7208,
+      "step": 3100
+    },
+    {
+      "epoch": 0.36026720882950913,
+      "grad_norm": 0.394388347864151,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 3101
+    },
+    {
+      "epoch": 0.3603833865814696,
+      "grad_norm": 0.3949473798274994,
+      "learning_rate": 0.0001,
+      "loss": 1.8397,
+      "step": 3102
+    },
+    {
+      "epoch": 0.36049956433343017,
+      "grad_norm": 0.41774412989616394,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 3103
+    },
+    {
+      "epoch": 0.36061574208539066,
+      "grad_norm": 0.3581482470035553,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 3104
+    },
+    {
+      "epoch": 0.36073191983735114,
+      "grad_norm": 0.41178280115127563,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 3105
+    },
+    {
+      "epoch": 0.36084809758931163,
+      "grad_norm": 0.4004935920238495,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 3106
+    },
+    {
+      "epoch": 0.3609642753412721,
+      "grad_norm": 0.403432697057724,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 3107
+    },
+    {
+      "epoch": 0.36108045309323267,
+      "grad_norm": 0.3860589861869812,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 3108
+    },
+    {
+      "epoch": 0.36119663084519316,
+      "grad_norm": 0.37853455543518066,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 3109
+    },
+    {
+      "epoch": 0.36131280859715365,
+      "grad_norm": 0.40292003750801086,
+      "learning_rate": 0.0001,
+      "loss": 1.8344,
+      "step": 3110
+    },
+    {
+      "epoch": 0.36142898634911413,
+      "grad_norm": 0.3984355926513672,
+      "learning_rate": 0.0001,
+      "loss": 1.879,
+      "step": 3111
+    },
+    {
+      "epoch": 0.3615451641010746,
+      "grad_norm": 0.44067731499671936,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 3112
+    },
+    {
+      "epoch": 0.36166134185303517,
+      "grad_norm": 0.3945563733577728,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 3113
+    },
+    {
+      "epoch": 0.36177751960499566,
+      "grad_norm": 0.4175066649913788,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 3114
+    },
+    {
+      "epoch": 0.36189369735695615,
+      "grad_norm": 0.41401904821395874,
+      "learning_rate": 0.0001,
+      "loss": 1.7809,
+      "step": 3115
+    },
+    {
+      "epoch": 0.36200987510891663,
+      "grad_norm": 0.3950018882751465,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3621260528608771,
+      "grad_norm": 0.3951791524887085,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 3117
+    },
+    {
+      "epoch": 0.36224223061283767,
+      "grad_norm": 0.42133089900016785,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 3118
+    },
+    {
+      "epoch": 0.36235840836479816,
+      "grad_norm": 0.423442542552948,
+      "learning_rate": 0.0001,
+      "loss": 1.9073,
+      "step": 3119
+    },
+    {
+      "epoch": 0.36247458611675865,
+      "grad_norm": 0.4069424271583557,
+      "learning_rate": 0.0001,
+      "loss": 1.8604,
+      "step": 3120
+    },
+    {
+      "epoch": 0.36259076386871913,
+      "grad_norm": 0.4289453625679016,
+      "learning_rate": 0.0001,
+      "loss": 1.7824,
+      "step": 3121
+    },
+    {
+      "epoch": 0.3627069416206796,
+      "grad_norm": 0.36343902349472046,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 3122
+    },
+    {
+      "epoch": 0.3628231193726401,
+      "grad_norm": 0.45105743408203125,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 3123
+    },
+    {
+      "epoch": 0.36293929712460066,
+      "grad_norm": 0.3925178050994873,
+      "learning_rate": 0.0001,
+      "loss": 1.6771,
+      "step": 3124
+    },
+    {
+      "epoch": 0.36305547487656115,
+      "grad_norm": 0.4135838449001312,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 3125
+    },
+    {
+      "epoch": 0.36317165262852164,
+      "grad_norm": 0.36576542258262634,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 3126
+    },
+    {
+      "epoch": 0.3632878303804821,
+      "grad_norm": 0.4198266565799713,
+      "learning_rate": 0.0001,
+      "loss": 1.8513,
+      "step": 3127
+    },
+    {
+      "epoch": 0.3634040081324426,
+      "grad_norm": 0.3670734763145447,
+      "learning_rate": 0.0001,
+      "loss": 1.6557,
+      "step": 3128
+    },
+    {
+      "epoch": 0.36352018588440316,
+      "grad_norm": 0.36761337518692017,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 3129
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.40638405084609985,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 3130
+    },
+    {
+      "epoch": 0.36375254138832414,
+      "grad_norm": 0.38102391362190247,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 3131
+    },
+    {
+      "epoch": 0.3638687191402846,
+      "grad_norm": 0.39593052864074707,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 3132
+    },
+    {
+      "epoch": 0.3639848968922451,
+      "grad_norm": 0.4067305624485016,
+      "learning_rate": 0.0001,
+      "loss": 1.7925,
+      "step": 3133
+    },
+    {
+      "epoch": 0.36410107464420566,
+      "grad_norm": 0.3692638576030731,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 3134
+    },
+    {
+      "epoch": 0.36421725239616615,
+      "grad_norm": 0.40837207436561584,
+      "learning_rate": 0.0001,
+      "loss": 1.8112,
+      "step": 3135
+    },
+    {
+      "epoch": 0.36433343014812664,
+      "grad_norm": 0.41324061155319214,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 3136
+    },
+    {
+      "epoch": 0.3644496079000871,
+      "grad_norm": 0.35416892170906067,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 3137
+    },
+    {
+      "epoch": 0.3645657856520476,
+      "grad_norm": 0.44099685549736023,
+      "learning_rate": 0.0001,
+      "loss": 1.918,
+      "step": 3138
+    },
+    {
+      "epoch": 0.36468196340400816,
+      "grad_norm": 0.3850747346878052,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 3139
+    },
+    {
+      "epoch": 0.36479814115596865,
+      "grad_norm": 0.4012189209461212,
+      "learning_rate": 0.0001,
+      "loss": 1.7401,
+      "step": 3140
+    },
+    {
+      "epoch": 0.36491431890792914,
+      "grad_norm": 0.41232171654701233,
+      "learning_rate": 0.0001,
+      "loss": 1.7973,
+      "step": 3141
+    },
+    {
+      "epoch": 0.3650304966598896,
+      "grad_norm": 0.37721219658851624,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 3142
+    },
+    {
+      "epoch": 0.3651466744118501,
+      "grad_norm": 0.3969804346561432,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 3143
+    },
+    {
+      "epoch": 0.3652628521638106,
+      "grad_norm": 0.4444282054901123,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 3144
+    },
+    {
+      "epoch": 0.36537902991577115,
+      "grad_norm": 0.4145006239414215,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 3145
+    },
+    {
+      "epoch": 0.36549520766773164,
+      "grad_norm": 0.3919104337692261,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 3146
+    },
+    {
+      "epoch": 0.3656113854196921,
+      "grad_norm": 0.4502047002315521,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 3147
+    },
+    {
+      "epoch": 0.3657275631716526,
+      "grad_norm": 0.3652731776237488,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 3148
+    },
+    {
+      "epoch": 0.3658437409236131,
+      "grad_norm": 0.37962082028388977,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 3149
+    },
+    {
+      "epoch": 0.36595991867557365,
+      "grad_norm": 0.39860934019088745,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 3150
+    },
+    {
+      "epoch": 0.36607609642753414,
+      "grad_norm": 0.42549842596054077,
+      "learning_rate": 0.0001,
+      "loss": 1.7802,
+      "step": 3151
+    },
+    {
+      "epoch": 0.3661922741794946,
+      "grad_norm": 0.37746211886405945,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 3152
+    },
+    {
+      "epoch": 0.3663084519314551,
+      "grad_norm": 0.3802013397216797,
+      "learning_rate": 0.0001,
+      "loss": 1.712,
+      "step": 3153
+    },
+    {
+      "epoch": 0.3664246296834156,
+      "grad_norm": 0.426530659198761,
+      "learning_rate": 0.0001,
+      "loss": 1.8369,
+      "step": 3154
+    },
+    {
+      "epoch": 0.36654080743537615,
+      "grad_norm": 0.4312988817691803,
+      "learning_rate": 0.0001,
+      "loss": 1.8648,
+      "step": 3155
+    },
+    {
+      "epoch": 0.36665698518733664,
+      "grad_norm": 0.41918662190437317,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 3156
+    },
+    {
+      "epoch": 0.3667731629392971,
+      "grad_norm": 0.3956180810928345,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 3157
+    },
+    {
+      "epoch": 0.3668893406912576,
+      "grad_norm": 0.3965786397457123,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 3158
+    },
+    {
+      "epoch": 0.3670055184432181,
+      "grad_norm": 0.38480865955352783,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 3159
+    },
+    {
+      "epoch": 0.36712169619517865,
+      "grad_norm": 0.3683672249317169,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 3160
+    },
+    {
+      "epoch": 0.36723787394713914,
+      "grad_norm": 0.3930373191833496,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 3161
+    },
+    {
+      "epoch": 0.3673540516990996,
+      "grad_norm": 0.4485227167606354,
+      "learning_rate": 0.0001,
+      "loss": 1.8452,
+      "step": 3162
+    },
+    {
+      "epoch": 0.3674702294510601,
+      "grad_norm": 0.3868573009967804,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3675864072030206,
+      "grad_norm": 0.35300782322883606,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 3164
+    },
+    {
+      "epoch": 0.36770258495498115,
+      "grad_norm": 0.38356491923332214,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 3165
+    },
+    {
+      "epoch": 0.36781876270694164,
+      "grad_norm": 0.4126032292842865,
+      "learning_rate": 0.0001,
+      "loss": 1.7307,
+      "step": 3166
+    },
+    {
+      "epoch": 0.3679349404589021,
+      "grad_norm": 0.3850405216217041,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 3167
+    },
+    {
+      "epoch": 0.3680511182108626,
+      "grad_norm": 0.40334662795066833,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 3168
+    },
+    {
+      "epoch": 0.3681672959628231,
+      "grad_norm": 0.3768133223056793,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 3169
+    },
+    {
+      "epoch": 0.3682834737147836,
+      "grad_norm": 0.43316081166267395,
+      "learning_rate": 0.0001,
+      "loss": 1.8014,
+      "step": 3170
+    },
+    {
+      "epoch": 0.36839965146674414,
+      "grad_norm": 0.39245787262916565,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 3171
+    },
+    {
+      "epoch": 0.3685158292187046,
+      "grad_norm": 0.42442598938941956,
+      "learning_rate": 0.0001,
+      "loss": 1.8705,
+      "step": 3172
+    },
+    {
+      "epoch": 0.3686320069706651,
+      "grad_norm": 0.384989857673645,
+      "learning_rate": 0.0001,
+      "loss": 1.8006,
+      "step": 3173
+    },
+    {
+      "epoch": 0.3687481847226256,
+      "grad_norm": 0.3639425039291382,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 3174
+    },
+    {
+      "epoch": 0.3688643624745861,
+      "grad_norm": 0.4113941490650177,
+      "learning_rate": 0.0001,
+      "loss": 1.7056,
+      "step": 3175
+    },
+    {
+      "epoch": 0.36898054022654664,
+      "grad_norm": 0.39703720808029175,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 3176
+    },
+    {
+      "epoch": 0.3690967179785071,
+      "grad_norm": 0.3902719020843506,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 3177
+    },
+    {
+      "epoch": 0.3692128957304676,
+      "grad_norm": 0.3856205344200134,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3693290734824281,
+      "grad_norm": 0.4505693018436432,
+      "learning_rate": 0.0001,
+      "loss": 1.7766,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3694452512343886,
+      "grad_norm": 0.4288894534111023,
+      "learning_rate": 0.0001,
+      "loss": 1.7079,
+      "step": 3180
+    },
+    {
+      "epoch": 0.36956142898634914,
+      "grad_norm": 0.39302563667297363,
+      "learning_rate": 0.0001,
+      "loss": 1.6753,
+      "step": 3181
+    },
+    {
+      "epoch": 0.3696776067383096,
+      "grad_norm": 0.3827257454395294,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 3182
+    },
+    {
+      "epoch": 0.3697937844902701,
+      "grad_norm": 0.38766050338745117,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 3183
+    },
+    {
+      "epoch": 0.3699099622422306,
+      "grad_norm": 0.38151904940605164,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3700261399941911,
+      "grad_norm": 0.3981049656867981,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 3185
+    },
+    {
+      "epoch": 0.37014231774615164,
+      "grad_norm": 0.38091933727264404,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 3186
+    },
+    {
+      "epoch": 0.37025849549811213,
+      "grad_norm": 0.34910398721694946,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 3187
+    },
+    {
+      "epoch": 0.3703746732500726,
+      "grad_norm": 0.3882802128791809,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 3188
+    },
+    {
+      "epoch": 0.3704908510020331,
+      "grad_norm": 0.40641340613365173,
+      "learning_rate": 0.0001,
+      "loss": 1.8477,
+      "step": 3189
+    },
+    {
+      "epoch": 0.3706070287539936,
+      "grad_norm": 0.3742848038673401,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 3190
+    },
+    {
+      "epoch": 0.3707232065059541,
+      "grad_norm": 0.39867183566093445,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 3191
+    },
+    {
+      "epoch": 0.37083938425791463,
+      "grad_norm": 0.38178515434265137,
+      "learning_rate": 0.0001,
+      "loss": 1.57,
+      "step": 3192
+    },
+    {
+      "epoch": 0.3709555620098751,
+      "grad_norm": 0.38367438316345215,
+      "learning_rate": 0.0001,
+      "loss": 1.7399,
+      "step": 3193
+    },
+    {
+      "epoch": 0.3710717397618356,
+      "grad_norm": 0.4279497563838959,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 3194
+    },
+    {
+      "epoch": 0.3711879175137961,
+      "grad_norm": 0.4051482379436493,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 3195
+    },
+    {
+      "epoch": 0.3713040952657566,
+      "grad_norm": 0.4077185392379761,
+      "learning_rate": 0.0001,
+      "loss": 1.7127,
+      "step": 3196
+    },
+    {
+      "epoch": 0.37142027301771713,
+      "grad_norm": 0.4203553795814514,
+      "learning_rate": 0.0001,
+      "loss": 1.8894,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3715364507696776,
+      "grad_norm": 0.40069088339805603,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 3198
+    },
+    {
+      "epoch": 0.3716526285216381,
+      "grad_norm": 0.4221431016921997,
+      "learning_rate": 0.0001,
+      "loss": 1.847,
+      "step": 3199
+    },
+    {
+      "epoch": 0.3717688062735986,
+      "grad_norm": 0.4118482768535614,
+      "learning_rate": 0.0001,
+      "loss": 1.7249,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3718849840255591,
+      "grad_norm": 0.39366352558135986,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 3201
+    },
+    {
+      "epoch": 0.37200116177751963,
+      "grad_norm": 0.38057029247283936,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 3202
+    },
+    {
+      "epoch": 0.3721173395294801,
+      "grad_norm": 0.41096463799476624,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 3203
+    },
+    {
+      "epoch": 0.3722335172814406,
+      "grad_norm": 0.39210009574890137,
+      "learning_rate": 0.0001,
+      "loss": 1.7374,
+      "step": 3204
+    },
+    {
+      "epoch": 0.3723496950334011,
+      "grad_norm": 0.404095858335495,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3724658727853616,
+      "grad_norm": 0.4096939265727997,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 3206
+    },
+    {
+      "epoch": 0.37258205053732213,
+      "grad_norm": 0.41257479786872864,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 3207
+    },
+    {
+      "epoch": 0.3726982282892826,
+      "grad_norm": 0.4033520221710205,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 3208
+    },
+    {
+      "epoch": 0.3728144060412431,
+      "grad_norm": 0.38651078939437866,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 3209
+    },
+    {
+      "epoch": 0.3729305837932036,
+      "grad_norm": 0.40973809361457825,
+      "learning_rate": 0.0001,
+      "loss": 1.822,
+      "step": 3210
+    },
+    {
+      "epoch": 0.3730467615451641,
+      "grad_norm": 0.40880417823791504,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 3211
+    },
+    {
+      "epoch": 0.3731629392971246,
+      "grad_norm": 0.4003618061542511,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3732791170490851,
+      "grad_norm": 0.38030943274497986,
+      "learning_rate": 0.0001,
+      "loss": 1.4813,
+      "step": 3213
+    },
+    {
+      "epoch": 0.3733952948010456,
+      "grad_norm": 0.4228413999080658,
+      "learning_rate": 0.0001,
+      "loss": 1.6994,
+      "step": 3214
+    },
+    {
+      "epoch": 0.3735114725530061,
+      "grad_norm": 0.3931327164173126,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 3215
+    },
+    {
+      "epoch": 0.3736276503049666,
+      "grad_norm": 0.39333489537239075,
+      "learning_rate": 0.0001,
+      "loss": 1.7909,
+      "step": 3216
+    },
+    {
+      "epoch": 0.3737438280569271,
+      "grad_norm": 0.4146861433982849,
+      "learning_rate": 0.0001,
+      "loss": 1.9366,
+      "step": 3217
+    },
+    {
+      "epoch": 0.3738600058088876,
+      "grad_norm": 0.3847891390323639,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 3218
+    },
+    {
+      "epoch": 0.3739761835608481,
+      "grad_norm": 0.3797701299190521,
+      "learning_rate": 0.0001,
+      "loss": 1.6363,
+      "step": 3219
+    },
+    {
+      "epoch": 0.3740923613128086,
+      "grad_norm": 0.38476166129112244,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3742085390647691,
+      "grad_norm": 0.39565321803092957,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 3221
+    },
+    {
+      "epoch": 0.3743247168167296,
+      "grad_norm": 0.38622844219207764,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 3222
+    },
+    {
+      "epoch": 0.3744408945686901,
+      "grad_norm": 0.39825373888015747,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 3223
+    },
+    {
+      "epoch": 0.3745570723206506,
+      "grad_norm": 0.35864919424057007,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 3224
+    },
+    {
+      "epoch": 0.3746732500726111,
+      "grad_norm": 0.40784189105033875,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 3225
+    },
+    {
+      "epoch": 0.3747894278245716,
+      "grad_norm": 0.41927802562713623,
+      "learning_rate": 0.0001,
+      "loss": 1.7068,
+      "step": 3226
+    },
+    {
+      "epoch": 0.3749056055765321,
+      "grad_norm": 0.44369855523109436,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 3227
+    },
+    {
+      "epoch": 0.3750217833284926,
+      "grad_norm": 0.36899685859680176,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3751379610804531,
+      "grad_norm": 0.3842264711856842,
+      "learning_rate": 0.0001,
+      "loss": 1.7682,
+      "step": 3229
+    },
+    {
+      "epoch": 0.3752541388324136,
+      "grad_norm": 0.41173362731933594,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 3230
+    },
+    {
+      "epoch": 0.3753703165843741,
+      "grad_norm": 0.37613826990127563,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 3231
+    },
+    {
+      "epoch": 0.3754864943363346,
+      "grad_norm": 0.4120997190475464,
+      "learning_rate": 0.0001,
+      "loss": 1.8258,
+      "step": 3232
+    },
+    {
+      "epoch": 0.37560267208829506,
+      "grad_norm": 0.4014083743095398,
+      "learning_rate": 0.0001,
+      "loss": 1.7189,
+      "step": 3233
+    },
+    {
+      "epoch": 0.3757188498402556,
+      "grad_norm": 0.38191670179367065,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 3234
+    },
+    {
+      "epoch": 0.3758350275922161,
+      "grad_norm": 0.3927380442619324,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3759512053441766,
+      "grad_norm": 0.39120250940322876,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 3236
+    },
+    {
+      "epoch": 0.3760673830961371,
+      "grad_norm": 0.36845624446868896,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 3237
+    },
+    {
+      "epoch": 0.37618356084809756,
+      "grad_norm": 0.39836472272872925,
+      "learning_rate": 0.0001,
+      "loss": 1.7282,
+      "step": 3238
+    },
+    {
+      "epoch": 0.3762997386000581,
+      "grad_norm": 0.3965427279472351,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 3239
+    },
+    {
+      "epoch": 0.3764159163520186,
+      "grad_norm": 0.4197937846183777,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3765320941039791,
+      "grad_norm": 0.42778176069259644,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 3241
+    },
+    {
+      "epoch": 0.3766482718559396,
+      "grad_norm": 0.3915843665599823,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 3242
+    },
+    {
+      "epoch": 0.37676444960790006,
+      "grad_norm": 0.401483416557312,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3768806273598606,
+      "grad_norm": 0.4040675163269043,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 3244
+    },
+    {
+      "epoch": 0.3769968051118211,
+      "grad_norm": 0.39297956228256226,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 3245
+    },
+    {
+      "epoch": 0.3771129828637816,
+      "grad_norm": 0.3960307240486145,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 3246
+    },
+    {
+      "epoch": 0.3772291606157421,
+      "grad_norm": 0.4098290205001831,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 3247
+    },
+    {
+      "epoch": 0.37734533836770257,
+      "grad_norm": 0.4199242889881134,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 3248
+    },
+    {
+      "epoch": 0.3774615161196631,
+      "grad_norm": 0.4626270830631256,
+      "learning_rate": 0.0001,
+      "loss": 1.9117,
+      "step": 3249
+    },
+    {
+      "epoch": 0.3775776938716236,
+      "grad_norm": 0.3856443464756012,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3776938716235841,
+      "grad_norm": 0.3848719000816345,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 3251
+    },
+    {
+      "epoch": 0.3778100493755446,
+      "grad_norm": 0.4036993384361267,
+      "learning_rate": 0.0001,
+      "loss": 1.8056,
+      "step": 3252
+    },
+    {
+      "epoch": 0.37792622712750507,
+      "grad_norm": 0.42733898758888245,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 3253
+    },
+    {
+      "epoch": 0.3780424048794656,
+      "grad_norm": 0.3956359922885895,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 3254
+    },
+    {
+      "epoch": 0.3781585826314261,
+      "grad_norm": 0.391928106546402,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3782747603833866,
+      "grad_norm": 0.4201536774635315,
+      "learning_rate": 0.0001,
+      "loss": 1.8002,
+      "step": 3256
+    },
+    {
+      "epoch": 0.3783909381353471,
+      "grad_norm": 0.41449272632598877,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 3257
+    },
+    {
+      "epoch": 0.37850711588730757,
+      "grad_norm": 0.4059070944786072,
+      "learning_rate": 0.0001,
+      "loss": 1.6892,
+      "step": 3258
+    },
+    {
+      "epoch": 0.37862329363926805,
+      "grad_norm": 0.40105417370796204,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 3259
+    },
+    {
+      "epoch": 0.3787394713912286,
+      "grad_norm": 0.4251152276992798,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 3260
+    },
+    {
+      "epoch": 0.3788556491431891,
+      "grad_norm": 0.3787403702735901,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 3261
+    },
+    {
+      "epoch": 0.3789718268951496,
+      "grad_norm": 0.39760053157806396,
+      "learning_rate": 0.0001,
+      "loss": 1.7866,
+      "step": 3262
+    },
+    {
+      "epoch": 0.37908800464711007,
+      "grad_norm": 0.3865306079387665,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 3263
+    },
+    {
+      "epoch": 0.37920418239907056,
+      "grad_norm": 0.40044572949409485,
+      "learning_rate": 0.0001,
+      "loss": 1.7403,
+      "step": 3264
+    },
+    {
+      "epoch": 0.3793203601510311,
+      "grad_norm": 0.36181673407554626,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 3265
+    },
+    {
+      "epoch": 0.3794365379029916,
+      "grad_norm": 0.4151827096939087,
+      "learning_rate": 0.0001,
+      "loss": 1.7351,
+      "step": 3266
+    },
+    {
+      "epoch": 0.3795527156549521,
+      "grad_norm": 0.3959139287471771,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 3267
+    },
+    {
+      "epoch": 0.37966889340691257,
+      "grad_norm": 0.40265771746635437,
+      "learning_rate": 0.0001,
+      "loss": 1.7255,
+      "step": 3268
+    },
+    {
+      "epoch": 0.37978507115887306,
+      "grad_norm": 0.39795050024986267,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 3269
+    },
+    {
+      "epoch": 0.3799012489108336,
+      "grad_norm": 0.38326337933540344,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3800174266627941,
+      "grad_norm": 0.4339217245578766,
+      "learning_rate": 0.0001,
+      "loss": 1.7723,
+      "step": 3271
+    },
+    {
+      "epoch": 0.3801336044147546,
+      "grad_norm": 0.391989141702652,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 3272
+    },
+    {
+      "epoch": 0.38024978216671507,
+      "grad_norm": 0.3686724901199341,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 3273
+    },
+    {
+      "epoch": 0.38036595991867556,
+      "grad_norm": 0.3889879584312439,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 3274
+    },
+    {
+      "epoch": 0.3804821376706361,
+      "grad_norm": 0.37887483835220337,
+      "learning_rate": 0.0001,
+      "loss": 1.64,
+      "step": 3275
+    },
+    {
+      "epoch": 0.3805983154225966,
+      "grad_norm": 0.3940137028694153,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 3276
+    },
+    {
+      "epoch": 0.3807144931745571,
+      "grad_norm": 0.38315024971961975,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 3277
+    },
+    {
+      "epoch": 0.38083067092651757,
+      "grad_norm": 0.43694132566452026,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 3278
+    },
+    {
+      "epoch": 0.38094684867847806,
+      "grad_norm": 0.37439560890197754,
+      "learning_rate": 0.0001,
+      "loss": 1.4972,
+      "step": 3279
+    },
+    {
+      "epoch": 0.38106302643043855,
+      "grad_norm": 0.39639008045196533,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 3280
+    },
+    {
+      "epoch": 0.3811792041823991,
+      "grad_norm": 0.41483819484710693,
+      "learning_rate": 0.0001,
+      "loss": 1.7358,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3812953819343596,
+      "grad_norm": 0.38614705204963684,
+      "learning_rate": 0.0001,
+      "loss": 1.884,
+      "step": 3282
+    },
+    {
+      "epoch": 0.38141155968632007,
+      "grad_norm": 0.4163734018802643,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 3283
+    },
+    {
+      "epoch": 0.38152773743828056,
+      "grad_norm": 0.3871447741985321,
+      "learning_rate": 0.0001,
+      "loss": 1.8271,
+      "step": 3284
+    },
+    {
+      "epoch": 0.38164391519024105,
+      "grad_norm": 0.4444115161895752,
+      "learning_rate": 0.0001,
+      "loss": 1.7751,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3817600929422016,
+      "grad_norm": 0.3978256285190582,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 3286
+    },
+    {
+      "epoch": 0.3818762706941621,
+      "grad_norm": 0.37365588545799255,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 3287
+    },
+    {
+      "epoch": 0.38199244844612257,
+      "grad_norm": 0.39886194467544556,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 3288
+    },
+    {
+      "epoch": 0.38210862619808306,
+      "grad_norm": 0.4288283586502075,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 3289
+    },
+    {
+      "epoch": 0.38222480395004355,
+      "grad_norm": 0.40688732266426086,
+      "learning_rate": 0.0001,
+      "loss": 1.8548,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3823409817020041,
+      "grad_norm": 0.3931783437728882,
+      "learning_rate": 0.0001,
+      "loss": 1.7435,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3824571594539646,
+      "grad_norm": 0.41494035720825195,
+      "learning_rate": 0.0001,
+      "loss": 1.7333,
+      "step": 3292
+    },
+    {
+      "epoch": 0.38257333720592507,
+      "grad_norm": 0.43234965205192566,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 3293
+    },
+    {
+      "epoch": 0.38268951495788556,
+      "grad_norm": 0.3797743022441864,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 3294
+    },
+    {
+      "epoch": 0.38280569270984605,
+      "grad_norm": 0.39665845036506653,
+      "learning_rate": 0.0001,
+      "loss": 1.6871,
+      "step": 3295
+    },
+    {
+      "epoch": 0.3829218704618066,
+      "grad_norm": 0.430963933467865,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 3296
+    },
+    {
+      "epoch": 0.3830380482137671,
+      "grad_norm": 0.3795839250087738,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 3297
+    },
+    {
+      "epoch": 0.38315422596572757,
+      "grad_norm": 0.3861173987388611,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 3298
+    },
+    {
+      "epoch": 0.38327040371768806,
+      "grad_norm": 0.4656057357788086,
+      "learning_rate": 0.0001,
+      "loss": 1.7191,
+      "step": 3299
+    },
+    {
+      "epoch": 0.38338658146964855,
+      "grad_norm": 0.42121168971061707,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 3300
+    },
+    {
+      "epoch": 0.38350275922160904,
+      "grad_norm": 0.4215461313724518,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 3301
+    },
+    {
+      "epoch": 0.3836189369735696,
+      "grad_norm": 0.3778843581676483,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 3302
+    },
+    {
+      "epoch": 0.38373511472553007,
+      "grad_norm": 0.372529000043869,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 3303
+    },
+    {
+      "epoch": 0.38385129247749056,
+      "grad_norm": 0.3816990852355957,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 3304
+    },
+    {
+      "epoch": 0.38396747022945105,
+      "grad_norm": 0.3976283669471741,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 3305
+    },
+    {
+      "epoch": 0.38408364798141154,
+      "grad_norm": 0.38981184363365173,
+      "learning_rate": 0.0001,
+      "loss": 1.7524,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3841998257333721,
+      "grad_norm": 0.4135308265686035,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 3307
+    },
+    {
+      "epoch": 0.38431600348533257,
+      "grad_norm": 0.39930155873298645,
+      "learning_rate": 0.0001,
+      "loss": 1.785,
+      "step": 3308
+    },
+    {
+      "epoch": 0.38443218123729306,
+      "grad_norm": 0.36703500151634216,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 3309
+    },
+    {
+      "epoch": 0.38454835898925355,
+      "grad_norm": 0.396085262298584,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 3310
+    },
+    {
+      "epoch": 0.38466453674121404,
+      "grad_norm": 0.4066247045993805,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 3311
+    },
+    {
+      "epoch": 0.3847807144931746,
+      "grad_norm": 0.4137192368507385,
+      "learning_rate": 0.0001,
+      "loss": 1.6474,
+      "step": 3312
+    },
+    {
+      "epoch": 0.38489689224513507,
+      "grad_norm": 0.44450923800468445,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 3313
+    },
+    {
+      "epoch": 0.38501306999709556,
+      "grad_norm": 0.4225256145000458,
+      "learning_rate": 0.0001,
+      "loss": 1.5993,
+      "step": 3314
+    },
+    {
+      "epoch": 0.38512924774905605,
+      "grad_norm": 0.37783291935920715,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 3315
+    },
+    {
+      "epoch": 0.38524542550101654,
+      "grad_norm": 0.3956874907016754,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 3316
+    },
+    {
+      "epoch": 0.3853616032529771,
+      "grad_norm": 0.39644670486450195,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 3317
+    },
+    {
+      "epoch": 0.38547778100493757,
+      "grad_norm": 0.4001430869102478,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 3318
+    },
+    {
+      "epoch": 0.38559395875689806,
+      "grad_norm": 0.42300862073898315,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 3319
+    },
+    {
+      "epoch": 0.38571013650885855,
+      "grad_norm": 0.4075738191604614,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 3320
+    },
+    {
+      "epoch": 0.38582631426081904,
+      "grad_norm": 0.40944305062294006,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 3321
+    },
+    {
+      "epoch": 0.3859424920127795,
+      "grad_norm": 0.41256004571914673,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 3322
+    },
+    {
+      "epoch": 0.38605866976474007,
+      "grad_norm": 0.45253586769104004,
+      "learning_rate": 0.0001,
+      "loss": 1.9485,
+      "step": 3323
+    },
+    {
+      "epoch": 0.38617484751670056,
+      "grad_norm": 0.3973802626132965,
+      "learning_rate": 0.0001,
+      "loss": 1.7537,
+      "step": 3324
+    },
+    {
+      "epoch": 0.38629102526866105,
+      "grad_norm": 0.3796943724155426,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 3325
+    },
+    {
+      "epoch": 0.38640720302062154,
+      "grad_norm": 0.38597339391708374,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 3326
+    },
+    {
+      "epoch": 0.386523380772582,
+      "grad_norm": 0.4044647514820099,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 3327
+    },
+    {
+      "epoch": 0.38663955852454257,
+      "grad_norm": 0.38887619972229004,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 3328
+    },
+    {
+      "epoch": 0.38675573627650306,
+      "grad_norm": 0.3996337354183197,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 3329
+    },
+    {
+      "epoch": 0.38687191402846355,
+      "grad_norm": 0.3903794288635254,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 3330
+    },
+    {
+      "epoch": 0.38698809178042404,
+      "grad_norm": 0.41847532987594604,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 3331
+    },
+    {
+      "epoch": 0.3871042695323845,
+      "grad_norm": 0.3963734805583954,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 3332
+    },
+    {
+      "epoch": 0.38722044728434507,
+      "grad_norm": 0.3818768858909607,
+      "learning_rate": 0.0001,
+      "loss": 1.6953,
+      "step": 3333
+    },
+    {
+      "epoch": 0.38733662503630556,
+      "grad_norm": 0.4042767286300659,
+      "learning_rate": 0.0001,
+      "loss": 1.6731,
+      "step": 3334
+    },
+    {
+      "epoch": 0.38745280278826605,
+      "grad_norm": 0.3764187693595886,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 3335
+    },
+    {
+      "epoch": 0.38756898054022654,
+      "grad_norm": 0.41739174723625183,
+      "learning_rate": 0.0001,
+      "loss": 1.771,
+      "step": 3336
+    },
+    {
+      "epoch": 0.387685158292187,
+      "grad_norm": 0.3916127383708954,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 3337
+    },
+    {
+      "epoch": 0.38780133604414757,
+      "grad_norm": 0.3758176863193512,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 3338
+    },
+    {
+      "epoch": 0.38791751379610806,
+      "grad_norm": 0.37665247917175293,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 3339
+    },
+    {
+      "epoch": 0.38803369154806855,
+      "grad_norm": 0.38842839002609253,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 3340
+    },
+    {
+      "epoch": 0.38814986930002904,
+      "grad_norm": 0.4044833183288574,
+      "learning_rate": 0.0001,
+      "loss": 1.7784,
+      "step": 3341
+    },
+    {
+      "epoch": 0.3882660470519895,
+      "grad_norm": 0.3983111083507538,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 3342
+    },
+    {
+      "epoch": 0.38838222480395007,
+      "grad_norm": 0.3763045072555542,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 3343
+    },
+    {
+      "epoch": 0.38849840255591056,
+      "grad_norm": 0.43065011501312256,
+      "learning_rate": 0.0001,
+      "loss": 1.7841,
+      "step": 3344
+    },
+    {
+      "epoch": 0.38861458030787105,
+      "grad_norm": 0.3971543312072754,
+      "learning_rate": 0.0001,
+      "loss": 1.7052,
+      "step": 3345
+    },
+    {
+      "epoch": 0.38873075805983154,
+      "grad_norm": 0.4392778277397156,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 3346
+    },
+    {
+      "epoch": 0.388846935811792,
+      "grad_norm": 0.41214969754219055,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 3347
+    },
+    {
+      "epoch": 0.3889631135637525,
+      "grad_norm": 0.3774077296257019,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 3348
+    },
+    {
+      "epoch": 0.38907929131571306,
+      "grad_norm": 0.4119464159011841,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 3349
+    },
+    {
+      "epoch": 0.38919546906767355,
+      "grad_norm": 0.4135492146015167,
+      "learning_rate": 0.0001,
+      "loss": 1.7139,
+      "step": 3350
+    },
+    {
+      "epoch": 0.38931164681963404,
+      "grad_norm": 0.41081520915031433,
+      "learning_rate": 0.0001,
+      "loss": 1.8803,
+      "step": 3351
+    },
+    {
+      "epoch": 0.3894278245715945,
+      "grad_norm": 0.4038920998573303,
+      "learning_rate": 0.0001,
+      "loss": 1.8064,
+      "step": 3352
+    },
+    {
+      "epoch": 0.389544002323555,
+      "grad_norm": 0.4117118716239929,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 3353
+    },
+    {
+      "epoch": 0.38966018007551556,
+      "grad_norm": 0.3965054452419281,
+      "learning_rate": 0.0001,
+      "loss": 1.7807,
+      "step": 3354
+    },
+    {
+      "epoch": 0.38977635782747605,
+      "grad_norm": 0.3699134886264801,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 3355
+    },
+    {
+      "epoch": 0.38989253557943654,
+      "grad_norm": 0.3890858590602875,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 3356
+    },
+    {
+      "epoch": 0.39000871333139703,
+      "grad_norm": 0.3924883008003235,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 3357
+    },
+    {
+      "epoch": 0.3901248910833575,
+      "grad_norm": 0.40697407722473145,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 3358
+    },
+    {
+      "epoch": 0.39024106883531806,
+      "grad_norm": 0.3949110507965088,
+      "learning_rate": 0.0001,
+      "loss": 1.7615,
+      "step": 3359
+    },
+    {
+      "epoch": 0.39035724658727855,
+      "grad_norm": 0.4023362100124359,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 3360
+    },
+    {
+      "epoch": 0.39047342433923904,
+      "grad_norm": 0.3989759385585785,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 3361
+    },
+    {
+      "epoch": 0.39058960209119953,
+      "grad_norm": 0.3800547420978546,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 3362
+    },
+    {
+      "epoch": 0.39070577984316,
+      "grad_norm": 0.408544659614563,
+      "learning_rate": 0.0001,
+      "loss": 1.7846,
+      "step": 3363
+    },
+    {
+      "epoch": 0.39082195759512056,
+      "grad_norm": 0.41592341661453247,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 3364
+    },
+    {
+      "epoch": 0.39093813534708105,
+      "grad_norm": 0.4047764539718628,
+      "learning_rate": 0.0001,
+      "loss": 1.88,
+      "step": 3365
+    },
+    {
+      "epoch": 0.39105431309904154,
+      "grad_norm": 0.3952164649963379,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 3366
+    },
+    {
+      "epoch": 0.39117049085100203,
+      "grad_norm": 0.39583536982536316,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3912866686029625,
+      "grad_norm": 0.40658921003341675,
+      "learning_rate": 0.0001,
+      "loss": 1.6708,
+      "step": 3368
+    },
+    {
+      "epoch": 0.391402846354923,
+      "grad_norm": 0.42151522636413574,
+      "learning_rate": 0.0001,
+      "loss": 1.7088,
+      "step": 3369
+    },
+    {
+      "epoch": 0.39151902410688355,
+      "grad_norm": 0.42344412207603455,
+      "learning_rate": 0.0001,
+      "loss": 1.7888,
+      "step": 3370
+    },
+    {
+      "epoch": 0.39163520185884404,
+      "grad_norm": 0.3720431625843048,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 3371
+    },
+    {
+      "epoch": 0.39175137961080453,
+      "grad_norm": 0.39411425590515137,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 3372
+    },
+    {
+      "epoch": 0.391867557362765,
+      "grad_norm": 0.3884545862674713,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 3373
+    },
+    {
+      "epoch": 0.3919837351147255,
+      "grad_norm": 0.4163500666618347,
+      "learning_rate": 0.0001,
+      "loss": 1.7583,
+      "step": 3374
+    },
+    {
+      "epoch": 0.39209991286668605,
+      "grad_norm": 0.38707318902015686,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 3375
+    },
+    {
+      "epoch": 0.39221609061864654,
+      "grad_norm": 0.4354010820388794,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 3376
+    },
+    {
+      "epoch": 0.39233226837060703,
+      "grad_norm": 0.39656829833984375,
+      "learning_rate": 0.0001,
+      "loss": 1.6507,
+      "step": 3377
+    },
+    {
+      "epoch": 0.3924484461225675,
+      "grad_norm": 0.39710286259651184,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 3378
+    },
+    {
+      "epoch": 0.392564623874528,
+      "grad_norm": 0.40320920944213867,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 3379
+    },
+    {
+      "epoch": 0.39268080162648855,
+      "grad_norm": 0.40849626064300537,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 3380
+    },
+    {
+      "epoch": 0.39279697937844904,
+      "grad_norm": 0.37604308128356934,
+      "learning_rate": 0.0001,
+      "loss": 1.6333,
+      "step": 3381
+    },
+    {
+      "epoch": 0.39291315713040953,
+      "grad_norm": 0.38765400648117065,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 3382
+    },
+    {
+      "epoch": 0.39302933488237,
+      "grad_norm": 0.38033145666122437,
+      "learning_rate": 0.0001,
+      "loss": 1.7555,
+      "step": 3383
+    },
+    {
+      "epoch": 0.3931455126343305,
+      "grad_norm": 0.39232638478279114,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 3384
+    },
+    {
+      "epoch": 0.39326169038629105,
+      "grad_norm": 0.41576313972473145,
+      "learning_rate": 0.0001,
+      "loss": 1.6984,
+      "step": 3385
+    },
+    {
+      "epoch": 0.39337786813825154,
+      "grad_norm": 0.40268710255622864,
+      "learning_rate": 0.0001,
+      "loss": 1.7465,
+      "step": 3386
+    },
+    {
+      "epoch": 0.39349404589021203,
+      "grad_norm": 0.3708171546459198,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3936102236421725,
+      "grad_norm": 0.3807376027107239,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 3388
+    },
+    {
+      "epoch": 0.393726401394133,
+      "grad_norm": 0.39858368039131165,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 3389
+    },
+    {
+      "epoch": 0.3938425791460935,
+      "grad_norm": 0.3780364692211151,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 3390
+    },
+    {
+      "epoch": 0.39395875689805404,
+      "grad_norm": 0.39685890078544617,
+      "learning_rate": 0.0001,
+      "loss": 1.7051,
+      "step": 3391
+    },
+    {
+      "epoch": 0.39407493465001453,
+      "grad_norm": 0.4291624128818512,
+      "learning_rate": 0.0001,
+      "loss": 1.8533,
+      "step": 3392
+    },
+    {
+      "epoch": 0.394191112401975,
+      "grad_norm": 0.366272509098053,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3943072901539355,
+      "grad_norm": 0.3910382091999054,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 3394
+    },
+    {
+      "epoch": 0.394423467905896,
+      "grad_norm": 0.4170825481414795,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 3395
+    },
+    {
+      "epoch": 0.39453964565785654,
+      "grad_norm": 0.3881252706050873,
+      "learning_rate": 0.0001,
+      "loss": 1.7354,
+      "step": 3396
+    },
+    {
+      "epoch": 0.39465582340981703,
+      "grad_norm": 0.3868289291858673,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3947720011617775,
+      "grad_norm": 0.38081884384155273,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 3398
+    },
+    {
+      "epoch": 0.394888178913738,
+      "grad_norm": 0.39945870637893677,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3950043566656985,
+      "grad_norm": 0.4090173840522766,
+      "learning_rate": 0.0001,
+      "loss": 1.7628,
+      "step": 3400
+    },
+    {
+      "epoch": 0.39512053441765904,
+      "grad_norm": 0.3833993971347809,
+      "learning_rate": 0.0001,
+      "loss": 1.4581,
+      "step": 3401
+    },
+    {
+      "epoch": 0.39523671216961953,
+      "grad_norm": 0.4201992154121399,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 3402
+    },
+    {
+      "epoch": 0.39535288992158,
+      "grad_norm": 0.4019543528556824,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3954690676735405,
+      "grad_norm": 0.3776283860206604,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 3404
+    },
+    {
+      "epoch": 0.395585245425501,
+      "grad_norm": 0.38334932923316956,
+      "learning_rate": 0.0001,
+      "loss": 1.3783,
+      "step": 3405
+    },
+    {
+      "epoch": 0.39570142317746154,
+      "grad_norm": 0.3982570767402649,
+      "learning_rate": 0.0001,
+      "loss": 1.6886,
+      "step": 3406
+    },
+    {
+      "epoch": 0.39581760092942203,
+      "grad_norm": 0.3854517340660095,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 3407
+    },
+    {
+      "epoch": 0.3959337786813825,
+      "grad_norm": 0.3867027461528778,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 3408
+    },
+    {
+      "epoch": 0.396049956433343,
+      "grad_norm": 0.39494284987449646,
+      "learning_rate": 0.0001,
+      "loss": 1.5789,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3961661341853035,
+      "grad_norm": 0.40102261304855347,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 3410
+    },
+    {
+      "epoch": 0.396282311937264,
+      "grad_norm": 0.38136184215545654,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 3411
+    },
+    {
+      "epoch": 0.39639848968922453,
+      "grad_norm": 0.40026000142097473,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 3412
+    },
+    {
+      "epoch": 0.396514667441185,
+      "grad_norm": 0.4199739694595337,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3966308451931455,
+      "grad_norm": 0.4212833344936371,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 3414
+    },
+    {
+      "epoch": 0.396747022945106,
+      "grad_norm": 0.39813485741615295,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3968632006970665,
+      "grad_norm": 0.40672633051872253,
+      "learning_rate": 0.0001,
+      "loss": 1.7394,
+      "step": 3416
+    },
+    {
+      "epoch": 0.39697937844902703,
+      "grad_norm": 0.42593738436698914,
+      "learning_rate": 0.0001,
+      "loss": 1.8138,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3970955562009875,
+      "grad_norm": 0.39474889636039734,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 3418
+    },
+    {
+      "epoch": 0.397211733952948,
+      "grad_norm": 0.4089927077293396,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 3419
+    },
+    {
+      "epoch": 0.3973279117049085,
+      "grad_norm": 0.3997913897037506,
+      "learning_rate": 0.0001,
+      "loss": 1.8294,
+      "step": 3420
+    },
+    {
+      "epoch": 0.397444089456869,
+      "grad_norm": 0.46322256326675415,
+      "learning_rate": 0.0001,
+      "loss": 1.7424,
+      "step": 3421
+    },
+    {
+      "epoch": 0.39756026720882953,
+      "grad_norm": 0.3604629337787628,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 3422
+    },
+    {
+      "epoch": 0.39767644496079,
+      "grad_norm": 0.3815920352935791,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3977926227127505,
+      "grad_norm": 0.37603339552879333,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 3424
+    },
+    {
+      "epoch": 0.397908800464711,
+      "grad_norm": 0.4085451364517212,
+      "learning_rate": 0.0001,
+      "loss": 1.7673,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3980249782166715,
+      "grad_norm": 0.46148502826690674,
+      "learning_rate": 0.0001,
+      "loss": 1.9431,
+      "step": 3426
+    },
+    {
+      "epoch": 0.39814115596863203,
+      "grad_norm": 0.3642321527004242,
+      "learning_rate": 0.0001,
+      "loss": 1.3867,
+      "step": 3427
+    },
+    {
+      "epoch": 0.3982573337205925,
+      "grad_norm": 0.39448490738868713,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 3428
+    },
+    {
+      "epoch": 0.398373511472553,
+      "grad_norm": 0.37852832674980164,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3984896892245135,
+      "grad_norm": 0.44772058725357056,
+      "learning_rate": 0.0001,
+      "loss": 1.7785,
+      "step": 3430
+    },
+    {
+      "epoch": 0.398605866976474,
+      "grad_norm": 0.39221101999282837,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3987220447284345,
+      "grad_norm": 0.38119301199913025,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 3432
+    },
+    {
+      "epoch": 0.398838222480395,
+      "grad_norm": 0.4090498685836792,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 3433
+    },
+    {
+      "epoch": 0.3989544002323555,
+      "grad_norm": 0.37487149238586426,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 3434
+    },
+    {
+      "epoch": 0.399070577984316,
+      "grad_norm": 0.39408043026924133,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 3435
+    },
+    {
+      "epoch": 0.3991867557362765,
+      "grad_norm": 0.38384729623794556,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 3436
+    },
+    {
+      "epoch": 0.399302933488237,
+      "grad_norm": 0.4172252416610718,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3994191112401975,
+      "grad_norm": 0.3903944790363312,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 3438
+    },
+    {
+      "epoch": 0.399535288992158,
+      "grad_norm": 0.3812950849533081,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3996514667441185,
+      "grad_norm": 0.4021446704864502,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 3440
+    },
+    {
+      "epoch": 0.399767644496079,
+      "grad_norm": 0.3971196711063385,
+      "learning_rate": 0.0001,
+      "loss": 1.7095,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3998838222480395,
+      "grad_norm": 0.40641874074935913,
+      "learning_rate": 0.0001,
+      "loss": 1.572,
+      "step": 3442
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.41259273886680603,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 3443
+    },
+    {
+      "epoch": 0.4001161777519605,
+      "grad_norm": 0.35140666365623474,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 3444
+    },
+    {
+      "epoch": 0.400232355503921,
+      "grad_norm": 0.3907198905944824,
+      "learning_rate": 0.0001,
+      "loss": 1.5896,
+      "step": 3445
+    },
+    {
+      "epoch": 0.4003485332558815,
+      "grad_norm": 0.41215452551841736,
+      "learning_rate": 0.0001,
+      "loss": 1.5177,
+      "step": 3446
+    },
+    {
+      "epoch": 0.400464711007842,
+      "grad_norm": 0.41600051522254944,
+      "learning_rate": 0.0001,
+      "loss": 1.9322,
+      "step": 3447
+    },
+    {
+      "epoch": 0.4005808887598025,
+      "grad_norm": 0.4017678201198578,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 3448
+    },
+    {
+      "epoch": 0.400697066511763,
+      "grad_norm": 0.4211594760417938,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 3449
+    },
+    {
+      "epoch": 0.4008132442637235,
+      "grad_norm": 0.4117183983325958,
+      "learning_rate": 0.0001,
+      "loss": 1.8139,
+      "step": 3450
+    },
+    {
+      "epoch": 0.400929422015684,
+      "grad_norm": 0.39731109142303467,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 3451
+    },
+    {
+      "epoch": 0.4010455997676445,
+      "grad_norm": 0.3891085088253021,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 3452
+    },
+    {
+      "epoch": 0.401161777519605,
+      "grad_norm": 0.3980685770511627,
+      "learning_rate": 0.0001,
+      "loss": 1.7527,
+      "step": 3453
+    },
+    {
+      "epoch": 0.4012779552715655,
+      "grad_norm": 0.43081504106521606,
+      "learning_rate": 0.0001,
+      "loss": 1.7178,
+      "step": 3454
+    },
+    {
+      "epoch": 0.401394133023526,
+      "grad_norm": 0.38269612193107605,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 3455
+    },
+    {
+      "epoch": 0.4015103107754865,
+      "grad_norm": 0.4062149226665497,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 3456
+    },
+    {
+      "epoch": 0.401626488527447,
+      "grad_norm": 0.38731974363327026,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 3457
+    },
+    {
+      "epoch": 0.40174266627940747,
+      "grad_norm": 0.4293443560600281,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 3458
+    },
+    {
+      "epoch": 0.401858844031368,
+      "grad_norm": 0.3938728868961334,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 3459
+    },
+    {
+      "epoch": 0.4019750217833285,
+      "grad_norm": 0.41334471106529236,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 3460
+    },
+    {
+      "epoch": 0.402091199535289,
+      "grad_norm": 0.39084866642951965,
+      "learning_rate": 0.0001,
+      "loss": 1.6759,
+      "step": 3461
+    },
+    {
+      "epoch": 0.4022073772872495,
+      "grad_norm": 0.4102165400981903,
+      "learning_rate": 0.0001,
+      "loss": 1.6758,
+      "step": 3462
+    },
+    {
+      "epoch": 0.40232355503920997,
+      "grad_norm": 0.4234514534473419,
+      "learning_rate": 0.0001,
+      "loss": 1.7546,
+      "step": 3463
+    },
+    {
+      "epoch": 0.4024397327911705,
+      "grad_norm": 0.38048434257507324,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 3464
+    },
+    {
+      "epoch": 0.402555910543131,
+      "grad_norm": 0.39564260840415955,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 3465
+    },
+    {
+      "epoch": 0.4026720882950915,
+      "grad_norm": 0.40055975317955017,
+      "learning_rate": 0.0001,
+      "loss": 1.7346,
+      "step": 3466
+    },
+    {
+      "epoch": 0.402788266047052,
+      "grad_norm": 0.424216628074646,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 3467
+    },
+    {
+      "epoch": 0.40290444379901247,
+      "grad_norm": 0.38856959342956543,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 3468
+    },
+    {
+      "epoch": 0.403020621550973,
+      "grad_norm": 0.4526919424533844,
+      "learning_rate": 0.0001,
+      "loss": 1.7348,
+      "step": 3469
+    },
+    {
+      "epoch": 0.4031367993029335,
+      "grad_norm": 0.4121178686618805,
+      "learning_rate": 0.0001,
+      "loss": 1.634,
+      "step": 3470
+    },
+    {
+      "epoch": 0.403252977054894,
+      "grad_norm": 0.41178447008132935,
+      "learning_rate": 0.0001,
+      "loss": 1.6574,
+      "step": 3471
+    },
+    {
+      "epoch": 0.4033691548068545,
+      "grad_norm": 0.3908507823944092,
+      "learning_rate": 0.0001,
+      "loss": 1.594,
+      "step": 3472
+    },
+    {
+      "epoch": 0.40348533255881497,
+      "grad_norm": 0.4167093336582184,
+      "learning_rate": 0.0001,
+      "loss": 1.7207,
+      "step": 3473
+    },
+    {
+      "epoch": 0.4036015103107755,
+      "grad_norm": 0.3978476822376251,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 3474
+    },
+    {
+      "epoch": 0.403717688062736,
+      "grad_norm": 0.4014563262462616,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 3475
+    },
+    {
+      "epoch": 0.4038338658146965,
+      "grad_norm": 0.42038458585739136,
+      "learning_rate": 0.0001,
+      "loss": 1.7169,
+      "step": 3476
+    },
+    {
+      "epoch": 0.403950043566657,
+      "grad_norm": 0.40541571378707886,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 3477
+    },
+    {
+      "epoch": 0.40406622131861747,
+      "grad_norm": 0.39469483494758606,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 3478
+    },
+    {
+      "epoch": 0.40418239907057796,
+      "grad_norm": 0.38308000564575195,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 3479
+    },
+    {
+      "epoch": 0.4042985768225385,
+      "grad_norm": 0.3976595103740692,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 3480
+    },
+    {
+      "epoch": 0.404414754574499,
+      "grad_norm": 0.4284829795360565,
+      "learning_rate": 0.0001,
+      "loss": 1.7053,
+      "step": 3481
+    },
+    {
+      "epoch": 0.4045309323264595,
+      "grad_norm": 0.36719101667404175,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 3482
+    },
+    {
+      "epoch": 0.40464711007841997,
+      "grad_norm": 0.4026053845882416,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 3483
+    },
+    {
+      "epoch": 0.40476328783038046,
+      "grad_norm": 0.40550971031188965,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 3484
+    },
+    {
+      "epoch": 0.404879465582341,
+      "grad_norm": 0.3787073493003845,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 3485
+    },
+    {
+      "epoch": 0.4049956433343015,
+      "grad_norm": 0.38911107182502747,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 3486
+    },
+    {
+      "epoch": 0.405111821086262,
+      "grad_norm": 0.420773446559906,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 3487
+    },
+    {
+      "epoch": 0.40522799883822247,
+      "grad_norm": 0.40293148159980774,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 3488
+    },
+    {
+      "epoch": 0.40534417659018296,
+      "grad_norm": 0.40412238240242004,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 3489
+    },
+    {
+      "epoch": 0.4054603543421435,
+      "grad_norm": 0.42694681882858276,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 3490
+    },
+    {
+      "epoch": 0.405576532094104,
+      "grad_norm": 0.3828969895839691,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 3491
+    },
+    {
+      "epoch": 0.4056927098460645,
+      "grad_norm": 0.4304792582988739,
+      "learning_rate": 0.0001,
+      "loss": 1.65,
+      "step": 3492
+    },
+    {
+      "epoch": 0.40580888759802497,
+      "grad_norm": 0.40881672501564026,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 3493
+    },
+    {
+      "epoch": 0.40592506534998546,
+      "grad_norm": 0.3886183202266693,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 3494
+    },
+    {
+      "epoch": 0.406041243101946,
+      "grad_norm": 0.3807865083217621,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 3495
+    },
+    {
+      "epoch": 0.4061574208539065,
+      "grad_norm": 0.4547371566295624,
+      "learning_rate": 0.0001,
+      "loss": 2.0104,
+      "step": 3496
+    },
+    {
+      "epoch": 0.406273598605867,
+      "grad_norm": 0.43122467398643494,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 3497
+    },
+    {
+      "epoch": 0.40638977635782747,
+      "grad_norm": 0.43641188740730286,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 3498
+    },
+    {
+      "epoch": 0.40650595410978796,
+      "grad_norm": 0.38105788826942444,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 3499
+    },
+    {
+      "epoch": 0.40662213186174845,
+      "grad_norm": 0.38278716802597046,
+      "learning_rate": 0.0001,
+      "loss": 1.6807,
+      "step": 3500
+    },
+    {
+      "epoch": 0.406738309613709,
+      "grad_norm": 0.4314371645450592,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 3501
+    },
+    {
+      "epoch": 0.4068544873656695,
+      "grad_norm": 0.39707180857658386,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 3502
+    },
+    {
+      "epoch": 0.40697066511762997,
+      "grad_norm": 0.40061885118484497,
+      "learning_rate": 0.0001,
+      "loss": 1.7397,
+      "step": 3503
+    },
+    {
+      "epoch": 0.40708684286959046,
+      "grad_norm": 0.39981821179389954,
+      "learning_rate": 0.0001,
+      "loss": 1.7075,
+      "step": 3504
+    },
+    {
+      "epoch": 0.40720302062155095,
+      "grad_norm": 0.41869446635246277,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 3505
+    },
+    {
+      "epoch": 0.4073191983735115,
+      "grad_norm": 0.4269830584526062,
+      "learning_rate": 0.0001,
+      "loss": 1.7423,
+      "step": 3506
+    },
+    {
+      "epoch": 0.407435376125472,
+      "grad_norm": 0.44403141736984253,
+      "learning_rate": 0.0001,
+      "loss": 1.7895,
+      "step": 3507
+    },
+    {
+      "epoch": 0.40755155387743247,
+      "grad_norm": 0.4114527404308319,
+      "learning_rate": 0.0001,
+      "loss": 1.6789,
+      "step": 3508
+    },
+    {
+      "epoch": 0.40766773162939296,
+      "grad_norm": 0.40180087089538574,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 3509
+    },
+    {
+      "epoch": 0.40778390938135345,
+      "grad_norm": 0.4387998580932617,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 3510
+    },
+    {
+      "epoch": 0.407900087133314,
+      "grad_norm": 0.43841442465782166,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 3511
+    },
+    {
+      "epoch": 0.4080162648852745,
+      "grad_norm": 0.41502171754837036,
+      "learning_rate": 0.0001,
+      "loss": 1.7085,
+      "step": 3512
+    },
+    {
+      "epoch": 0.40813244263723497,
+      "grad_norm": 0.40264812111854553,
+      "learning_rate": 0.0001,
+      "loss": 1.7709,
+      "step": 3513
+    },
+    {
+      "epoch": 0.40824862038919546,
+      "grad_norm": 0.39349353313446045,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 3514
+    },
+    {
+      "epoch": 0.40836479814115595,
+      "grad_norm": 0.42213505506515503,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 3515
+    },
+    {
+      "epoch": 0.4084809758931165,
+      "grad_norm": 0.382960706949234,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 3516
+    },
+    {
+      "epoch": 0.408597153645077,
+      "grad_norm": 0.4227701723575592,
+      "learning_rate": 0.0001,
+      "loss": 1.7499,
+      "step": 3517
+    },
+    {
+      "epoch": 0.40871333139703747,
+      "grad_norm": 0.4224139451980591,
+      "learning_rate": 0.0001,
+      "loss": 1.6512,
+      "step": 3518
+    },
+    {
+      "epoch": 0.40882950914899796,
+      "grad_norm": 0.3891499638557434,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 3519
+    },
+    {
+      "epoch": 0.40894568690095845,
+      "grad_norm": 0.40314075350761414,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 3520
+    },
+    {
+      "epoch": 0.40906186465291894,
+      "grad_norm": 0.3946097791194916,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 3521
+    },
+    {
+      "epoch": 0.4091780424048795,
+      "grad_norm": 0.41073909401893616,
+      "learning_rate": 0.0001,
+      "loss": 1.7195,
+      "step": 3522
+    },
+    {
+      "epoch": 0.40929422015683997,
+      "grad_norm": 0.3862104117870331,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 3523
+    },
+    {
+      "epoch": 0.40941039790880046,
+      "grad_norm": 0.38965824246406555,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 3524
+    },
+    {
+      "epoch": 0.40952657566076095,
+      "grad_norm": 0.4076133370399475,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 3525
+    },
+    {
+      "epoch": 0.40964275341272144,
+      "grad_norm": 0.39782625436782837,
+      "learning_rate": 0.0001,
+      "loss": 1.6744,
+      "step": 3526
+    },
+    {
+      "epoch": 0.409758931164682,
+      "grad_norm": 0.4191223680973053,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 3527
+    },
+    {
+      "epoch": 0.40987510891664247,
+      "grad_norm": 0.41905683279037476,
+      "learning_rate": 0.0001,
+      "loss": 1.7366,
+      "step": 3528
+    },
+    {
+      "epoch": 0.40999128666860296,
+      "grad_norm": 0.47418078780174255,
+      "learning_rate": 0.0001,
+      "loss": 1.8641,
+      "step": 3529
+    },
+    {
+      "epoch": 0.41010746442056345,
+      "grad_norm": 0.4219752252101898,
+      "learning_rate": 0.0001,
+      "loss": 1.773,
+      "step": 3530
+    },
+    {
+      "epoch": 0.41022364217252394,
+      "grad_norm": 0.39003801345825195,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 3531
+    },
+    {
+      "epoch": 0.4103398199244845,
+      "grad_norm": 0.39231449365615845,
+      "learning_rate": 0.0001,
+      "loss": 1.7753,
+      "step": 3532
+    },
+    {
+      "epoch": 0.410455997676445,
+      "grad_norm": 0.39529526233673096,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 3533
+    },
+    {
+      "epoch": 0.41057217542840546,
+      "grad_norm": 0.4106643497943878,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 3534
+    },
+    {
+      "epoch": 0.41068835318036595,
+      "grad_norm": 0.3763425648212433,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 3535
+    },
+    {
+      "epoch": 0.41080453093232644,
+      "grad_norm": 0.4354191720485687,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 3536
+    },
+    {
+      "epoch": 0.410920708684287,
+      "grad_norm": 0.4091987609863281,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 3537
+    },
+    {
+      "epoch": 0.4110368864362475,
+      "grad_norm": 0.3786587417125702,
+      "learning_rate": 0.0001,
+      "loss": 1.617,
+      "step": 3538
+    },
+    {
+      "epoch": 0.41115306418820796,
+      "grad_norm": 0.3957653343677521,
+      "learning_rate": 0.0001,
+      "loss": 1.6741,
+      "step": 3539
+    },
+    {
+      "epoch": 0.41126924194016845,
+      "grad_norm": 0.41153082251548767,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 3540
+    },
+    {
+      "epoch": 0.41138541969212894,
+      "grad_norm": 0.408596932888031,
+      "learning_rate": 0.0001,
+      "loss": 1.6669,
+      "step": 3541
+    },
+    {
+      "epoch": 0.4115015974440895,
+      "grad_norm": 0.4165953993797302,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 3542
+    },
+    {
+      "epoch": 0.41161777519605,
+      "grad_norm": 0.4159640371799469,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 3543
+    },
+    {
+      "epoch": 0.41173395294801046,
+      "grad_norm": 0.4020317494869232,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 3544
+    },
+    {
+      "epoch": 0.41185013069997095,
+      "grad_norm": 0.40720680356025696,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 3545
+    },
+    {
+      "epoch": 0.41196630845193144,
+      "grad_norm": 0.3976491689682007,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 3546
+    },
+    {
+      "epoch": 0.41208248620389193,
+      "grad_norm": 0.40788426995277405,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 3547
+    },
+    {
+      "epoch": 0.4121986639558525,
+      "grad_norm": 0.43293899297714233,
+      "learning_rate": 0.0001,
+      "loss": 1.7645,
+      "step": 3548
+    },
+    {
+      "epoch": 0.41231484170781296,
+      "grad_norm": 0.4064108729362488,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 3549
+    },
+    {
+      "epoch": 0.41243101945977345,
+      "grad_norm": 0.38206931948661804,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 3550
+    },
+    {
+      "epoch": 0.41254719721173394,
+      "grad_norm": 0.41803139448165894,
+      "learning_rate": 0.0001,
+      "loss": 1.7538,
+      "step": 3551
+    },
+    {
+      "epoch": 0.41266337496369443,
+      "grad_norm": 0.4098378121852875,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 3552
+    },
+    {
+      "epoch": 0.412779552715655,
+      "grad_norm": 0.40097564458847046,
+      "learning_rate": 0.0001,
+      "loss": 1.7684,
+      "step": 3553
+    },
+    {
+      "epoch": 0.41289573046761546,
+      "grad_norm": 0.4043814241886139,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 3554
+    },
+    {
+      "epoch": 0.41301190821957595,
+      "grad_norm": 0.41799452900886536,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 3555
+    },
+    {
+      "epoch": 0.41312808597153644,
+      "grad_norm": 0.41283276677131653,
+      "learning_rate": 0.0001,
+      "loss": 1.6325,
+      "step": 3556
+    },
+    {
+      "epoch": 0.41324426372349693,
+      "grad_norm": 0.3751561641693115,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 3557
+    },
+    {
+      "epoch": 0.4133604414754575,
+      "grad_norm": 0.4034547805786133,
+      "learning_rate": 0.0001,
+      "loss": 1.6932,
+      "step": 3558
+    },
+    {
+      "epoch": 0.41347661922741796,
+      "grad_norm": 0.3995891511440277,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 3559
+    },
+    {
+      "epoch": 0.41359279697937845,
+      "grad_norm": 0.4040803909301758,
+      "learning_rate": 0.0001,
+      "loss": 1.6819,
+      "step": 3560
+    },
+    {
+      "epoch": 0.41370897473133894,
+      "grad_norm": 0.3844342827796936,
+      "learning_rate": 0.0001,
+      "loss": 1.8967,
+      "step": 3561
+    },
+    {
+      "epoch": 0.41382515248329943,
+      "grad_norm": 0.41226911544799805,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 3562
+    },
+    {
+      "epoch": 0.41394133023526,
+      "grad_norm": 0.4063775837421417,
+      "learning_rate": 0.0001,
+      "loss": 1.7052,
+      "step": 3563
+    },
+    {
+      "epoch": 0.41405750798722046,
+      "grad_norm": 0.3665454089641571,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 3564
+    },
+    {
+      "epoch": 0.41417368573918095,
+      "grad_norm": 0.3731880784034729,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 3565
+    },
+    {
+      "epoch": 0.41428986349114144,
+      "grad_norm": 0.4028816223144531,
+      "learning_rate": 0.0001,
+      "loss": 1.602,
+      "step": 3566
+    },
+    {
+      "epoch": 0.41440604124310193,
+      "grad_norm": 0.39572829008102417,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 3567
+    },
+    {
+      "epoch": 0.4145222189950624,
+      "grad_norm": 0.3968917727470398,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 3568
+    },
+    {
+      "epoch": 0.41463839674702296,
+      "grad_norm": 0.42275354266166687,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 3569
+    },
+    {
+      "epoch": 0.41475457449898345,
+      "grad_norm": 0.4133806526660919,
+      "learning_rate": 0.0001,
+      "loss": 1.7573,
+      "step": 3570
+    },
+    {
+      "epoch": 0.41487075225094394,
+      "grad_norm": 0.4026901125907898,
+      "learning_rate": 0.0001,
+      "loss": 1.7445,
+      "step": 3571
+    },
+    {
+      "epoch": 0.41498693000290443,
+      "grad_norm": 0.40076354146003723,
+      "learning_rate": 0.0001,
+      "loss": 1.6168,
+      "step": 3572
+    },
+    {
+      "epoch": 0.4151031077548649,
+      "grad_norm": 0.3952670991420746,
+      "learning_rate": 0.0001,
+      "loss": 1.7818,
+      "step": 3573
+    },
+    {
+      "epoch": 0.41521928550682546,
+      "grad_norm": 0.396936297416687,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 3574
+    },
+    {
+      "epoch": 0.41533546325878595,
+      "grad_norm": 0.38129517436027527,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 3575
+    },
+    {
+      "epoch": 0.41545164101074644,
+      "grad_norm": 0.42438849806785583,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 3576
+    },
+    {
+      "epoch": 0.41556781876270693,
+      "grad_norm": 0.41612252593040466,
+      "learning_rate": 0.0001,
+      "loss": 1.8019,
+      "step": 3577
+    },
+    {
+      "epoch": 0.4156839965146674,
+      "grad_norm": 0.4326874017715454,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 3578
+    },
+    {
+      "epoch": 0.41580017426662796,
+      "grad_norm": 0.4148086905479431,
+      "learning_rate": 0.0001,
+      "loss": 1.6639,
+      "step": 3579
+    },
+    {
+      "epoch": 0.41591635201858845,
+      "grad_norm": 0.4372271001338959,
+      "learning_rate": 0.0001,
+      "loss": 1.8974,
+      "step": 3580
+    },
+    {
+      "epoch": 0.41603252977054894,
+      "grad_norm": 0.42456239461898804,
+      "learning_rate": 0.0001,
+      "loss": 1.8067,
+      "step": 3581
+    },
+    {
+      "epoch": 0.41614870752250943,
+      "grad_norm": 0.3889367878437042,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 3582
+    },
+    {
+      "epoch": 0.4162648852744699,
+      "grad_norm": 0.4123779833316803,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 3583
+    },
+    {
+      "epoch": 0.41638106302643046,
+      "grad_norm": 0.45006388425827026,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 3584
+    },
+    {
+      "epoch": 0.41649724077839095,
+      "grad_norm": 0.3993321359157562,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 3585
+    },
+    {
+      "epoch": 0.41661341853035144,
+      "grad_norm": 0.39892643690109253,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 3586
+    },
+    {
+      "epoch": 0.41672959628231193,
+      "grad_norm": 0.4181758463382721,
+      "learning_rate": 0.0001,
+      "loss": 1.7197,
+      "step": 3587
+    },
+    {
+      "epoch": 0.4168457740342724,
+      "grad_norm": 0.3975769877433777,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 3588
+    },
+    {
+      "epoch": 0.4169619517862329,
+      "grad_norm": 0.3831265866756439,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 3589
+    },
+    {
+      "epoch": 0.41707812953819345,
+      "grad_norm": 0.41471192240715027,
+      "learning_rate": 0.0001,
+      "loss": 1.7917,
+      "step": 3590
+    },
+    {
+      "epoch": 0.41719430729015394,
+      "grad_norm": 0.3869475722312927,
+      "learning_rate": 0.0001,
+      "loss": 1.7876,
+      "step": 3591
+    },
+    {
+      "epoch": 0.41731048504211443,
+      "grad_norm": 0.37546539306640625,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 3592
+    },
+    {
+      "epoch": 0.4174266627940749,
+      "grad_norm": 0.38693931698799133,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 3593
+    },
+    {
+      "epoch": 0.4175428405460354,
+      "grad_norm": 0.40979790687561035,
+      "learning_rate": 0.0001,
+      "loss": 1.7553,
+      "step": 3594
+    },
+    {
+      "epoch": 0.41765901829799595,
+      "grad_norm": 0.4143114686012268,
+      "learning_rate": 0.0001,
+      "loss": 1.875,
+      "step": 3595
+    },
+    {
+      "epoch": 0.41777519604995644,
+      "grad_norm": 0.4477331340312958,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 3596
+    },
+    {
+      "epoch": 0.41789137380191693,
+      "grad_norm": 0.4350159168243408,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 3597
+    },
+    {
+      "epoch": 0.4180075515538774,
+      "grad_norm": 0.40103811025619507,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 3598
+    },
+    {
+      "epoch": 0.4181237293058379,
+      "grad_norm": 0.42773351073265076,
+      "learning_rate": 0.0001,
+      "loss": 1.8818,
+      "step": 3599
+    },
+    {
+      "epoch": 0.41823990705779845,
+      "grad_norm": 0.3894450068473816,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 3600
+    },
+    {
+      "epoch": 0.41835608480975894,
+      "grad_norm": 0.3698801100254059,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 3601
+    },
+    {
+      "epoch": 0.41847226256171943,
+      "grad_norm": 0.3938926160335541,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 3602
+    },
+    {
+      "epoch": 0.4185884403136799,
+      "grad_norm": 0.40313947200775146,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 3603
+    },
+    {
+      "epoch": 0.4187046180656404,
+      "grad_norm": 0.4046363830566406,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 3604
+    },
+    {
+      "epoch": 0.41882079581760095,
+      "grad_norm": 0.4111688435077667,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 3605
+    },
+    {
+      "epoch": 0.41893697356956144,
+      "grad_norm": 0.41641494631767273,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 3606
+    },
+    {
+      "epoch": 0.41905315132152193,
+      "grad_norm": 0.3970726728439331,
+      "learning_rate": 0.0001,
+      "loss": 1.6719,
+      "step": 3607
+    },
+    {
+      "epoch": 0.4191693290734824,
+      "grad_norm": 0.40690597891807556,
+      "learning_rate": 0.0001,
+      "loss": 1.7636,
+      "step": 3608
+    },
+    {
+      "epoch": 0.4192855068254429,
+      "grad_norm": 0.41410157084465027,
+      "learning_rate": 0.0001,
+      "loss": 1.794,
+      "step": 3609
+    },
+    {
+      "epoch": 0.4194016845774034,
+      "grad_norm": 0.38639214634895325,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 3610
+    },
+    {
+      "epoch": 0.41951786232936394,
+      "grad_norm": 0.3689410090446472,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 3611
+    },
+    {
+      "epoch": 0.41963404008132443,
+      "grad_norm": 0.40414369106292725,
+      "learning_rate": 0.0001,
+      "loss": 1.8223,
+      "step": 3612
+    },
+    {
+      "epoch": 0.4197502178332849,
+      "grad_norm": 0.38138678669929504,
+      "learning_rate": 0.0001,
+      "loss": 1.3491,
+      "step": 3613
+    },
+    {
+      "epoch": 0.4198663955852454,
+      "grad_norm": 0.37447166442871094,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 3614
+    },
+    {
+      "epoch": 0.4199825733372059,
+      "grad_norm": 0.40835610032081604,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 3615
+    },
+    {
+      "epoch": 0.42009875108916644,
+      "grad_norm": 0.4299129545688629,
+      "learning_rate": 0.0001,
+      "loss": 1.7448,
+      "step": 3616
+    },
+    {
+      "epoch": 0.42021492884112693,
+      "grad_norm": 0.3932313621044159,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 3617
+    },
+    {
+      "epoch": 0.4203311065930874,
+      "grad_norm": 0.3847653865814209,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 3618
+    },
+    {
+      "epoch": 0.4204472843450479,
+      "grad_norm": 0.435201495885849,
+      "learning_rate": 0.0001,
+      "loss": 1.8684,
+      "step": 3619
+    },
+    {
+      "epoch": 0.4205634620970084,
+      "grad_norm": 0.3746712803840637,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 3620
+    },
+    {
+      "epoch": 0.42067963984896894,
+      "grad_norm": 0.37663495540618896,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 3621
+    },
+    {
+      "epoch": 0.42079581760092943,
+      "grad_norm": 0.37208837270736694,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 3622
+    },
+    {
+      "epoch": 0.4209119953528899,
+      "grad_norm": 0.3866986334323883,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 3623
+    },
+    {
+      "epoch": 0.4210281731048504,
+      "grad_norm": 0.40129542350769043,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 3624
+    },
+    {
+      "epoch": 0.4211443508568109,
+      "grad_norm": 0.38859081268310547,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 3625
+    },
+    {
+      "epoch": 0.42126052860877145,
+      "grad_norm": 0.43616414070129395,
+      "learning_rate": 0.0001,
+      "loss": 1.7725,
+      "step": 3626
+    },
+    {
+      "epoch": 0.42137670636073193,
+      "grad_norm": 0.43202710151672363,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 3627
+    },
+    {
+      "epoch": 0.4214928841126924,
+      "grad_norm": 0.39463871717453003,
+      "learning_rate": 0.0001,
+      "loss": 1.4335,
+      "step": 3628
+    },
+    {
+      "epoch": 0.4216090618646529,
+      "grad_norm": 0.4237414002418518,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 3629
+    },
+    {
+      "epoch": 0.4217252396166134,
+      "grad_norm": 0.39005210995674133,
+      "learning_rate": 0.0001,
+      "loss": 1.48,
+      "step": 3630
+    },
+    {
+      "epoch": 0.42184141736857395,
+      "grad_norm": 0.3883463144302368,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 3631
+    },
+    {
+      "epoch": 0.42195759512053443,
+      "grad_norm": 0.3896929621696472,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 3632
+    },
+    {
+      "epoch": 0.4220737728724949,
+      "grad_norm": 0.41953974962234497,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 3633
+    },
+    {
+      "epoch": 0.4221899506244554,
+      "grad_norm": 0.3946000039577484,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 3634
+    },
+    {
+      "epoch": 0.4223061283764159,
+      "grad_norm": 0.4022957384586334,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 3635
+    },
+    {
+      "epoch": 0.4224223061283764,
+      "grad_norm": 0.40276673436164856,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 3636
+    },
+    {
+      "epoch": 0.42253848388033693,
+      "grad_norm": 0.41087186336517334,
+      "learning_rate": 0.0001,
+      "loss": 1.7919,
+      "step": 3637
+    },
+    {
+      "epoch": 0.4226546616322974,
+      "grad_norm": 0.4109443426132202,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 3638
+    },
+    {
+      "epoch": 0.4227708393842579,
+      "grad_norm": 0.3984861671924591,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 3639
+    },
+    {
+      "epoch": 0.4228870171362184,
+      "grad_norm": 0.42480790615081787,
+      "learning_rate": 0.0001,
+      "loss": 1.7443,
+      "step": 3640
+    },
+    {
+      "epoch": 0.4230031948881789,
+      "grad_norm": 0.37020301818847656,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 3641
+    },
+    {
+      "epoch": 0.42311937264013944,
+      "grad_norm": 0.3878079950809479,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 3642
+    },
+    {
+      "epoch": 0.4232355503920999,
+      "grad_norm": 0.4088906943798065,
+      "learning_rate": 0.0001,
+      "loss": 1.7003,
+      "step": 3643
+    },
+    {
+      "epoch": 0.4233517281440604,
+      "grad_norm": 0.39940375089645386,
+      "learning_rate": 0.0001,
+      "loss": 1.673,
+      "step": 3644
+    },
+    {
+      "epoch": 0.4234679058960209,
+      "grad_norm": 0.41945314407348633,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 3645
+    },
+    {
+      "epoch": 0.4235840836479814,
+      "grad_norm": 0.43986397981643677,
+      "learning_rate": 0.0001,
+      "loss": 1.7769,
+      "step": 3646
+    },
+    {
+      "epoch": 0.42370026139994194,
+      "grad_norm": 0.41708502173423767,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 3647
+    },
+    {
+      "epoch": 0.4238164391519024,
+      "grad_norm": 0.3953828513622284,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 3648
+    },
+    {
+      "epoch": 0.4239326169038629,
+      "grad_norm": 0.3876919746398926,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 3649
+    },
+    {
+      "epoch": 0.4240487946558234,
+      "grad_norm": 0.4110132157802582,
+      "learning_rate": 0.0001,
+      "loss": 1.8118,
+      "step": 3650
+    },
+    {
+      "epoch": 0.4241649724077839,
+      "grad_norm": 0.4448528587818146,
+      "learning_rate": 0.0001,
+      "loss": 1.8269,
+      "step": 3651
+    },
+    {
+      "epoch": 0.42428115015974444,
+      "grad_norm": 0.4393356740474701,
+      "learning_rate": 0.0001,
+      "loss": 1.8647,
+      "step": 3652
+    },
+    {
+      "epoch": 0.4243973279117049,
+      "grad_norm": 0.4252372086048126,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 3653
+    },
+    {
+      "epoch": 0.4245135056636654,
+      "grad_norm": 0.41888418793678284,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 3654
+    },
+    {
+      "epoch": 0.4246296834156259,
+      "grad_norm": 0.39156967401504517,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 3655
+    },
+    {
+      "epoch": 0.4247458611675864,
+      "grad_norm": 0.3865427076816559,
+      "learning_rate": 0.0001,
+      "loss": 1.6541,
+      "step": 3656
+    },
+    {
+      "epoch": 0.4248620389195469,
+      "grad_norm": 0.37357524037361145,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 3657
+    },
+    {
+      "epoch": 0.4249782166715074,
+      "grad_norm": 0.3771733045578003,
+      "learning_rate": 0.0001,
+      "loss": 1.59,
+      "step": 3658
+    },
+    {
+      "epoch": 0.4250943944234679,
+      "grad_norm": 0.40311431884765625,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 3659
+    },
+    {
+      "epoch": 0.4252105721754284,
+      "grad_norm": 0.39886149764060974,
+      "learning_rate": 0.0001,
+      "loss": 1.8507,
+      "step": 3660
+    },
+    {
+      "epoch": 0.4253267499273889,
+      "grad_norm": 0.4184694290161133,
+      "learning_rate": 0.0001,
+      "loss": 1.7392,
+      "step": 3661
+    },
+    {
+      "epoch": 0.4254429276793494,
+      "grad_norm": 0.4278115928173065,
+      "learning_rate": 0.0001,
+      "loss": 1.8403,
+      "step": 3662
+    },
+    {
+      "epoch": 0.4255591054313099,
+      "grad_norm": 0.4227597117424011,
+      "learning_rate": 0.0001,
+      "loss": 1.7257,
+      "step": 3663
+    },
+    {
+      "epoch": 0.4256752831832704,
+      "grad_norm": 0.3717440068721771,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 3664
+    },
+    {
+      "epoch": 0.4257914609352309,
+      "grad_norm": 0.3993324935436249,
+      "learning_rate": 0.0001,
+      "loss": 1.6323,
+      "step": 3665
+    },
+    {
+      "epoch": 0.4259076386871914,
+      "grad_norm": 0.40730348229408264,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 3666
+    },
+    {
+      "epoch": 0.4260238164391519,
+      "grad_norm": 0.38585323095321655,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 3667
+    },
+    {
+      "epoch": 0.4261399941911124,
+      "grad_norm": 0.4284312129020691,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 3668
+    },
+    {
+      "epoch": 0.4262561719430729,
+      "grad_norm": 0.417501300573349,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 3669
+    },
+    {
+      "epoch": 0.4263723496950334,
+      "grad_norm": 0.3896951973438263,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 3670
+    },
+    {
+      "epoch": 0.4264885274469939,
+      "grad_norm": 0.4131647050380707,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 3671
+    },
+    {
+      "epoch": 0.4266047051989544,
+      "grad_norm": 0.36741092801094055,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 3672
+    },
+    {
+      "epoch": 0.4267208829509149,
+      "grad_norm": 0.41601473093032837,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 3673
+    },
+    {
+      "epoch": 0.4268370607028754,
+      "grad_norm": 0.41685929894447327,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 3674
+    },
+    {
+      "epoch": 0.4269532384548359,
+      "grad_norm": 0.41353273391723633,
+      "learning_rate": 0.0001,
+      "loss": 1.7881,
+      "step": 3675
+    },
+    {
+      "epoch": 0.4270694162067964,
+      "grad_norm": 0.40456423163414,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 3676
+    },
+    {
+      "epoch": 0.4271855939587569,
+      "grad_norm": 0.457220196723938,
+      "learning_rate": 0.0001,
+      "loss": 1.7856,
+      "step": 3677
+    },
+    {
+      "epoch": 0.42730177171071737,
+      "grad_norm": 0.4161781668663025,
+      "learning_rate": 0.0001,
+      "loss": 1.6692,
+      "step": 3678
+    },
+    {
+      "epoch": 0.4274179494626779,
+      "grad_norm": 0.40317872166633606,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 3679
+    },
+    {
+      "epoch": 0.4275341272146384,
+      "grad_norm": 0.4100804328918457,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 3680
+    },
+    {
+      "epoch": 0.4276503049665989,
+      "grad_norm": 0.40741920471191406,
+      "learning_rate": 0.0001,
+      "loss": 1.6931,
+      "step": 3681
+    },
+    {
+      "epoch": 0.4277664827185594,
+      "grad_norm": 0.3968175947666168,
+      "learning_rate": 0.0001,
+      "loss": 1.5824,
+      "step": 3682
+    },
+    {
+      "epoch": 0.42788266047051987,
+      "grad_norm": 0.3978244662284851,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 3683
+    },
+    {
+      "epoch": 0.4279988382224804,
+      "grad_norm": 0.39147892594337463,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 3684
+    },
+    {
+      "epoch": 0.4281150159744409,
+      "grad_norm": 0.410375714302063,
+      "learning_rate": 0.0001,
+      "loss": 1.7164,
+      "step": 3685
+    },
+    {
+      "epoch": 0.4282311937264014,
+      "grad_norm": 0.4207121729850769,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 3686
+    },
+    {
+      "epoch": 0.4283473714783619,
+      "grad_norm": 0.446114718914032,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 3687
+    },
+    {
+      "epoch": 0.42846354923032237,
+      "grad_norm": 0.429855614900589,
+      "learning_rate": 0.0001,
+      "loss": 1.8145,
+      "step": 3688
+    },
+    {
+      "epoch": 0.4285797269822829,
+      "grad_norm": 0.4075881838798523,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 3689
+    },
+    {
+      "epoch": 0.4286959047342434,
+      "grad_norm": 0.42212194204330444,
+      "learning_rate": 0.0001,
+      "loss": 1.8045,
+      "step": 3690
+    },
+    {
+      "epoch": 0.4288120824862039,
+      "grad_norm": 0.3913438022136688,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 3691
+    },
+    {
+      "epoch": 0.4289282602381644,
+      "grad_norm": 0.4033333659172058,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 3692
+    },
+    {
+      "epoch": 0.42904443799012487,
+      "grad_norm": 0.41124168038368225,
+      "learning_rate": 0.0001,
+      "loss": 1.7919,
+      "step": 3693
+    },
+    {
+      "epoch": 0.4291606157420854,
+      "grad_norm": 0.4082317352294922,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 3694
+    },
+    {
+      "epoch": 0.4292767934940459,
+      "grad_norm": 0.43243342638015747,
+      "learning_rate": 0.0001,
+      "loss": 1.8346,
+      "step": 3695
+    },
+    {
+      "epoch": 0.4293929712460064,
+      "grad_norm": 0.4029131829738617,
+      "learning_rate": 0.0001,
+      "loss": 1.6845,
+      "step": 3696
+    },
+    {
+      "epoch": 0.4295091489979669,
+      "grad_norm": 0.39119669795036316,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 3697
+    },
+    {
+      "epoch": 0.42962532674992737,
+      "grad_norm": 0.4251865744590759,
+      "learning_rate": 0.0001,
+      "loss": 1.7231,
+      "step": 3698
+    },
+    {
+      "epoch": 0.42974150450188786,
+      "grad_norm": 0.41602757573127747,
+      "learning_rate": 0.0001,
+      "loss": 1.6948,
+      "step": 3699
+    },
+    {
+      "epoch": 0.4298576822538484,
+      "grad_norm": 0.41048288345336914,
+      "learning_rate": 0.0001,
+      "loss": 1.7455,
+      "step": 3700
+    },
+    {
+      "epoch": 0.4299738600058089,
+      "grad_norm": 0.39157551527023315,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 3701
+    },
+    {
+      "epoch": 0.4300900377577694,
+      "grad_norm": 0.40497899055480957,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 3702
+    },
+    {
+      "epoch": 0.4302062155097299,
+      "grad_norm": 0.36876180768013,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 3703
+    },
+    {
+      "epoch": 0.43032239326169036,
+      "grad_norm": 0.39832401275634766,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 3704
+    },
+    {
+      "epoch": 0.4304385710136509,
+      "grad_norm": 0.40576431155204773,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 3705
+    },
+    {
+      "epoch": 0.4305547487656114,
+      "grad_norm": 0.41582056879997253,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 3706
+    },
+    {
+      "epoch": 0.4306709265175719,
+      "grad_norm": 0.37532898783683777,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 3707
+    },
+    {
+      "epoch": 0.4307871042695324,
+      "grad_norm": 0.4363482594490051,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 3708
+    },
+    {
+      "epoch": 0.43090328202149286,
+      "grad_norm": 0.38650989532470703,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 3709
+    },
+    {
+      "epoch": 0.4310194597734534,
+      "grad_norm": 0.4108797013759613,
+      "learning_rate": 0.0001,
+      "loss": 1.8012,
+      "step": 3710
+    },
+    {
+      "epoch": 0.4311356375254139,
+      "grad_norm": 0.3996245563030243,
+      "learning_rate": 0.0001,
+      "loss": 1.7163,
+      "step": 3711
+    },
+    {
+      "epoch": 0.4312518152773744,
+      "grad_norm": 0.40178465843200684,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 3712
+    },
+    {
+      "epoch": 0.4313679930293349,
+      "grad_norm": 0.38844192028045654,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 3713
+    },
+    {
+      "epoch": 0.43148417078129536,
+      "grad_norm": 0.38284438848495483,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 3714
+    },
+    {
+      "epoch": 0.4316003485332559,
+      "grad_norm": 0.4122409224510193,
+      "learning_rate": 0.0001,
+      "loss": 1.7502,
+      "step": 3715
+    },
+    {
+      "epoch": 0.4317165262852164,
+      "grad_norm": 0.4218509793281555,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 3716
+    },
+    {
+      "epoch": 0.4318327040371769,
+      "grad_norm": 0.4130255877971649,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 3717
+    },
+    {
+      "epoch": 0.4319488817891374,
+      "grad_norm": 0.4026492238044739,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 3718
+    },
+    {
+      "epoch": 0.43206505954109786,
+      "grad_norm": 0.4177059531211853,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 3719
+    },
+    {
+      "epoch": 0.4321812372930584,
+      "grad_norm": 0.4138438105583191,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 3720
+    },
+    {
+      "epoch": 0.4322974150450189,
+      "grad_norm": 0.4228832721710205,
+      "learning_rate": 0.0001,
+      "loss": 1.8092,
+      "step": 3721
+    },
+    {
+      "epoch": 0.4324135927969794,
+      "grad_norm": 0.4101725220680237,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 3722
+    },
+    {
+      "epoch": 0.4325297705489399,
+      "grad_norm": 0.38750308752059937,
+      "learning_rate": 0.0001,
+      "loss": 1.4348,
+      "step": 3723
+    },
+    {
+      "epoch": 0.43264594830090036,
+      "grad_norm": 0.43653422594070435,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 3724
+    },
+    {
+      "epoch": 0.43276212605286085,
+      "grad_norm": 0.39200592041015625,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 3725
+    },
+    {
+      "epoch": 0.4328783038048214,
+      "grad_norm": 0.40859901905059814,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 3726
+    },
+    {
+      "epoch": 0.4329944815567819,
+      "grad_norm": 0.41595181822776794,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 3727
+    },
+    {
+      "epoch": 0.4331106593087424,
+      "grad_norm": 0.391353577375412,
+      "learning_rate": 0.0001,
+      "loss": 1.5171,
+      "step": 3728
+    },
+    {
+      "epoch": 0.43322683706070286,
+      "grad_norm": 0.40748852491378784,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 3729
+    },
+    {
+      "epoch": 0.43334301481266335,
+      "grad_norm": 0.3781512975692749,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 3730
+    },
+    {
+      "epoch": 0.4334591925646239,
+      "grad_norm": 0.36677882075309753,
+      "learning_rate": 0.0001,
+      "loss": 1.3444,
+      "step": 3731
+    },
+    {
+      "epoch": 0.4335753703165844,
+      "grad_norm": 0.38588741421699524,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 3732
+    },
+    {
+      "epoch": 0.4336915480685449,
+      "grad_norm": 0.4158620834350586,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 3733
+    },
+    {
+      "epoch": 0.43380772582050536,
+      "grad_norm": 0.4044388234615326,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 3734
+    },
+    {
+      "epoch": 0.43392390357246585,
+      "grad_norm": 0.4197262227535248,
+      "learning_rate": 0.0001,
+      "loss": 1.7444,
+      "step": 3735
+    },
+    {
+      "epoch": 0.4340400813244264,
+      "grad_norm": 0.40574586391448975,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 3736
+    },
+    {
+      "epoch": 0.4341562590763869,
+      "grad_norm": 0.397217333316803,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 3737
+    },
+    {
+      "epoch": 0.4342724368283474,
+      "grad_norm": 0.4184318482875824,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 3738
+    },
+    {
+      "epoch": 0.43438861458030786,
+      "grad_norm": 0.3860454857349396,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 3739
+    },
+    {
+      "epoch": 0.43450479233226835,
+      "grad_norm": 0.41592180728912354,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 3740
+    },
+    {
+      "epoch": 0.4346209700842289,
+      "grad_norm": 0.3821113407611847,
+      "learning_rate": 0.0001,
+      "loss": 1.5594,
+      "step": 3741
+    },
+    {
+      "epoch": 0.4347371478361894,
+      "grad_norm": 0.4071682095527649,
+      "learning_rate": 0.0001,
+      "loss": 1.5585,
+      "step": 3742
+    },
+    {
+      "epoch": 0.4348533255881499,
+      "grad_norm": 0.41301366686820984,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 3743
+    },
+    {
+      "epoch": 0.43496950334011036,
+      "grad_norm": 0.4541078209877014,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 3744
+    },
+    {
+      "epoch": 0.43508568109207085,
+      "grad_norm": 0.4197753369808197,
+      "learning_rate": 0.0001,
+      "loss": 1.7419,
+      "step": 3745
+    },
+    {
+      "epoch": 0.43520185884403134,
+      "grad_norm": 0.40982601046562195,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 3746
+    },
+    {
+      "epoch": 0.4353180365959919,
+      "grad_norm": 0.439325213432312,
+      "learning_rate": 0.0001,
+      "loss": 1.8748,
+      "step": 3747
+    },
+    {
+      "epoch": 0.4354342143479524,
+      "grad_norm": 0.4326147139072418,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 3748
+    },
+    {
+      "epoch": 0.43555039209991286,
+      "grad_norm": 0.43868646025657654,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 3749
+    },
+    {
+      "epoch": 0.43566656985187335,
+      "grad_norm": 0.4201911389827728,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 3750
+    },
+    {
+      "epoch": 0.43578274760383384,
+      "grad_norm": 0.3979409337043762,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 3751
+    },
+    {
+      "epoch": 0.4358989253557944,
+      "grad_norm": 0.4207451045513153,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 3752
+    },
+    {
+      "epoch": 0.4360151031077549,
+      "grad_norm": 0.3865394592285156,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 3753
+    },
+    {
+      "epoch": 0.43613128085971536,
+      "grad_norm": 0.4036276340484619,
+      "learning_rate": 0.0001,
+      "loss": 1.7083,
+      "step": 3754
+    },
+    {
+      "epoch": 0.43624745861167585,
+      "grad_norm": 0.39419230818748474,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 3755
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 0.42213842272758484,
+      "learning_rate": 0.0001,
+      "loss": 1.7595,
+      "step": 3756
+    },
+    {
+      "epoch": 0.4364798141155969,
+      "grad_norm": 0.4202689528465271,
+      "learning_rate": 0.0001,
+      "loss": 1.7564,
+      "step": 3757
+    },
+    {
+      "epoch": 0.4365959918675574,
+      "grad_norm": 0.40122315287590027,
+      "learning_rate": 0.0001,
+      "loss": 1.7285,
+      "step": 3758
+    },
+    {
+      "epoch": 0.43671216961951786,
+      "grad_norm": 0.41420555114746094,
+      "learning_rate": 0.0001,
+      "loss": 1.7936,
+      "step": 3759
+    },
+    {
+      "epoch": 0.43682834737147835,
+      "grad_norm": 0.46364185214042664,
+      "learning_rate": 0.0001,
+      "loss": 1.7911,
+      "step": 3760
+    },
+    {
+      "epoch": 0.43694452512343884,
+      "grad_norm": 0.404472678899765,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4370607028753994,
+      "grad_norm": 0.4069894552230835,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4371768806273599,
+      "grad_norm": 0.41052114963531494,
+      "learning_rate": 0.0001,
+      "loss": 1.5779,
+      "step": 3763
+    },
+    {
+      "epoch": 0.43729305837932037,
+      "grad_norm": 0.41978520154953003,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 3764
+    },
+    {
+      "epoch": 0.43740923613128085,
+      "grad_norm": 0.45879876613616943,
+      "learning_rate": 0.0001,
+      "loss": 1.7743,
+      "step": 3765
+    },
+    {
+      "epoch": 0.43752541388324134,
+      "grad_norm": 0.42506903409957886,
+      "learning_rate": 0.0001,
+      "loss": 1.706,
+      "step": 3766
+    },
+    {
+      "epoch": 0.43764159163520183,
+      "grad_norm": 0.4226793944835663,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 3767
+    },
+    {
+      "epoch": 0.4377577693871624,
+      "grad_norm": 0.38549569249153137,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 3768
+    },
+    {
+      "epoch": 0.43787394713912287,
+      "grad_norm": 0.3775811791419983,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 3769
+    },
+    {
+      "epoch": 0.43799012489108335,
+      "grad_norm": 0.4261288046836853,
+      "learning_rate": 0.0001,
+      "loss": 1.7852,
+      "step": 3770
+    },
+    {
+      "epoch": 0.43810630264304384,
+      "grad_norm": 0.4037782847881317,
+      "learning_rate": 0.0001,
+      "loss": 1.7035,
+      "step": 3771
+    },
+    {
+      "epoch": 0.43822248039500433,
+      "grad_norm": 0.3799671530723572,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 3772
+    },
+    {
+      "epoch": 0.4383386581469649,
+      "grad_norm": 0.4238581955432892,
+      "learning_rate": 0.0001,
+      "loss": 1.7204,
+      "step": 3773
+    },
+    {
+      "epoch": 0.43845483589892537,
+      "grad_norm": 0.39733898639678955,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 3774
+    },
+    {
+      "epoch": 0.43857101365088585,
+      "grad_norm": 0.386123925447464,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 3775
+    },
+    {
+      "epoch": 0.43868719140284634,
+      "grad_norm": 0.42280828952789307,
+      "learning_rate": 0.0001,
+      "loss": 1.7637,
+      "step": 3776
+    },
+    {
+      "epoch": 0.43880336915480683,
+      "grad_norm": 0.4024019241333008,
+      "learning_rate": 0.0001,
+      "loss": 1.7998,
+      "step": 3777
+    },
+    {
+      "epoch": 0.4389195469067674,
+      "grad_norm": 0.36499494314193726,
+      "learning_rate": 0.0001,
+      "loss": 1.408,
+      "step": 3778
+    },
+    {
+      "epoch": 0.43903572465872787,
+      "grad_norm": 0.4545857906341553,
+      "learning_rate": 0.0001,
+      "loss": 1.9895,
+      "step": 3779
+    },
+    {
+      "epoch": 0.43915190241068836,
+      "grad_norm": 0.4315713047981262,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 3780
+    },
+    {
+      "epoch": 0.43926808016264884,
+      "grad_norm": 0.4134562015533447,
+      "learning_rate": 0.0001,
+      "loss": 1.6684,
+      "step": 3781
+    },
+    {
+      "epoch": 0.43938425791460933,
+      "grad_norm": 0.39927372336387634,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 3782
+    },
+    {
+      "epoch": 0.4395004356665699,
+      "grad_norm": 0.407842755317688,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 3783
+    },
+    {
+      "epoch": 0.43961661341853037,
+      "grad_norm": 0.4489937126636505,
+      "learning_rate": 0.0001,
+      "loss": 1.7954,
+      "step": 3784
+    },
+    {
+      "epoch": 0.43973279117049086,
+      "grad_norm": 0.41381195187568665,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 3785
+    },
+    {
+      "epoch": 0.43984896892245134,
+      "grad_norm": 0.4299207627773285,
+      "learning_rate": 0.0001,
+      "loss": 1.7795,
+      "step": 3786
+    },
+    {
+      "epoch": 0.43996514667441183,
+      "grad_norm": 0.4247708320617676,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 3787
+    },
+    {
+      "epoch": 0.4400813244263723,
+      "grad_norm": 0.45429563522338867,
+      "learning_rate": 0.0001,
+      "loss": 1.6822,
+      "step": 3788
+    },
+    {
+      "epoch": 0.44019750217833287,
+      "grad_norm": 0.39192894101142883,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 3789
+    },
+    {
+      "epoch": 0.44031367993029336,
+      "grad_norm": 0.4126262664794922,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 3790
+    },
+    {
+      "epoch": 0.44042985768225384,
+      "grad_norm": 0.3963555693626404,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 3791
+    },
+    {
+      "epoch": 0.44054603543421433,
+      "grad_norm": 0.3770107924938202,
+      "learning_rate": 0.0001,
+      "loss": 1.3031,
+      "step": 3792
+    },
+    {
+      "epoch": 0.4406622131861748,
+      "grad_norm": 0.40233081579208374,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 3793
+    },
+    {
+      "epoch": 0.44077839093813537,
+      "grad_norm": 0.4272160232067108,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 3794
+    },
+    {
+      "epoch": 0.44089456869009586,
+      "grad_norm": 0.4221998453140259,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 3795
+    },
+    {
+      "epoch": 0.44101074644205635,
+      "grad_norm": 0.3974875509738922,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 3796
+    },
+    {
+      "epoch": 0.44112692419401683,
+      "grad_norm": 0.462016224861145,
+      "learning_rate": 0.0001,
+      "loss": 1.8508,
+      "step": 3797
+    },
+    {
+      "epoch": 0.4412431019459773,
+      "grad_norm": 0.412485271692276,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 3798
+    },
+    {
+      "epoch": 0.44135927969793787,
+      "grad_norm": 0.40575462579727173,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 3799
+    },
+    {
+      "epoch": 0.44147545744989836,
+      "grad_norm": 0.393636018037796,
+      "learning_rate": 0.0001,
+      "loss": 1.5162,
+      "step": 3800
+    },
+    {
+      "epoch": 0.44159163520185885,
+      "grad_norm": 0.39681223034858704,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 3801
+    },
+    {
+      "epoch": 0.44170781295381933,
+      "grad_norm": 0.41064146161079407,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4418239907057798,
+      "grad_norm": 0.4206371605396271,
+      "learning_rate": 0.0001,
+      "loss": 1.7796,
+      "step": 3803
+    },
+    {
+      "epoch": 0.44194016845774037,
+      "grad_norm": 0.41671186685562134,
+      "learning_rate": 0.0001,
+      "loss": 1.7832,
+      "step": 3804
+    },
+    {
+      "epoch": 0.44205634620970086,
+      "grad_norm": 0.38577428460121155,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 3805
+    },
+    {
+      "epoch": 0.44217252396166135,
+      "grad_norm": 0.3924373388290405,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 3806
+    },
+    {
+      "epoch": 0.44228870171362183,
+      "grad_norm": 0.44141262769699097,
+      "learning_rate": 0.0001,
+      "loss": 1.7508,
+      "step": 3807
+    },
+    {
+      "epoch": 0.4424048794655823,
+      "grad_norm": 0.40393298864364624,
+      "learning_rate": 0.0001,
+      "loss": 1.7571,
+      "step": 3808
+    },
+    {
+      "epoch": 0.44252105721754287,
+      "grad_norm": 0.42056989669799805,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 3809
+    },
+    {
+      "epoch": 0.44263723496950336,
+      "grad_norm": 0.38610607385635376,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 3810
+    },
+    {
+      "epoch": 0.44275341272146385,
+      "grad_norm": 0.408483624458313,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 3811
+    },
+    {
+      "epoch": 0.44286959047342433,
+      "grad_norm": 0.3991870880126953,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 3812
+    },
+    {
+      "epoch": 0.4429857682253848,
+      "grad_norm": 0.4130445718765259,
+      "learning_rate": 0.0001,
+      "loss": 1.7266,
+      "step": 3813
+    },
+    {
+      "epoch": 0.4431019459773453,
+      "grad_norm": 0.4013391137123108,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 3814
+    },
+    {
+      "epoch": 0.44321812372930586,
+      "grad_norm": 0.4077419638633728,
+      "learning_rate": 0.0001,
+      "loss": 1.7261,
+      "step": 3815
+    },
+    {
+      "epoch": 0.44333430148126635,
+      "grad_norm": 0.40052545070648193,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 3816
+    },
+    {
+      "epoch": 0.44345047923322684,
+      "grad_norm": 0.38764074444770813,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4435666569851873,
+      "grad_norm": 0.41951900720596313,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 3818
+    },
+    {
+      "epoch": 0.4436828347371478,
+      "grad_norm": 0.42698657512664795,
+      "learning_rate": 0.0001,
+      "loss": 1.7343,
+      "step": 3819
+    },
+    {
+      "epoch": 0.44379901248910836,
+      "grad_norm": 0.42899012565612793,
+      "learning_rate": 0.0001,
+      "loss": 1.8282,
+      "step": 3820
+    },
+    {
+      "epoch": 0.44391519024106885,
+      "grad_norm": 0.3827013373374939,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 3821
+    },
+    {
+      "epoch": 0.44403136799302934,
+      "grad_norm": 0.42411962151527405,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4441475457449898,
+      "grad_norm": 0.4320610463619232,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 3823
+    },
+    {
+      "epoch": 0.4442637234969503,
+      "grad_norm": 0.4369167685508728,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 3824
+    },
+    {
+      "epoch": 0.44437990124891086,
+      "grad_norm": 0.3931221663951874,
+      "learning_rate": 0.0001,
+      "loss": 1.7523,
+      "step": 3825
+    },
+    {
+      "epoch": 0.44449607900087135,
+      "grad_norm": 0.4219224750995636,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 3826
+    },
+    {
+      "epoch": 0.44461225675283184,
+      "grad_norm": 0.4289875030517578,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 3827
+    },
+    {
+      "epoch": 0.4447284345047923,
+      "grad_norm": 0.3924271762371063,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 3828
+    },
+    {
+      "epoch": 0.4448446122567528,
+      "grad_norm": 0.3892977237701416,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 3829
+    },
+    {
+      "epoch": 0.44496079000871336,
+      "grad_norm": 0.4215613305568695,
+      "learning_rate": 0.0001,
+      "loss": 1.8097,
+      "step": 3830
+    },
+    {
+      "epoch": 0.44507696776067385,
+      "grad_norm": 0.38903629779815674,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 3831
+    },
+    {
+      "epoch": 0.44519314551263434,
+      "grad_norm": 0.3972281217575073,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 3832
+    },
+    {
+      "epoch": 0.4453093232645948,
+      "grad_norm": 0.43121159076690674,
+      "learning_rate": 0.0001,
+      "loss": 1.7917,
+      "step": 3833
+    },
+    {
+      "epoch": 0.4454255010165553,
+      "grad_norm": 0.3963190019130707,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4455416787685158,
+      "grad_norm": 0.4058324992656708,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 3835
+    },
+    {
+      "epoch": 0.44565785652047635,
+      "grad_norm": 0.38773831725120544,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 3836
+    },
+    {
+      "epoch": 0.44577403427243684,
+      "grad_norm": 0.4000050723552704,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4458902120243973,
+      "grad_norm": 0.4279811382293701,
+      "learning_rate": 0.0001,
+      "loss": 1.7576,
+      "step": 3838
+    },
+    {
+      "epoch": 0.4460063897763578,
+      "grad_norm": 0.3969043493270874,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 3839
+    },
+    {
+      "epoch": 0.4461225675283183,
+      "grad_norm": 0.3967093825340271,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 3840
+    },
+    {
+      "epoch": 0.44623874528027885,
+      "grad_norm": 0.4337485432624817,
+      "learning_rate": 0.0001,
+      "loss": 1.7033,
+      "step": 3841
+    },
+    {
+      "epoch": 0.44635492303223934,
+      "grad_norm": 0.4148968756198883,
+      "learning_rate": 0.0001,
+      "loss": 1.7227,
+      "step": 3842
+    },
+    {
+      "epoch": 0.4464711007841998,
+      "grad_norm": 0.4100302457809448,
+      "learning_rate": 0.0001,
+      "loss": 1.7214,
+      "step": 3843
+    },
+    {
+      "epoch": 0.4465872785361603,
+      "grad_norm": 0.430477112531662,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 3844
+    },
+    {
+      "epoch": 0.4467034562881208,
+      "grad_norm": 0.424699068069458,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 3845
+    },
+    {
+      "epoch": 0.44681963404008135,
+      "grad_norm": 0.39795196056365967,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 3846
+    },
+    {
+      "epoch": 0.44693581179204184,
+      "grad_norm": 0.4175790548324585,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 3847
+    },
+    {
+      "epoch": 0.4470519895440023,
+      "grad_norm": 0.3961549699306488,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 3848
+    },
+    {
+      "epoch": 0.4471681672959628,
+      "grad_norm": 0.3898450434207916,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4472843450479233,
+      "grad_norm": 0.41821977496147156,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 3850
+    },
+    {
+      "epoch": 0.44740052279988385,
+      "grad_norm": 0.38670074939727783,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 3851
+    },
+    {
+      "epoch": 0.44751670055184434,
+      "grad_norm": 0.4136195182800293,
+      "learning_rate": 0.0001,
+      "loss": 1.8563,
+      "step": 3852
+    },
+    {
+      "epoch": 0.4476328783038048,
+      "grad_norm": 0.4174270033836365,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4477490560557653,
+      "grad_norm": 0.41631799936294556,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 3854
+    },
+    {
+      "epoch": 0.4478652338077258,
+      "grad_norm": 0.443198561668396,
+      "learning_rate": 0.0001,
+      "loss": 1.6342,
+      "step": 3855
+    },
+    {
+      "epoch": 0.4479814115596863,
+      "grad_norm": 0.4108307659626007,
+      "learning_rate": 0.0001,
+      "loss": 1.6264,
+      "step": 3856
+    },
+    {
+      "epoch": 0.44809758931164684,
+      "grad_norm": 0.36928313970565796,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4482137670636073,
+      "grad_norm": 0.42408835887908936,
+      "learning_rate": 0.0001,
+      "loss": 1.909,
+      "step": 3858
+    },
+    {
+      "epoch": 0.4483299448155678,
+      "grad_norm": 0.4042268991470337,
+      "learning_rate": 0.0001,
+      "loss": 1.6949,
+      "step": 3859
+    },
+    {
+      "epoch": 0.4484461225675283,
+      "grad_norm": 0.40864208340644836,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 3860
+    },
+    {
+      "epoch": 0.4485623003194888,
+      "grad_norm": 0.38092851638793945,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 3861
+    },
+    {
+      "epoch": 0.44867847807144934,
+      "grad_norm": 0.3783656656742096,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 3862
+    },
+    {
+      "epoch": 0.4487946558234098,
+      "grad_norm": 0.3975265920162201,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4489108335753703,
+      "grad_norm": 0.42539462447166443,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 3864
+    },
+    {
+      "epoch": 0.4490270113273308,
+      "grad_norm": 0.4166495203971863,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 3865
+    },
+    {
+      "epoch": 0.4491431890792913,
+      "grad_norm": 0.39668262004852295,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 3866
+    },
+    {
+      "epoch": 0.44925936683125184,
+      "grad_norm": 0.4214429557323456,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 3867
+    },
+    {
+      "epoch": 0.4493755445832123,
+      "grad_norm": 0.43809008598327637,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4494917223351728,
+      "grad_norm": 0.42034274339675903,
+      "learning_rate": 0.0001,
+      "loss": 1.794,
+      "step": 3869
+    },
+    {
+      "epoch": 0.4496079000871333,
+      "grad_norm": 0.3921016454696655,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 3870
+    },
+    {
+      "epoch": 0.4497240778390938,
+      "grad_norm": 0.4140496253967285,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 3871
+    },
+    {
+      "epoch": 0.44984025559105434,
+      "grad_norm": 0.43890661001205444,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 3872
+    },
+    {
+      "epoch": 0.4499564333430148,
+      "grad_norm": 0.40606656670570374,
+      "learning_rate": 0.0001,
+      "loss": 1.5126,
+      "step": 3873
+    },
+    {
+      "epoch": 0.4500726110949753,
+      "grad_norm": 0.45156994462013245,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 3874
+    },
+    {
+      "epoch": 0.4501887888469358,
+      "grad_norm": 0.4368317425251007,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4503049665988963,
+      "grad_norm": 0.41252076625823975,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 3876
+    },
+    {
+      "epoch": 0.4504211443508568,
+      "grad_norm": 0.46063879132270813,
+      "learning_rate": 0.0001,
+      "loss": 1.7607,
+      "step": 3877
+    },
+    {
+      "epoch": 0.45053732210281733,
+      "grad_norm": 0.4769091308116913,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 3878
+    },
+    {
+      "epoch": 0.4506534998547778,
+      "grad_norm": 0.3981684446334839,
+      "learning_rate": 0.0001,
+      "loss": 1.7009,
+      "step": 3879
+    },
+    {
+      "epoch": 0.4507696776067383,
+      "grad_norm": 0.4136994779109955,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4508858553586988,
+      "grad_norm": 0.394010454416275,
+      "learning_rate": 0.0001,
+      "loss": 1.6413,
+      "step": 3881
+    },
+    {
+      "epoch": 0.4510020331106593,
+      "grad_norm": 0.3855630159378052,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 3882
+    },
+    {
+      "epoch": 0.45111821086261983,
+      "grad_norm": 0.41295209527015686,
+      "learning_rate": 0.0001,
+      "loss": 1.7063,
+      "step": 3883
+    },
+    {
+      "epoch": 0.4512343886145803,
+      "grad_norm": 0.36102786660194397,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4513505663665408,
+      "grad_norm": 0.42346733808517456,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4514667441185013,
+      "grad_norm": 0.40333321690559387,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 3886
+    },
+    {
+      "epoch": 0.4515829218704618,
+      "grad_norm": 0.38405728340148926,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 3887
+    },
+    {
+      "epoch": 0.45169909962242233,
+      "grad_norm": 0.4108724594116211,
+      "learning_rate": 0.0001,
+      "loss": 1.6753,
+      "step": 3888
+    },
+    {
+      "epoch": 0.4518152773743828,
+      "grad_norm": 0.42868340015411377,
+      "learning_rate": 0.0001,
+      "loss": 1.7265,
+      "step": 3889
+    },
+    {
+      "epoch": 0.4519314551263433,
+      "grad_norm": 0.4091024100780487,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4520476328783038,
+      "grad_norm": 0.4288122355937958,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4521638106302643,
+      "grad_norm": 0.42493435740470886,
+      "learning_rate": 0.0001,
+      "loss": 1.8673,
+      "step": 3892
+    },
+    {
+      "epoch": 0.45227998838222483,
+      "grad_norm": 0.3990475833415985,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 3893
+    },
+    {
+      "epoch": 0.4523961661341853,
+      "grad_norm": 0.39818140864372253,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 3894
+    },
+    {
+      "epoch": 0.4525123438861458,
+      "grad_norm": 0.39904457330703735,
+      "learning_rate": 0.0001,
+      "loss": 1.648,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4526285216381063,
+      "grad_norm": 0.41773471236228943,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 3896
+    },
+    {
+      "epoch": 0.4527446993900668,
+      "grad_norm": 0.42894163727760315,
+      "learning_rate": 0.0001,
+      "loss": 1.805,
+      "step": 3897
+    },
+    {
+      "epoch": 0.45286087714202733,
+      "grad_norm": 0.4194585084915161,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4529770548939878,
+      "grad_norm": 0.40379759669303894,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 3899
+    },
+    {
+      "epoch": 0.4530932326459483,
+      "grad_norm": 0.4571918845176697,
+      "learning_rate": 0.0001,
+      "loss": 1.8804,
+      "step": 3900
+    },
+    {
+      "epoch": 0.4532094103979088,
+      "grad_norm": 0.3897266685962677,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 3901
+    },
+    {
+      "epoch": 0.4533255881498693,
+      "grad_norm": 0.4049232304096222,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 3902
+    },
+    {
+      "epoch": 0.4534417659018298,
+      "grad_norm": 0.43378233909606934,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 3903
+    },
+    {
+      "epoch": 0.4535579436537903,
+      "grad_norm": 0.43286213278770447,
+      "learning_rate": 0.0001,
+      "loss": 1.7369,
+      "step": 3904
+    },
+    {
+      "epoch": 0.4536741214057508,
+      "grad_norm": 0.41819891333580017,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 3905
+    },
+    {
+      "epoch": 0.4537902991577113,
+      "grad_norm": 0.4100865125656128,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4539064769096718,
+      "grad_norm": 0.3872487545013428,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 3907
+    },
+    {
+      "epoch": 0.4540226546616323,
+      "grad_norm": 0.4244650602340698,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 3908
+    },
+    {
+      "epoch": 0.4541388324135928,
+      "grad_norm": 0.38467180728912354,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 3909
+    },
+    {
+      "epoch": 0.4542550101655533,
+      "grad_norm": 0.3707321286201477,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4543711879175138,
+      "grad_norm": 0.412838876247406,
+      "learning_rate": 0.0001,
+      "loss": 1.6973,
+      "step": 3911
+    },
+    {
+      "epoch": 0.4544873656694743,
+      "grad_norm": 0.4340794086456299,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 3912
+    },
+    {
+      "epoch": 0.4546035434214348,
+      "grad_norm": 0.38097426295280457,
+      "learning_rate": 0.0001,
+      "loss": 1.4161,
+      "step": 3913
+    },
+    {
+      "epoch": 0.4547197211733953,
+      "grad_norm": 0.418896347284317,
+      "learning_rate": 0.0001,
+      "loss": 1.7463,
+      "step": 3914
+    },
+    {
+      "epoch": 0.4548358989253558,
+      "grad_norm": 0.427630752325058,
+      "learning_rate": 0.0001,
+      "loss": 1.8143,
+      "step": 3915
+    },
+    {
+      "epoch": 0.4549520766773163,
+      "grad_norm": 0.4016280174255371,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 3916
+    },
+    {
+      "epoch": 0.4550682544292768,
+      "grad_norm": 0.46618133783340454,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 3917
+    },
+    {
+      "epoch": 0.4551844321812373,
+      "grad_norm": 0.40906664729118347,
+      "learning_rate": 0.0001,
+      "loss": 1.5765,
+      "step": 3918
+    },
+    {
+      "epoch": 0.4553006099331978,
+      "grad_norm": 0.40268227458000183,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 3919
+    },
+    {
+      "epoch": 0.4554167876851583,
+      "grad_norm": 0.408345490694046,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 3920
+    },
+    {
+      "epoch": 0.4555329654371188,
+      "grad_norm": 0.39653435349464417,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 3921
+    },
+    {
+      "epoch": 0.4556491431890793,
+      "grad_norm": 0.39751750230789185,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 3922
+    },
+    {
+      "epoch": 0.4557653209410398,
+      "grad_norm": 0.3973672688007355,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 3923
+    },
+    {
+      "epoch": 0.45588149869300026,
+      "grad_norm": 0.43432149291038513,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 3924
+    },
+    {
+      "epoch": 0.4559976764449608,
+      "grad_norm": 0.43509483337402344,
+      "learning_rate": 0.0001,
+      "loss": 1.7848,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4561138541969213,
+      "grad_norm": 0.407279908657074,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 3926
+    },
+    {
+      "epoch": 0.4562300319488818,
+      "grad_norm": 0.40425196290016174,
+      "learning_rate": 0.0001,
+      "loss": 1.6857,
+      "step": 3927
+    },
+    {
+      "epoch": 0.4563462097008423,
+      "grad_norm": 0.4185945689678192,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 3928
+    },
+    {
+      "epoch": 0.45646238745280276,
+      "grad_norm": 0.39828792214393616,
+      "learning_rate": 0.0001,
+      "loss": 1.7497,
+      "step": 3929
+    },
+    {
+      "epoch": 0.4565785652047633,
+      "grad_norm": 0.40944311022758484,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4566947429567238,
+      "grad_norm": 0.3925982117652893,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 3931
+    },
+    {
+      "epoch": 0.4568109207086843,
+      "grad_norm": 0.413461834192276,
+      "learning_rate": 0.0001,
+      "loss": 1.7878,
+      "step": 3932
+    },
+    {
+      "epoch": 0.4569270984606448,
+      "grad_norm": 0.4017687737941742,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 3933
+    },
+    {
+      "epoch": 0.45704327621260527,
+      "grad_norm": 0.3962691128253937,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 3934
+    },
+    {
+      "epoch": 0.4571594539645658,
+      "grad_norm": 0.4170648753643036,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4572756317165263,
+      "grad_norm": 0.4325627386569977,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 3936
+    },
+    {
+      "epoch": 0.4573918094684868,
+      "grad_norm": 0.40576669573783875,
+      "learning_rate": 0.0001,
+      "loss": 1.5092,
+      "step": 3937
+    },
+    {
+      "epoch": 0.4575079872204473,
+      "grad_norm": 0.40341272950172424,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 3938
+    },
+    {
+      "epoch": 0.45762416497240777,
+      "grad_norm": 0.3955957293510437,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 3939
+    },
+    {
+      "epoch": 0.4577403427243683,
+      "grad_norm": 0.3895145356655121,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 3940
+    },
+    {
+      "epoch": 0.4578565204763288,
+      "grad_norm": 0.4147211015224457,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 3941
+    },
+    {
+      "epoch": 0.4579726982282893,
+      "grad_norm": 0.39672261476516724,
+      "learning_rate": 0.0001,
+      "loss": 1.5186,
+      "step": 3942
+    },
+    {
+      "epoch": 0.4580888759802498,
+      "grad_norm": 0.4300733506679535,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 3943
+    },
+    {
+      "epoch": 0.45820505373221027,
+      "grad_norm": 0.4062694013118744,
+      "learning_rate": 0.0001,
+      "loss": 1.8041,
+      "step": 3944
+    },
+    {
+      "epoch": 0.45832123148417075,
+      "grad_norm": 0.4132828414440155,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 3945
+    },
+    {
+      "epoch": 0.4584374092361313,
+      "grad_norm": 0.42365893721580505,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 3946
+    },
+    {
+      "epoch": 0.4585535869880918,
+      "grad_norm": 0.3917977511882782,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 3947
+    },
+    {
+      "epoch": 0.4586697647400523,
+      "grad_norm": 0.4075486361980438,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 3948
+    },
+    {
+      "epoch": 0.45878594249201277,
+      "grad_norm": 0.41571909189224243,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 3949
+    },
+    {
+      "epoch": 0.45890212024397325,
+      "grad_norm": 0.44082123041152954,
+      "learning_rate": 0.0001,
+      "loss": 1.7793,
+      "step": 3950
+    },
+    {
+      "epoch": 0.4590182979959338,
+      "grad_norm": 0.40916866064071655,
+      "learning_rate": 0.0001,
+      "loss": 1.7366,
+      "step": 3951
+    },
+    {
+      "epoch": 0.4591344757478943,
+      "grad_norm": 0.3818245530128479,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 3952
+    },
+    {
+      "epoch": 0.4592506534998548,
+      "grad_norm": 0.4646684527397156,
+      "learning_rate": 0.0001,
+      "loss": 1.8009,
+      "step": 3953
+    },
+    {
+      "epoch": 0.45936683125181527,
+      "grad_norm": 0.4245597720146179,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 3954
+    },
+    {
+      "epoch": 0.45948300900377576,
+      "grad_norm": 0.42128875851631165,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 3955
+    },
+    {
+      "epoch": 0.4595991867557363,
+      "grad_norm": 0.41281989216804504,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 3956
+    },
+    {
+      "epoch": 0.4597153645076968,
+      "grad_norm": 0.4056452810764313,
+      "learning_rate": 0.0001,
+      "loss": 1.6749,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4598315422596573,
+      "grad_norm": 0.4163060188293457,
+      "learning_rate": 0.0001,
+      "loss": 1.6496,
+      "step": 3958
+    },
+    {
+      "epoch": 0.45994772001161777,
+      "grad_norm": 0.4370429813861847,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 3959
+    },
+    {
+      "epoch": 0.46006389776357826,
+      "grad_norm": 0.4308282136917114,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4601800755155388,
+      "grad_norm": 0.42895281314849854,
+      "learning_rate": 0.0001,
+      "loss": 1.9235,
+      "step": 3961
+    },
+    {
+      "epoch": 0.4602962532674993,
+      "grad_norm": 0.3952750563621521,
+      "learning_rate": 0.0001,
+      "loss": 1.7455,
+      "step": 3962
+    },
+    {
+      "epoch": 0.4604124310194598,
+      "grad_norm": 0.4076629877090454,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 3963
+    },
+    {
+      "epoch": 0.46052860877142027,
+      "grad_norm": 0.41401106119155884,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 3964
+    },
+    {
+      "epoch": 0.46064478652338076,
+      "grad_norm": 0.39911845326423645,
+      "learning_rate": 0.0001,
+      "loss": 1.7128,
+      "step": 3965
+    },
+    {
+      "epoch": 0.46076096427534124,
+      "grad_norm": 0.3996291160583496,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 3966
+    },
+    {
+      "epoch": 0.4608771420273018,
+      "grad_norm": 0.4146043360233307,
+      "learning_rate": 0.0001,
+      "loss": 1.7652,
+      "step": 3967
+    },
+    {
+      "epoch": 0.4609933197792623,
+      "grad_norm": 0.4056430459022522,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 3968
+    },
+    {
+      "epoch": 0.46110949753122277,
+      "grad_norm": 0.3950080871582031,
+      "learning_rate": 0.0001,
+      "loss": 1.6484,
+      "step": 3969
+    },
+    {
+      "epoch": 0.46122567528318326,
+      "grad_norm": 0.3902670443058014,
+      "learning_rate": 0.0001,
+      "loss": 1.5769,
+      "step": 3970
+    },
+    {
+      "epoch": 0.46134185303514375,
+      "grad_norm": 0.408426970243454,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 3971
+    },
+    {
+      "epoch": 0.4614580307871043,
+      "grad_norm": 0.42066359519958496,
+      "learning_rate": 0.0001,
+      "loss": 1.8631,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4615742085390648,
+      "grad_norm": 0.4041798412799835,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 3973
+    },
+    {
+      "epoch": 0.46169038629102527,
+      "grad_norm": 0.4431467056274414,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 3974
+    },
+    {
+      "epoch": 0.46180656404298576,
+      "grad_norm": 0.3934585154056549,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 3975
+    },
+    {
+      "epoch": 0.46192274179494625,
+      "grad_norm": 0.40387487411499023,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 3976
+    },
+    {
+      "epoch": 0.4620389195469068,
+      "grad_norm": 0.39581024646759033,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 3977
+    },
+    {
+      "epoch": 0.4621550972988673,
+      "grad_norm": 0.43566346168518066,
+      "learning_rate": 0.0001,
+      "loss": 1.7688,
+      "step": 3978
+    },
+    {
+      "epoch": 0.46227127505082777,
+      "grad_norm": 0.399808794260025,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 3979
+    },
+    {
+      "epoch": 0.46238745280278826,
+      "grad_norm": 0.42738527059555054,
+      "learning_rate": 0.0001,
+      "loss": 1.6639,
+      "step": 3980
+    },
+    {
+      "epoch": 0.46250363055474875,
+      "grad_norm": 0.3896215260028839,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 3981
+    },
+    {
+      "epoch": 0.4626198083067093,
+      "grad_norm": 0.41741159558296204,
+      "learning_rate": 0.0001,
+      "loss": 1.7703,
+      "step": 3982
+    },
+    {
+      "epoch": 0.4627359860586698,
+      "grad_norm": 0.40396156907081604,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 3983
+    },
+    {
+      "epoch": 0.46285216381063027,
+      "grad_norm": 0.38009852170944214,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 3984
+    },
+    {
+      "epoch": 0.46296834156259076,
+      "grad_norm": 0.43098655343055725,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 3985
+    },
+    {
+      "epoch": 0.46308451931455125,
+      "grad_norm": 0.461058109998703,
+      "learning_rate": 0.0001,
+      "loss": 1.6804,
+      "step": 3986
+    },
+    {
+      "epoch": 0.4632006970665118,
+      "grad_norm": 0.45243731141090393,
+      "learning_rate": 0.0001,
+      "loss": 1.9556,
+      "step": 3987
+    },
+    {
+      "epoch": 0.4633168748184723,
+      "grad_norm": 0.41540998220443726,
+      "learning_rate": 0.0001,
+      "loss": 1.5549,
+      "step": 3988
+    },
+    {
+      "epoch": 0.46343305257043277,
+      "grad_norm": 0.4656381905078888,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 3989
+    },
+    {
+      "epoch": 0.46354923032239326,
+      "grad_norm": 0.4463013708591461,
+      "learning_rate": 0.0001,
+      "loss": 1.7644,
+      "step": 3990
+    },
+    {
+      "epoch": 0.46366540807435375,
+      "grad_norm": 0.40116289258003235,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 3991
+    },
+    {
+      "epoch": 0.46378158582631424,
+      "grad_norm": 0.41295450925827026,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 3992
+    },
+    {
+      "epoch": 0.4638977635782748,
+      "grad_norm": 0.43854475021362305,
+      "learning_rate": 0.0001,
+      "loss": 1.8506,
+      "step": 3993
+    },
+    {
+      "epoch": 0.46401394133023527,
+      "grad_norm": 0.39046186208724976,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 3994
+    },
+    {
+      "epoch": 0.46413011908219576,
+      "grad_norm": 0.4226478338241577,
+      "learning_rate": 0.0001,
+      "loss": 1.7984,
+      "step": 3995
+    },
+    {
+      "epoch": 0.46424629683415625,
+      "grad_norm": 0.40615102648735046,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 3996
+    },
+    {
+      "epoch": 0.46436247458611674,
+      "grad_norm": 0.41897422075271606,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 3997
+    },
+    {
+      "epoch": 0.4644786523380773,
+      "grad_norm": 0.3868965804576874,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 3998
+    },
+    {
+      "epoch": 0.46459483009003777,
+      "grad_norm": 0.40913960337638855,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 3999
+    },
+    {
+      "epoch": 0.46471100784199826,
+      "grad_norm": 0.44357261061668396,
+      "learning_rate": 0.0001,
+      "loss": 1.5254,
+      "step": 4000
+    },
+    {
+      "epoch": 0.46482718559395875,
+      "grad_norm": 0.42295897006988525,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 4001
+    },
+    {
+      "epoch": 0.46494336334591924,
+      "grad_norm": 0.38419994711875916,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 4002
+    },
+    {
+      "epoch": 0.4650595410978798,
+      "grad_norm": 0.4167155623435974,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 4003
+    },
+    {
+      "epoch": 0.46517571884984027,
+      "grad_norm": 0.394581139087677,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 4004
+    },
+    {
+      "epoch": 0.46529189660180076,
+      "grad_norm": 0.4233587384223938,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 4005
+    },
+    {
+      "epoch": 0.46540807435376125,
+      "grad_norm": 0.3914411962032318,
+      "learning_rate": 0.0001,
+      "loss": 1.5213,
+      "step": 4006
+    },
+    {
+      "epoch": 0.46552425210572174,
+      "grad_norm": 0.4535432457923889,
+      "learning_rate": 0.0001,
+      "loss": 1.798,
+      "step": 4007
+    },
+    {
+      "epoch": 0.4656404298576823,
+      "grad_norm": 0.39255577325820923,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 4008
+    },
+    {
+      "epoch": 0.46575660760964277,
+      "grad_norm": 0.38806408643722534,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 4009
+    },
+    {
+      "epoch": 0.46587278536160326,
+      "grad_norm": 0.4313128888607025,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 4010
+    },
+    {
+      "epoch": 0.46598896311356375,
+      "grad_norm": 0.39540693163871765,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 4011
+    },
+    {
+      "epoch": 0.46610514086552424,
+      "grad_norm": 0.40360215306282043,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 4012
+    },
+    {
+      "epoch": 0.4662213186174847,
+      "grad_norm": 0.4235546886920929,
+      "learning_rate": 0.0001,
+      "loss": 1.7763,
+      "step": 4013
+    },
+    {
+      "epoch": 0.46633749636944527,
+      "grad_norm": 0.4328942596912384,
+      "learning_rate": 0.0001,
+      "loss": 1.8291,
+      "step": 4014
+    },
+    {
+      "epoch": 0.46645367412140576,
+      "grad_norm": 0.40199771523475647,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 4015
+    },
+    {
+      "epoch": 0.46656985187336625,
+      "grad_norm": 0.3917251527309418,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 4016
+    },
+    {
+      "epoch": 0.46668602962532674,
+      "grad_norm": 0.4347023665904999,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 4017
+    },
+    {
+      "epoch": 0.4668022073772872,
+      "grad_norm": 0.41650915145874023,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 4018
+    },
+    {
+      "epoch": 0.46691838512924777,
+      "grad_norm": 0.39249107241630554,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 4019
+    },
+    {
+      "epoch": 0.46703456288120826,
+      "grad_norm": 0.42523854970932007,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 4020
+    },
+    {
+      "epoch": 0.46715074063316875,
+      "grad_norm": 0.385466068983078,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 4021
+    },
+    {
+      "epoch": 0.46726691838512924,
+      "grad_norm": 0.39372915029525757,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 4022
+    },
+    {
+      "epoch": 0.4673830961370897,
+      "grad_norm": 0.41617611050605774,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 4023
+    },
+    {
+      "epoch": 0.46749927388905027,
+      "grad_norm": 0.4050281047821045,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 4024
+    },
+    {
+      "epoch": 0.46761545164101076,
+      "grad_norm": 0.41811496019363403,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 4025
+    },
+    {
+      "epoch": 0.46773162939297125,
+      "grad_norm": 0.4445625841617584,
+      "learning_rate": 0.0001,
+      "loss": 1.6577,
+      "step": 4026
+    },
+    {
+      "epoch": 0.46784780714493174,
+      "grad_norm": 0.3947649598121643,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 4027
+    },
+    {
+      "epoch": 0.4679639848968922,
+      "grad_norm": 0.39127346873283386,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 4028
+    },
+    {
+      "epoch": 0.46808016264885277,
+      "grad_norm": 0.4156622588634491,
+      "learning_rate": 0.0001,
+      "loss": 1.782,
+      "step": 4029
+    },
+    {
+      "epoch": 0.46819634040081326,
+      "grad_norm": 0.40657055377960205,
+      "learning_rate": 0.0001,
+      "loss": 1.7922,
+      "step": 4030
+    },
+    {
+      "epoch": 0.46831251815277375,
+      "grad_norm": 0.4071350693702698,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 4031
+    },
+    {
+      "epoch": 0.46842869590473424,
+      "grad_norm": 0.41567331552505493,
+      "learning_rate": 0.0001,
+      "loss": 1.703,
+      "step": 4032
+    },
+    {
+      "epoch": 0.4685448736566947,
+      "grad_norm": 0.4297296106815338,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 4033
+    },
+    {
+      "epoch": 0.4686610514086552,
+      "grad_norm": 0.46219146251678467,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 4034
+    },
+    {
+      "epoch": 0.46877722916061576,
+      "grad_norm": 0.38711977005004883,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 4035
+    },
+    {
+      "epoch": 0.46889340691257625,
+      "grad_norm": 0.39784133434295654,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 4036
+    },
+    {
+      "epoch": 0.46900958466453674,
+      "grad_norm": 0.4138612747192383,
+      "learning_rate": 0.0001,
+      "loss": 1.7207,
+      "step": 4037
+    },
+    {
+      "epoch": 0.4691257624164972,
+      "grad_norm": 0.42039382457733154,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 4038
+    },
+    {
+      "epoch": 0.4692419401684577,
+      "grad_norm": 0.40157628059387207,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 4039
+    },
+    {
+      "epoch": 0.46935811792041826,
+      "grad_norm": 0.40044525265693665,
+      "learning_rate": 0.0001,
+      "loss": 1.7727,
+      "step": 4040
+    },
+    {
+      "epoch": 0.46947429567237875,
+      "grad_norm": 0.4081622064113617,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 4041
+    },
+    {
+      "epoch": 0.46959047342433924,
+      "grad_norm": 0.40018370747566223,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 4042
+    },
+    {
+      "epoch": 0.4697066511762997,
+      "grad_norm": 0.3724619746208191,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 4043
+    },
+    {
+      "epoch": 0.4698228289282602,
+      "grad_norm": 0.386092871427536,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 4044
+    },
+    {
+      "epoch": 0.46993900668022076,
+      "grad_norm": 0.44518908858299255,
+      "learning_rate": 0.0001,
+      "loss": 1.8437,
+      "step": 4045
+    },
+    {
+      "epoch": 0.47005518443218125,
+      "grad_norm": 0.4211956560611725,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 4046
+    },
+    {
+      "epoch": 0.47017136218414174,
+      "grad_norm": 0.4604097306728363,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 4047
+    },
+    {
+      "epoch": 0.47028753993610223,
+      "grad_norm": 0.4125956892967224,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 4048
+    },
+    {
+      "epoch": 0.4704037176880627,
+      "grad_norm": 0.46292823553085327,
+      "learning_rate": 0.0001,
+      "loss": 1.877,
+      "step": 4049
+    },
+    {
+      "epoch": 0.47051989544002326,
+      "grad_norm": 0.4355732202529907,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 4050
+    },
+    {
+      "epoch": 0.47063607319198375,
+      "grad_norm": 0.4154967665672302,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 4051
+    },
+    {
+      "epoch": 0.47075225094394424,
+      "grad_norm": 0.4346443712711334,
+      "learning_rate": 0.0001,
+      "loss": 1.6655,
+      "step": 4052
+    },
+    {
+      "epoch": 0.47086842869590473,
+      "grad_norm": 0.41704896092414856,
+      "learning_rate": 0.0001,
+      "loss": 1.8386,
+      "step": 4053
+    },
+    {
+      "epoch": 0.4709846064478652,
+      "grad_norm": 0.4196739196777344,
+      "learning_rate": 0.0001,
+      "loss": 1.6717,
+      "step": 4054
+    },
+    {
+      "epoch": 0.4711007841998257,
+      "grad_norm": 0.43008390069007874,
+      "learning_rate": 0.0001,
+      "loss": 1.7694,
+      "step": 4055
+    },
+    {
+      "epoch": 0.47121696195178625,
+      "grad_norm": 0.37963777780532837,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 4056
+    },
+    {
+      "epoch": 0.47133313970374674,
+      "grad_norm": 0.40089571475982666,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 4057
+    },
+    {
+      "epoch": 0.47144931745570723,
+      "grad_norm": 0.4169794023036957,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 4058
+    },
+    {
+      "epoch": 0.4715654952076677,
+      "grad_norm": 0.4090938866138458,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 4059
+    },
+    {
+      "epoch": 0.4716816729596282,
+      "grad_norm": 0.4060511589050293,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 4060
+    },
+    {
+      "epoch": 0.47179785071158875,
+      "grad_norm": 0.4162571132183075,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 4061
+    },
+    {
+      "epoch": 0.47191402846354924,
+      "grad_norm": 0.39749133586883545,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 4062
+    },
+    {
+      "epoch": 0.47203020621550973,
+      "grad_norm": 0.4047934412956238,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 4063
+    },
+    {
+      "epoch": 0.4721463839674702,
+      "grad_norm": 0.436172217130661,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 4064
+    },
+    {
+      "epoch": 0.4722625617194307,
+      "grad_norm": 0.4056571125984192,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 4065
+    },
+    {
+      "epoch": 0.47237873947139125,
+      "grad_norm": 0.4182429909706116,
+      "learning_rate": 0.0001,
+      "loss": 1.7614,
+      "step": 4066
+    },
+    {
+      "epoch": 0.47249491722335174,
+      "grad_norm": 0.3990522623062134,
+      "learning_rate": 0.0001,
+      "loss": 1.5811,
+      "step": 4067
+    },
+    {
+      "epoch": 0.47261109497531223,
+      "grad_norm": 0.4048585593700409,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 4068
+    },
+    {
+      "epoch": 0.4727272727272727,
+      "grad_norm": 0.3930312693119049,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4728434504792332,
+      "grad_norm": 0.4067966639995575,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 4070
+    },
+    {
+      "epoch": 0.47295962823119375,
+      "grad_norm": 0.4243100881576538,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 4071
+    },
+    {
+      "epoch": 0.47307580598315424,
+      "grad_norm": 0.4516909718513489,
+      "learning_rate": 0.0001,
+      "loss": 1.7275,
+      "step": 4072
+    },
+    {
+      "epoch": 0.47319198373511473,
+      "grad_norm": 0.4081507921218872,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 4073
+    },
+    {
+      "epoch": 0.4733081614870752,
+      "grad_norm": 0.43128862977027893,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 4074
+    },
+    {
+      "epoch": 0.4734243392390357,
+      "grad_norm": 0.43954023718833923,
+      "learning_rate": 0.0001,
+      "loss": 1.8691,
+      "step": 4075
+    },
+    {
+      "epoch": 0.47354051699099625,
+      "grad_norm": 0.422748863697052,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 4076
+    },
+    {
+      "epoch": 0.47365669474295674,
+      "grad_norm": 0.40614694356918335,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 4077
+    },
+    {
+      "epoch": 0.47377287249491723,
+      "grad_norm": 0.4623175263404846,
+      "learning_rate": 0.0001,
+      "loss": 1.8329,
+      "step": 4078
+    },
+    {
+      "epoch": 0.4738890502468777,
+      "grad_norm": 0.44638627767562866,
+      "learning_rate": 0.0001,
+      "loss": 1.9259,
+      "step": 4079
+    },
+    {
+      "epoch": 0.4740052279988382,
+      "grad_norm": 0.4093332588672638,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4741214057507987,
+      "grad_norm": 0.4484642446041107,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 4081
+    },
+    {
+      "epoch": 0.47423758350275924,
+      "grad_norm": 0.4261915981769562,
+      "learning_rate": 0.0001,
+      "loss": 1.7479,
+      "step": 4082
+    },
+    {
+      "epoch": 0.47435376125471973,
+      "grad_norm": 0.3942796587944031,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 4083
+    },
+    {
+      "epoch": 0.4744699390066802,
+      "grad_norm": 0.4067613184452057,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 4084
+    },
+    {
+      "epoch": 0.4745861167586407,
+      "grad_norm": 0.42234787344932556,
+      "learning_rate": 0.0001,
+      "loss": 1.6912,
+      "step": 4085
+    },
+    {
+      "epoch": 0.4747022945106012,
+      "grad_norm": 0.38540762662887573,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 4086
+    },
+    {
+      "epoch": 0.47481847226256174,
+      "grad_norm": 0.386514276266098,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 4087
+    },
+    {
+      "epoch": 0.47493465001452223,
+      "grad_norm": 0.4075110852718353,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 4088
+    },
+    {
+      "epoch": 0.4750508277664827,
+      "grad_norm": 0.39759472012519836,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 4089
+    },
+    {
+      "epoch": 0.4751670055184432,
+      "grad_norm": 0.3987995386123657,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 4090
+    },
+    {
+      "epoch": 0.4752831832704037,
+      "grad_norm": 0.3973774313926697,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 4091
+    },
+    {
+      "epoch": 0.47539936102236424,
+      "grad_norm": 0.4317288398742676,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 4092
+    },
+    {
+      "epoch": 0.47551553877432473,
+      "grad_norm": 0.4416712522506714,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 4093
+    },
+    {
+      "epoch": 0.4756317165262852,
+      "grad_norm": 0.43052783608436584,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 4094
+    },
+    {
+      "epoch": 0.4757478942782457,
+      "grad_norm": 0.4163714647293091,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 4095
+    },
+    {
+      "epoch": 0.4758640720302062,
+      "grad_norm": 0.41067805886268616,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 4096
+    },
+    {
+      "epoch": 0.47598024978216674,
+      "grad_norm": 0.4482041001319885,
+      "learning_rate": 0.0001,
+      "loss": 1.8553,
+      "step": 4097
+    },
+    {
+      "epoch": 0.47609642753412723,
+      "grad_norm": 0.4115692973136902,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 4098
+    },
+    {
+      "epoch": 0.4762126052860877,
+      "grad_norm": 0.4347507953643799,
+      "learning_rate": 0.0001,
+      "loss": 1.7567,
+      "step": 4099
+    },
+    {
+      "epoch": 0.4763287830380482,
+      "grad_norm": 0.409196674823761,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4764449607900087,
+      "grad_norm": 0.46702349185943604,
+      "learning_rate": 0.0001,
+      "loss": 1.7937,
+      "step": 4101
+    },
+    {
+      "epoch": 0.4765611385419692,
+      "grad_norm": 0.3931210935115814,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 4102
+    },
+    {
+      "epoch": 0.47667731629392973,
+      "grad_norm": 0.42548468708992004,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 4103
+    },
+    {
+      "epoch": 0.4767934940458902,
+      "grad_norm": 0.690558910369873,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 4104
+    },
+    {
+      "epoch": 0.4769096717978507,
+      "grad_norm": 0.41380855441093445,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 4105
+    },
+    {
+      "epoch": 0.4770258495498112,
+      "grad_norm": 0.37842079997062683,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 4106
+    },
+    {
+      "epoch": 0.4771420273017717,
+      "grad_norm": 0.4153887629508972,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 4107
+    },
+    {
+      "epoch": 0.47725820505373223,
+      "grad_norm": 0.408482164144516,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 4108
+    },
+    {
+      "epoch": 0.4773743828056927,
+      "grad_norm": 0.4116607904434204,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 4109
+    },
+    {
+      "epoch": 0.4774905605576532,
+      "grad_norm": 0.3892625570297241,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4776067383096137,
+      "grad_norm": 0.41425132751464844,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 4111
+    },
+    {
+      "epoch": 0.4777229160615742,
+      "grad_norm": 0.4124647080898285,
+      "learning_rate": 0.0001,
+      "loss": 1.5062,
+      "step": 4112
+    },
+    {
+      "epoch": 0.47783909381353473,
+      "grad_norm": 0.43545451760292053,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 4113
+    },
+    {
+      "epoch": 0.4779552715654952,
+      "grad_norm": 0.4274523854255676,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 4114
+    },
+    {
+      "epoch": 0.4780714493174557,
+      "grad_norm": 0.41416046023368835,
+      "learning_rate": 0.0001,
+      "loss": 1.6724,
+      "step": 4115
+    },
+    {
+      "epoch": 0.4781876270694162,
+      "grad_norm": 0.4206220805644989,
+      "learning_rate": 0.0001,
+      "loss": 1.7191,
+      "step": 4116
+    },
+    {
+      "epoch": 0.4783038048213767,
+      "grad_norm": 0.4083358645439148,
+      "learning_rate": 0.0001,
+      "loss": 1.7273,
+      "step": 4117
+    },
+    {
+      "epoch": 0.47841998257333723,
+      "grad_norm": 0.4127695560455322,
+      "learning_rate": 0.0001,
+      "loss": 1.8423,
+      "step": 4118
+    },
+    {
+      "epoch": 0.4785361603252977,
+      "grad_norm": 0.441363662481308,
+      "learning_rate": 0.0001,
+      "loss": 1.8589,
+      "step": 4119
+    },
+    {
+      "epoch": 0.4786523380772582,
+      "grad_norm": 0.41235068440437317,
+      "learning_rate": 0.0001,
+      "loss": 1.7405,
+      "step": 4120
+    },
+    {
+      "epoch": 0.4787685158292187,
+      "grad_norm": 0.41113796830177307,
+      "learning_rate": 0.0001,
+      "loss": 1.6793,
+      "step": 4121
+    },
+    {
+      "epoch": 0.4788846935811792,
+      "grad_norm": 0.44025418162345886,
+      "learning_rate": 0.0001,
+      "loss": 1.7407,
+      "step": 4122
+    },
+    {
+      "epoch": 0.4790008713331397,
+      "grad_norm": 0.3890633285045624,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 4123
+    },
+    {
+      "epoch": 0.4791170490851002,
+      "grad_norm": 0.40119051933288574,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 4124
+    },
+    {
+      "epoch": 0.4792332268370607,
+      "grad_norm": 0.4198724329471588,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 4125
+    },
+    {
+      "epoch": 0.4793494045890212,
+      "grad_norm": 0.4038214385509491,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 4126
+    },
+    {
+      "epoch": 0.4794655823409817,
+      "grad_norm": 0.4278080463409424,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 4127
+    },
+    {
+      "epoch": 0.4795817600929422,
+      "grad_norm": 0.41188672184944153,
+      "learning_rate": 0.0001,
+      "loss": 1.621,
+      "step": 4128
+    },
+    {
+      "epoch": 0.4796979378449027,
+      "grad_norm": 0.43129464983940125,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 4129
+    },
+    {
+      "epoch": 0.4798141155968632,
+      "grad_norm": 0.4218551814556122,
+      "learning_rate": 0.0001,
+      "loss": 1.742,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4799302933488237,
+      "grad_norm": 0.37498828768730164,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 4131
+    },
+    {
+      "epoch": 0.4800464711007842,
+      "grad_norm": 0.4059140682220459,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 4132
+    },
+    {
+      "epoch": 0.4801626488527447,
+      "grad_norm": 0.4361055791378021,
+      "learning_rate": 0.0001,
+      "loss": 1.9007,
+      "step": 4133
+    },
+    {
+      "epoch": 0.4802788266047052,
+      "grad_norm": 0.40483996272087097,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 4134
+    },
+    {
+      "epoch": 0.4803950043566657,
+      "grad_norm": 0.4151374101638794,
+      "learning_rate": 0.0001,
+      "loss": 1.8005,
+      "step": 4135
+    },
+    {
+      "epoch": 0.4805111821086262,
+      "grad_norm": 0.4022291302680969,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 4136
+    },
+    {
+      "epoch": 0.4806273598605867,
+      "grad_norm": 0.40195754170417786,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 4137
+    },
+    {
+      "epoch": 0.4807435376125472,
+      "grad_norm": 0.41277408599853516,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 4138
+    },
+    {
+      "epoch": 0.4808597153645077,
+      "grad_norm": 0.4332813024520874,
+      "learning_rate": 0.0001,
+      "loss": 1.9132,
+      "step": 4139
+    },
+    {
+      "epoch": 0.4809758931164682,
+      "grad_norm": 0.3997774124145508,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 4140
+    },
+    {
+      "epoch": 0.4810920708684287,
+      "grad_norm": 0.4188762605190277,
+      "learning_rate": 0.0001,
+      "loss": 1.6811,
+      "step": 4141
+    },
+    {
+      "epoch": 0.4812082486203892,
+      "grad_norm": 0.40828222036361694,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 4142
+    },
+    {
+      "epoch": 0.4813244263723497,
+      "grad_norm": 0.4262183904647827,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 4143
+    },
+    {
+      "epoch": 0.48144060412431017,
+      "grad_norm": 0.4175308346748352,
+      "learning_rate": 0.0001,
+      "loss": 1.7043,
+      "step": 4144
+    },
+    {
+      "epoch": 0.4815567818762707,
+      "grad_norm": 0.40894895792007446,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 4145
+    },
+    {
+      "epoch": 0.4816729596282312,
+      "grad_norm": 0.4432525038719177,
+      "learning_rate": 0.0001,
+      "loss": 1.6228,
+      "step": 4146
+    },
+    {
+      "epoch": 0.4817891373801917,
+      "grad_norm": 0.412423700094223,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 4147
+    },
+    {
+      "epoch": 0.4819053151321522,
+      "grad_norm": 0.3926151990890503,
+      "learning_rate": 0.0001,
+      "loss": 1.697,
+      "step": 4148
+    },
+    {
+      "epoch": 0.48202149288411267,
+      "grad_norm": 0.4282263517379761,
+      "learning_rate": 0.0001,
+      "loss": 1.7589,
+      "step": 4149
+    },
+    {
+      "epoch": 0.4821376706360732,
+      "grad_norm": 0.3978181481361389,
+      "learning_rate": 0.0001,
+      "loss": 1.4873,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4822538483880337,
+      "grad_norm": 0.39814430475234985,
+      "learning_rate": 0.0001,
+      "loss": 1.6877,
+      "step": 4151
+    },
+    {
+      "epoch": 0.4823700261399942,
+      "grad_norm": 0.40920761227607727,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 4152
+    },
+    {
+      "epoch": 0.4824862038919547,
+      "grad_norm": 0.43666741251945496,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 4153
+    },
+    {
+      "epoch": 0.48260238164391517,
+      "grad_norm": 0.449294775724411,
+      "learning_rate": 0.0001,
+      "loss": 1.7839,
+      "step": 4154
+    },
+    {
+      "epoch": 0.4827185593958757,
+      "grad_norm": 0.4310729205608368,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 4155
+    },
+    {
+      "epoch": 0.4828347371478362,
+      "grad_norm": 0.4184435307979584,
+      "learning_rate": 0.0001,
+      "loss": 1.7023,
+      "step": 4156
+    },
+    {
+      "epoch": 0.4829509148997967,
+      "grad_norm": 0.4592064321041107,
+      "learning_rate": 0.0001,
+      "loss": 1.853,
+      "step": 4157
+    },
+    {
+      "epoch": 0.4830670926517572,
+      "grad_norm": 0.41902077198028564,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 4158
+    },
+    {
+      "epoch": 0.48318327040371767,
+      "grad_norm": 0.40385156869888306,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 4159
+    },
+    {
+      "epoch": 0.4832994481556782,
+      "grad_norm": 0.4063291549682617,
+      "learning_rate": 0.0001,
+      "loss": 1.444,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4834156259076387,
+      "grad_norm": 0.4716438055038452,
+      "learning_rate": 0.0001,
+      "loss": 1.8141,
+      "step": 4161
+    },
+    {
+      "epoch": 0.4835318036595992,
+      "grad_norm": 0.3783859312534332,
+      "learning_rate": 0.0001,
+      "loss": 1.3083,
+      "step": 4162
+    },
+    {
+      "epoch": 0.4836479814115597,
+      "grad_norm": 0.4354921281337738,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 4163
+    },
+    {
+      "epoch": 0.48376415916352017,
+      "grad_norm": 0.4501488506793976,
+      "learning_rate": 0.0001,
+      "loss": 1.7144,
+      "step": 4164
+    },
+    {
+      "epoch": 0.48388033691548066,
+      "grad_norm": 0.42613235116004944,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 4165
+    },
+    {
+      "epoch": 0.4839965146674412,
+      "grad_norm": 0.41294065117836,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 4166
+    },
+    {
+      "epoch": 0.4841126924194017,
+      "grad_norm": 0.399277001619339,
+      "learning_rate": 0.0001,
+      "loss": 1.4283,
+      "step": 4167
+    },
+    {
+      "epoch": 0.4842288701713622,
+      "grad_norm": 0.4134821891784668,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 4168
+    },
+    {
+      "epoch": 0.48434504792332267,
+      "grad_norm": 0.4123948812484741,
+      "learning_rate": 0.0001,
+      "loss": 1.6264,
+      "step": 4169
+    },
+    {
+      "epoch": 0.48446122567528316,
+      "grad_norm": 0.4469812214374542,
+      "learning_rate": 0.0001,
+      "loss": 1.8222,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4845774034272437,
+      "grad_norm": 0.4323446452617645,
+      "learning_rate": 0.0001,
+      "loss": 1.761,
+      "step": 4171
+    },
+    {
+      "epoch": 0.4846935811792042,
+      "grad_norm": 0.4458593428134918,
+      "learning_rate": 0.0001,
+      "loss": 1.7806,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4848097589311647,
+      "grad_norm": 0.4241911470890045,
+      "learning_rate": 0.0001,
+      "loss": 1.7529,
+      "step": 4173
+    },
+    {
+      "epoch": 0.48492593668312517,
+      "grad_norm": 0.43590041995048523,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 4174
+    },
+    {
+      "epoch": 0.48504211443508566,
+      "grad_norm": 0.4520244002342224,
+      "learning_rate": 0.0001,
+      "loss": 1.764,
+      "step": 4175
+    },
+    {
+      "epoch": 0.4851582921870462,
+      "grad_norm": 0.40647611021995544,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 4176
+    },
+    {
+      "epoch": 0.4852744699390067,
+      "grad_norm": 0.4175068438053131,
+      "learning_rate": 0.0001,
+      "loss": 1.7583,
+      "step": 4177
+    },
+    {
+      "epoch": 0.4853906476909672,
+      "grad_norm": 0.4129279553890228,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 4178
+    },
+    {
+      "epoch": 0.48550682544292767,
+      "grad_norm": 0.42088520526885986,
+      "learning_rate": 0.0001,
+      "loss": 1.756,
+      "step": 4179
+    },
+    {
+      "epoch": 0.48562300319488816,
+      "grad_norm": 0.4570264220237732,
+      "learning_rate": 0.0001,
+      "loss": 1.7849,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4857391809468487,
+      "grad_norm": 0.4151148200035095,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 4181
+    },
+    {
+      "epoch": 0.4858553586988092,
+      "grad_norm": 0.4177875518798828,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 4182
+    },
+    {
+      "epoch": 0.4859715364507697,
+      "grad_norm": 0.4347212016582489,
+      "learning_rate": 0.0001,
+      "loss": 1.6551,
+      "step": 4183
+    },
+    {
+      "epoch": 0.48608771420273017,
+      "grad_norm": 0.406599760055542,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 4184
+    },
+    {
+      "epoch": 0.48620389195469066,
+      "grad_norm": 0.4066656529903412,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 4185
+    },
+    {
+      "epoch": 0.4863200697066512,
+      "grad_norm": 0.4207620918750763,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 4186
+    },
+    {
+      "epoch": 0.4864362474586117,
+      "grad_norm": 0.43291810154914856,
+      "learning_rate": 0.0001,
+      "loss": 1.7537,
+      "step": 4187
+    },
+    {
+      "epoch": 0.4865524252105722,
+      "grad_norm": 0.44666996598243713,
+      "learning_rate": 0.0001,
+      "loss": 1.7468,
+      "step": 4188
+    },
+    {
+      "epoch": 0.48666860296253267,
+      "grad_norm": 0.394218772649765,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 4189
+    },
+    {
+      "epoch": 0.48678478071449316,
+      "grad_norm": 0.41351521015167236,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 4190
+    },
+    {
+      "epoch": 0.48690095846645365,
+      "grad_norm": 0.412653386592865,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 4191
+    },
+    {
+      "epoch": 0.4870171362184142,
+      "grad_norm": 0.4062814712524414,
+      "learning_rate": 0.0001,
+      "loss": 1.5941,
+      "step": 4192
+    },
+    {
+      "epoch": 0.4871333139703747,
+      "grad_norm": 0.41630494594573975,
+      "learning_rate": 0.0001,
+      "loss": 1.7655,
+      "step": 4193
+    },
+    {
+      "epoch": 0.48724949172233517,
+      "grad_norm": 0.41663774847984314,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 4194
+    },
+    {
+      "epoch": 0.48736566947429566,
+      "grad_norm": 0.42713698744773865,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 4195
+    },
+    {
+      "epoch": 0.48748184722625615,
+      "grad_norm": 0.43906137347221375,
+      "learning_rate": 0.0001,
+      "loss": 1.7459,
+      "step": 4196
+    },
+    {
+      "epoch": 0.4875980249782167,
+      "grad_norm": 0.3749426305294037,
+      "learning_rate": 0.0001,
+      "loss": 1.6356,
+      "step": 4197
+    },
+    {
+      "epoch": 0.4877142027301772,
+      "grad_norm": 0.5166771411895752,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 4198
+    },
+    {
+      "epoch": 0.48783038048213767,
+      "grad_norm": 0.42134204506874084,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 4199
+    },
+    {
+      "epoch": 0.48794655823409816,
+      "grad_norm": 0.4114912152290344,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 4200
+    },
+    {
+      "epoch": 0.48806273598605865,
+      "grad_norm": 0.4079913794994354,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 4201
+    },
+    {
+      "epoch": 0.4881789137380192,
+      "grad_norm": 0.4527676999568939,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 4202
+    },
+    {
+      "epoch": 0.4882950914899797,
+      "grad_norm": 0.4323418438434601,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 4203
+    },
+    {
+      "epoch": 0.48841126924194017,
+      "grad_norm": 0.43545210361480713,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 4204
+    },
+    {
+      "epoch": 0.48852744699390066,
+      "grad_norm": 0.4371805191040039,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 4205
+    },
+    {
+      "epoch": 0.48864362474586115,
+      "grad_norm": 0.42473122477531433,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 4206
+    },
+    {
+      "epoch": 0.4887598024978217,
+      "grad_norm": 0.42178717255592346,
+      "learning_rate": 0.0001,
+      "loss": 1.7745,
+      "step": 4207
+    },
+    {
+      "epoch": 0.4888759802497822,
+      "grad_norm": 0.39883777499198914,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 4208
+    },
+    {
+      "epoch": 0.48899215800174267,
+      "grad_norm": 0.39283865690231323,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 4209
+    },
+    {
+      "epoch": 0.48910833575370316,
+      "grad_norm": 0.390767902135849,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 4210
+    },
+    {
+      "epoch": 0.48922451350566365,
+      "grad_norm": 0.40290915966033936,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 4211
+    },
+    {
+      "epoch": 0.48934069125762414,
+      "grad_norm": 0.3867753744125366,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 4212
+    },
+    {
+      "epoch": 0.4894568690095847,
+      "grad_norm": 0.40270060300827026,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 4213
+    },
+    {
+      "epoch": 0.48957304676154517,
+      "grad_norm": 0.40884700417518616,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 4214
+    },
+    {
+      "epoch": 0.48968922451350566,
+      "grad_norm": 0.4496394395828247,
+      "learning_rate": 0.0001,
+      "loss": 1.7212,
+      "step": 4215
+    },
+    {
+      "epoch": 0.48980540226546615,
+      "grad_norm": 0.4228356182575226,
+      "learning_rate": 0.0001,
+      "loss": 1.7374,
+      "step": 4216
+    },
+    {
+      "epoch": 0.48992158001742664,
+      "grad_norm": 0.3831014633178711,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 4217
+    },
+    {
+      "epoch": 0.4900377577693872,
+      "grad_norm": 0.4408629834651947,
+      "learning_rate": 0.0001,
+      "loss": 1.661,
+      "step": 4218
+    },
+    {
+      "epoch": 0.4901539355213477,
+      "grad_norm": 0.4310610890388489,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 4219
+    },
+    {
+      "epoch": 0.49027011327330816,
+      "grad_norm": 0.4541778266429901,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 4220
+    },
+    {
+      "epoch": 0.49038629102526865,
+      "grad_norm": 0.4351007342338562,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 4221
+    },
+    {
+      "epoch": 0.49050246877722914,
+      "grad_norm": 0.4092995226383209,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 4222
+    },
+    {
+      "epoch": 0.4906186465291897,
+      "grad_norm": 0.41761350631713867,
+      "learning_rate": 0.0001,
+      "loss": 1.6881,
+      "step": 4223
+    },
+    {
+      "epoch": 0.4907348242811502,
+      "grad_norm": 0.42286214232444763,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 4224
+    },
+    {
+      "epoch": 0.49085100203311066,
+      "grad_norm": 0.44822847843170166,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 4225
+    },
+    {
+      "epoch": 0.49096717978507115,
+      "grad_norm": 0.40160536766052246,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 4226
+    },
+    {
+      "epoch": 0.49108335753703164,
+      "grad_norm": 0.39959850907325745,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 4227
+    },
+    {
+      "epoch": 0.4911995352889922,
+      "grad_norm": 0.4233490228652954,
+      "learning_rate": 0.0001,
+      "loss": 1.7888,
+      "step": 4228
+    },
+    {
+      "epoch": 0.4913157130409527,
+      "grad_norm": 0.41371530294418335,
+      "learning_rate": 0.0001,
+      "loss": 1.7971,
+      "step": 4229
+    },
+    {
+      "epoch": 0.49143189079291316,
+      "grad_norm": 0.4132305085659027,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 4230
+    },
+    {
+      "epoch": 0.49154806854487365,
+      "grad_norm": 0.4752264618873596,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 4231
+    },
+    {
+      "epoch": 0.49166424629683414,
+      "grad_norm": 0.37759026885032654,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 4232
+    },
+    {
+      "epoch": 0.49178042404879463,
+      "grad_norm": 0.423034131526947,
+      "learning_rate": 0.0001,
+      "loss": 1.7351,
+      "step": 4233
+    },
+    {
+      "epoch": 0.4918966018007552,
+      "grad_norm": 0.42618465423583984,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 4234
+    },
+    {
+      "epoch": 0.49201277955271566,
+      "grad_norm": 0.4361681044101715,
+      "learning_rate": 0.0001,
+      "loss": 1.7888,
+      "step": 4235
+    },
+    {
+      "epoch": 0.49212895730467615,
+      "grad_norm": 0.4297569692134857,
+      "learning_rate": 0.0001,
+      "loss": 1.8216,
+      "step": 4236
+    },
+    {
+      "epoch": 0.49224513505663664,
+      "grad_norm": 0.4081512987613678,
+      "learning_rate": 0.0001,
+      "loss": 1.6625,
+      "step": 4237
+    },
+    {
+      "epoch": 0.49236131280859713,
+      "grad_norm": 0.4123179614543915,
+      "learning_rate": 0.0001,
+      "loss": 1.7767,
+      "step": 4238
+    },
+    {
+      "epoch": 0.4924774905605577,
+      "grad_norm": 0.42507731914520264,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 4239
+    },
+    {
+      "epoch": 0.49259366831251816,
+      "grad_norm": 0.4431770145893097,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 4240
+    },
+    {
+      "epoch": 0.49270984606447865,
+      "grad_norm": 0.4066873788833618,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 4241
+    },
+    {
+      "epoch": 0.49282602381643914,
+      "grad_norm": 0.39505207538604736,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 4242
+    },
+    {
+      "epoch": 0.49294220156839963,
+      "grad_norm": 0.3953791856765747,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 4243
+    },
+    {
+      "epoch": 0.4930583793203602,
+      "grad_norm": 0.4070848822593689,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 4244
+    },
+    {
+      "epoch": 0.49317455707232066,
+      "grad_norm": 0.3953931927680969,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 4245
+    },
+    {
+      "epoch": 0.49329073482428115,
+      "grad_norm": 0.41489389538764954,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 4246
+    },
+    {
+      "epoch": 0.49340691257624164,
+      "grad_norm": 0.381734699010849,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 4247
+    },
+    {
+      "epoch": 0.49352309032820213,
+      "grad_norm": 0.4128166139125824,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 4248
+    },
+    {
+      "epoch": 0.4936392680801627,
+      "grad_norm": 0.4323165714740753,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 4249
+    },
+    {
+      "epoch": 0.49375544583212316,
+      "grad_norm": 0.4313991069793701,
+      "learning_rate": 0.0001,
+      "loss": 1.74,
+      "step": 4250
+    },
+    {
+      "epoch": 0.49387162358408365,
+      "grad_norm": 0.41360074281692505,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 4251
+    },
+    {
+      "epoch": 0.49398780133604414,
+      "grad_norm": 0.41541731357574463,
+      "learning_rate": 0.0001,
+      "loss": 1.7381,
+      "step": 4252
+    },
+    {
+      "epoch": 0.49410397908800463,
+      "grad_norm": 0.3932528495788574,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 4253
+    },
+    {
+      "epoch": 0.4942201568399651,
+      "grad_norm": 0.41322755813598633,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 4254
+    },
+    {
+      "epoch": 0.49433633459192566,
+      "grad_norm": 0.39592114090919495,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 4255
+    },
+    {
+      "epoch": 0.49445251234388615,
+      "grad_norm": 0.4194466173648834,
+      "learning_rate": 0.0001,
+      "loss": 1.7409,
+      "step": 4256
+    },
+    {
+      "epoch": 0.49456869009584664,
+      "grad_norm": 0.4130099415779114,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 4257
+    },
+    {
+      "epoch": 0.49468486784780713,
+      "grad_norm": 0.39909616112709045,
+      "learning_rate": 0.0001,
+      "loss": 1.3854,
+      "step": 4258
+    },
+    {
+      "epoch": 0.4948010455997676,
+      "grad_norm": 0.4121648371219635,
+      "learning_rate": 0.0001,
+      "loss": 1.6729,
+      "step": 4259
+    },
+    {
+      "epoch": 0.49491722335172816,
+      "grad_norm": 0.4459930956363678,
+      "learning_rate": 0.0001,
+      "loss": 1.8308,
+      "step": 4260
+    },
+    {
+      "epoch": 0.49503340110368865,
+      "grad_norm": 0.42174288630485535,
+      "learning_rate": 0.0001,
+      "loss": 1.7415,
+      "step": 4261
+    },
+    {
+      "epoch": 0.49514957885564914,
+      "grad_norm": 0.4145379364490509,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 4262
+    },
+    {
+      "epoch": 0.49526575660760963,
+      "grad_norm": 0.40851959586143494,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 4263
+    },
+    {
+      "epoch": 0.4953819343595701,
+      "grad_norm": 0.42193418741226196,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 4264
+    },
+    {
+      "epoch": 0.49549811211153066,
+      "grad_norm": 0.4060492515563965,
+      "learning_rate": 0.0001,
+      "loss": 1.7783,
+      "step": 4265
+    },
+    {
+      "epoch": 0.49561428986349115,
+      "grad_norm": 0.3899538815021515,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 4266
+    },
+    {
+      "epoch": 0.49573046761545164,
+      "grad_norm": 0.42852500081062317,
+      "learning_rate": 0.0001,
+      "loss": 1.7705,
+      "step": 4267
+    },
+    {
+      "epoch": 0.49584664536741213,
+      "grad_norm": 0.41384875774383545,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 4268
+    },
+    {
+      "epoch": 0.4959628231193726,
+      "grad_norm": 0.41500815749168396,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 4269
+    },
+    {
+      "epoch": 0.49607900087133316,
+      "grad_norm": 0.40487122535705566,
+      "learning_rate": 0.0001,
+      "loss": 1.5961,
+      "step": 4270
+    },
+    {
+      "epoch": 0.49619517862329365,
+      "grad_norm": 0.4144662022590637,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 4271
+    },
+    {
+      "epoch": 0.49631135637525414,
+      "grad_norm": 0.40673938393592834,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 4272
+    },
+    {
+      "epoch": 0.49642753412721463,
+      "grad_norm": 0.4252079725265503,
+      "learning_rate": 0.0001,
+      "loss": 1.5973,
+      "step": 4273
+    },
+    {
+      "epoch": 0.4965437118791751,
+      "grad_norm": 0.4067811071872711,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 4274
+    },
+    {
+      "epoch": 0.49665988963113566,
+      "grad_norm": 0.4027066230773926,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 4275
+    },
+    {
+      "epoch": 0.49677606738309615,
+      "grad_norm": 0.4006246030330658,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 4276
+    },
+    {
+      "epoch": 0.49689224513505664,
+      "grad_norm": 0.42645466327667236,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 4277
+    },
+    {
+      "epoch": 0.49700842288701713,
+      "grad_norm": 0.3970547616481781,
+      "learning_rate": 0.0001,
+      "loss": 1.71,
+      "step": 4278
+    },
+    {
+      "epoch": 0.4971246006389776,
+      "grad_norm": 0.43710431456565857,
+      "learning_rate": 0.0001,
+      "loss": 1.8507,
+      "step": 4279
+    },
+    {
+      "epoch": 0.4972407783909381,
+      "grad_norm": 0.4387757182121277,
+      "learning_rate": 0.0001,
+      "loss": 1.7658,
+      "step": 4280
+    },
+    {
+      "epoch": 0.49735695614289865,
+      "grad_norm": 0.3968219757080078,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 4281
+    },
+    {
+      "epoch": 0.49747313389485914,
+      "grad_norm": 0.42177560925483704,
+      "learning_rate": 0.0001,
+      "loss": 1.6699,
+      "step": 4282
+    },
+    {
+      "epoch": 0.49758931164681963,
+      "grad_norm": 0.46521419286727905,
+      "learning_rate": 0.0001,
+      "loss": 1.8792,
+      "step": 4283
+    },
+    {
+      "epoch": 0.4977054893987801,
+      "grad_norm": 0.45769378542900085,
+      "learning_rate": 0.0001,
+      "loss": 1.9134,
+      "step": 4284
+    },
+    {
+      "epoch": 0.4978216671507406,
+      "grad_norm": 0.415170282125473,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 4285
+    },
+    {
+      "epoch": 0.49793784490270115,
+      "grad_norm": 0.4076900780200958,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 4286
+    },
+    {
+      "epoch": 0.49805402265466164,
+      "grad_norm": 0.42825424671173096,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 4287
+    },
+    {
+      "epoch": 0.49817020040662213,
+      "grad_norm": 0.39369601011276245,
+      "learning_rate": 0.0001,
+      "loss": 1.5995,
+      "step": 4288
+    },
+    {
+      "epoch": 0.4982863781585826,
+      "grad_norm": 0.415863037109375,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 4289
+    },
+    {
+      "epoch": 0.4984025559105431,
+      "grad_norm": 0.4149499833583832,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 4290
+    },
+    {
+      "epoch": 0.49851873366250365,
+      "grad_norm": 0.4311138391494751,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 4291
+    },
+    {
+      "epoch": 0.49863491141446414,
+      "grad_norm": 0.4341351389884949,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 4292
+    },
+    {
+      "epoch": 0.49875108916642463,
+      "grad_norm": 0.43341371417045593,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 4293
+    },
+    {
+      "epoch": 0.4988672669183851,
+      "grad_norm": 0.4445037543773651,
+      "learning_rate": 0.0001,
+      "loss": 1.8165,
+      "step": 4294
+    },
+    {
+      "epoch": 0.4989834446703456,
+      "grad_norm": 0.4031905233860016,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 4295
+    },
+    {
+      "epoch": 0.49909962242230616,
+      "grad_norm": 0.3780546188354492,
+      "learning_rate": 0.0001,
+      "loss": 1.3956,
+      "step": 4296
+    },
+    {
+      "epoch": 0.49921580017426664,
+      "grad_norm": 0.4312121868133545,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 4297
+    },
+    {
+      "epoch": 0.49933197792622713,
+      "grad_norm": 0.42564424872398376,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 4298
+    },
+    {
+      "epoch": 0.4994481556781876,
+      "grad_norm": 0.41907092928886414,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 4299
+    },
+    {
+      "epoch": 0.4995643334301481,
+      "grad_norm": 0.44234657287597656,
+      "learning_rate": 0.0001,
+      "loss": 1.8374,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4996805111821086,
+      "grad_norm": 0.43759170174598694,
+      "learning_rate": 0.0001,
+      "loss": 1.7536,
+      "step": 4301
+    },
+    {
+      "epoch": 0.49979668893406914,
+      "grad_norm": 0.4182339012622833,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 4302
+    },
+    {
+      "epoch": 0.49991286668602963,
+      "grad_norm": 0.4217366874217987,
+      "learning_rate": 0.0001,
+      "loss": 1.7919,
+      "step": 4303
+    },
+    {
+      "epoch": 0.5000290444379901,
+      "grad_norm": 0.4132644236087799,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 4304
+    },
+    {
+      "epoch": 0.5001452221899506,
+      "grad_norm": 0.4324856996536255,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 4305
+    },
+    {
+      "epoch": 0.5002613999419111,
+      "grad_norm": 0.4039803445339203,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 4306
+    },
+    {
+      "epoch": 0.5003775776938716,
+      "grad_norm": 0.4950650930404663,
+      "learning_rate": 0.0001,
+      "loss": 1.8132,
+      "step": 4307
+    },
+    {
+      "epoch": 0.5004937554458321,
+      "grad_norm": 0.4059164524078369,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 4308
+    },
+    {
+      "epoch": 0.5006099331977927,
+      "grad_norm": 0.4175871014595032,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 4309
+    },
+    {
+      "epoch": 0.5007261109497532,
+      "grad_norm": 0.40468019247055054,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 4310
+    },
+    {
+      "epoch": 0.5008422887017137,
+      "grad_norm": 0.45055514574050903,
+      "learning_rate": 0.0001,
+      "loss": 1.8287,
+      "step": 4311
+    },
+    {
+      "epoch": 0.5009584664536741,
+      "grad_norm": 0.4032573997974396,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 4312
+    },
+    {
+      "epoch": 0.5010746442056346,
+      "grad_norm": 0.4106045067310333,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 4313
+    },
+    {
+      "epoch": 0.5011908219575951,
+      "grad_norm": 0.4448246955871582,
+      "learning_rate": 0.0001,
+      "loss": 1.6811,
+      "step": 4314
+    },
+    {
+      "epoch": 0.5013069997095556,
+      "grad_norm": 0.45284712314605713,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 4315
+    },
+    {
+      "epoch": 0.5014231774615161,
+      "grad_norm": 0.4348433017730713,
+      "learning_rate": 0.0001,
+      "loss": 1.6896,
+      "step": 4316
+    },
+    {
+      "epoch": 0.5015393552134766,
+      "grad_norm": 0.4291326701641083,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 4317
+    },
+    {
+      "epoch": 0.5016555329654371,
+      "grad_norm": 0.42308685183525085,
+      "learning_rate": 0.0001,
+      "loss": 1.7336,
+      "step": 4318
+    },
+    {
+      "epoch": 0.5017717107173976,
+      "grad_norm": 0.4419673681259155,
+      "learning_rate": 0.0001,
+      "loss": 1.8071,
+      "step": 4319
+    },
+    {
+      "epoch": 0.5018878884693582,
+      "grad_norm": 0.4059154987335205,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 4320
+    },
+    {
+      "epoch": 0.5020040662213187,
+      "grad_norm": 0.44748127460479736,
+      "learning_rate": 0.0001,
+      "loss": 1.8027,
+      "step": 4321
+    },
+    {
+      "epoch": 0.5021202439732791,
+      "grad_norm": 0.4353194832801819,
+      "learning_rate": 0.0001,
+      "loss": 1.77,
+      "step": 4322
+    },
+    {
+      "epoch": 0.5022364217252396,
+      "grad_norm": 0.43615394830703735,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 4323
+    },
+    {
+      "epoch": 0.5023525994772001,
+      "grad_norm": 0.4044416546821594,
+      "learning_rate": 0.0001,
+      "loss": 1.4778,
+      "step": 4324
+    },
+    {
+      "epoch": 0.5024687772291606,
+      "grad_norm": 0.4381449222564697,
+      "learning_rate": 0.0001,
+      "loss": 1.6863,
+      "step": 4325
+    },
+    {
+      "epoch": 0.5025849549811211,
+      "grad_norm": 0.41832980513572693,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 4326
+    },
+    {
+      "epoch": 0.5027011327330816,
+      "grad_norm": 0.41191017627716064,
+      "learning_rate": 0.0001,
+      "loss": 1.7425,
+      "step": 4327
+    },
+    {
+      "epoch": 0.5028173104850421,
+      "grad_norm": 0.4307502508163452,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 4328
+    },
+    {
+      "epoch": 0.5029334882370026,
+      "grad_norm": 0.39051318168640137,
+      "learning_rate": 0.0001,
+      "loss": 1.4484,
+      "step": 4329
+    },
+    {
+      "epoch": 0.5030496659889632,
+      "grad_norm": 0.4072178602218628,
+      "learning_rate": 0.0001,
+      "loss": 1.6725,
+      "step": 4330
+    },
+    {
+      "epoch": 0.5031658437409237,
+      "grad_norm": 0.4136503040790558,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 4331
+    },
+    {
+      "epoch": 0.5032820214928841,
+      "grad_norm": 0.4073225259780884,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 4332
+    },
+    {
+      "epoch": 0.5033981992448446,
+      "grad_norm": 0.45762255787849426,
+      "learning_rate": 0.0001,
+      "loss": 1.8642,
+      "step": 4333
+    },
+    {
+      "epoch": 0.5035143769968051,
+      "grad_norm": 0.3969933092594147,
+      "learning_rate": 0.0001,
+      "loss": 1.4036,
+      "step": 4334
+    },
+    {
+      "epoch": 0.5036305547487656,
+      "grad_norm": 0.4575958251953125,
+      "learning_rate": 0.0001,
+      "loss": 1.7049,
+      "step": 4335
+    },
+    {
+      "epoch": 0.5037467325007261,
+      "grad_norm": 0.4271569550037384,
+      "learning_rate": 0.0001,
+      "loss": 1.786,
+      "step": 4336
+    },
+    {
+      "epoch": 0.5038629102526866,
+      "grad_norm": 0.41905540227890015,
+      "learning_rate": 0.0001,
+      "loss": 1.4688,
+      "step": 4337
+    },
+    {
+      "epoch": 0.5039790880046471,
+      "grad_norm": 0.4295593202114105,
+      "learning_rate": 0.0001,
+      "loss": 1.7429,
+      "step": 4338
+    },
+    {
+      "epoch": 0.5040952657566076,
+      "grad_norm": 0.4510742425918579,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 4339
+    },
+    {
+      "epoch": 0.5042114435085681,
+      "grad_norm": 0.38910943269729614,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 4340
+    },
+    {
+      "epoch": 0.5043276212605287,
+      "grad_norm": 0.43599119782447815,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 4341
+    },
+    {
+      "epoch": 0.5044437990124891,
+      "grad_norm": 0.4247783124446869,
+      "learning_rate": 0.0001,
+      "loss": 1.7167,
+      "step": 4342
+    },
+    {
+      "epoch": 0.5045599767644496,
+      "grad_norm": 0.41217729449272156,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 4343
+    },
+    {
+      "epoch": 0.5046761545164101,
+      "grad_norm": 0.4111458361148834,
+      "learning_rate": 0.0001,
+      "loss": 1.7107,
+      "step": 4344
+    },
+    {
+      "epoch": 0.5047923322683706,
+      "grad_norm": 0.44936275482177734,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 4345
+    },
+    {
+      "epoch": 0.5049085100203311,
+      "grad_norm": 0.39809176325798035,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 4346
+    },
+    {
+      "epoch": 0.5050246877722916,
+      "grad_norm": 0.45946526527404785,
+      "learning_rate": 0.0001,
+      "loss": 1.7883,
+      "step": 4347
+    },
+    {
+      "epoch": 0.5051408655242521,
+      "grad_norm": 0.4194871485233307,
+      "learning_rate": 0.0001,
+      "loss": 1.6288,
+      "step": 4348
+    },
+    {
+      "epoch": 0.5052570432762126,
+      "grad_norm": 0.4171569049358368,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 4349
+    },
+    {
+      "epoch": 0.5053732210281731,
+      "grad_norm": 0.4225428104400635,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 4350
+    },
+    {
+      "epoch": 0.5054893987801337,
+      "grad_norm": 0.410523921251297,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 4351
+    },
+    {
+      "epoch": 0.5056055765320941,
+      "grad_norm": 0.4116719663143158,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 4352
+    },
+    {
+      "epoch": 0.5057217542840546,
+      "grad_norm": 0.405364453792572,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 4353
+    },
+    {
+      "epoch": 0.5058379320360151,
+      "grad_norm": 0.38687193393707275,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 4354
+    },
+    {
+      "epoch": 0.5059541097879756,
+      "grad_norm": 0.4160784184932709,
+      "learning_rate": 0.0001,
+      "loss": 1.6666,
+      "step": 4355
+    },
+    {
+      "epoch": 0.5060702875399361,
+      "grad_norm": 0.4402058720588684,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 4356
+    },
+    {
+      "epoch": 0.5061864652918966,
+      "grad_norm": 0.4149058759212494,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 4357
+    },
+    {
+      "epoch": 0.5063026430438571,
+      "grad_norm": 0.39352482557296753,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 4358
+    },
+    {
+      "epoch": 0.5064188207958176,
+      "grad_norm": 0.3932999074459076,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 4359
+    },
+    {
+      "epoch": 0.5065349985477781,
+      "grad_norm": 0.41762053966522217,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 4360
+    },
+    {
+      "epoch": 0.5066511762997385,
+      "grad_norm": 0.39525556564331055,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 4361
+    },
+    {
+      "epoch": 0.5067673540516991,
+      "grad_norm": 0.41103050112724304,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 4362
+    },
+    {
+      "epoch": 0.5068835318036596,
+      "grad_norm": 0.4429750144481659,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 4363
+    },
+    {
+      "epoch": 0.5069997095556201,
+      "grad_norm": 0.42924562096595764,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 4364
+    },
+    {
+      "epoch": 0.5071158873075806,
+      "grad_norm": 0.4067068099975586,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 4365
+    },
+    {
+      "epoch": 0.5072320650595411,
+      "grad_norm": 0.42301809787750244,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 4366
+    },
+    {
+      "epoch": 0.5073482428115016,
+      "grad_norm": 0.4366927146911621,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 4367
+    },
+    {
+      "epoch": 0.5074644205634621,
+      "grad_norm": 0.48699912428855896,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 4368
+    },
+    {
+      "epoch": 0.5075805983154226,
+      "grad_norm": 0.4013606309890747,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 4369
+    },
+    {
+      "epoch": 0.5076967760673831,
+      "grad_norm": 0.42257198691368103,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 4370
+    },
+    {
+      "epoch": 0.5078129538193435,
+      "grad_norm": 0.41376587748527527,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 4371
+    },
+    {
+      "epoch": 0.5079291315713041,
+      "grad_norm": 0.42543458938598633,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 4372
+    },
+    {
+      "epoch": 0.5080453093232646,
+      "grad_norm": 0.40931135416030884,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 4373
+    },
+    {
+      "epoch": 0.5081614870752251,
+      "grad_norm": 0.3892346918582916,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 4374
+    },
+    {
+      "epoch": 0.5082776648271856,
+      "grad_norm": 0.3831053674221039,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 4375
+    },
+    {
+      "epoch": 0.5083938425791461,
+      "grad_norm": 0.3641822338104248,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 4376
+    },
+    {
+      "epoch": 0.5085100203311066,
+      "grad_norm": 0.40960368514060974,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 4377
+    },
+    {
+      "epoch": 0.5086261980830671,
+      "grad_norm": 0.42281374335289,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 4378
+    },
+    {
+      "epoch": 0.5087423758350276,
+      "grad_norm": 0.4280136227607727,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 4379
+    },
+    {
+      "epoch": 0.5088585535869881,
+      "grad_norm": 0.4118686020374298,
+      "learning_rate": 0.0001,
+      "loss": 1.6772,
+      "step": 4380
+    },
+    {
+      "epoch": 0.5089747313389485,
+      "grad_norm": 0.45106491446495056,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 4381
+    },
+    {
+      "epoch": 0.509090909090909,
+      "grad_norm": 0.3965935707092285,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 4382
+    },
+    {
+      "epoch": 0.5092070868428696,
+      "grad_norm": 0.3852759003639221,
+      "learning_rate": 0.0001,
+      "loss": 1.3832,
+      "step": 4383
+    },
+    {
+      "epoch": 0.5093232645948301,
+      "grad_norm": 0.4219793677330017,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 4384
+    },
+    {
+      "epoch": 0.5094394423467906,
+      "grad_norm": 0.4372718632221222,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 4385
+    },
+    {
+      "epoch": 0.5095556200987511,
+      "grad_norm": 0.40098169445991516,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 4386
+    },
+    {
+      "epoch": 0.5096717978507116,
+      "grad_norm": 0.4264896810054779,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 4387
+    },
+    {
+      "epoch": 0.5097879756026721,
+      "grad_norm": 0.4068213105201721,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 4388
+    },
+    {
+      "epoch": 0.5099041533546326,
+      "grad_norm": 0.4088718593120575,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 4389
+    },
+    {
+      "epoch": 0.5100203311065931,
+      "grad_norm": 0.42276012897491455,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 4390
+    },
+    {
+      "epoch": 0.5101365088585536,
+      "grad_norm": 0.4298572540283203,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 4391
+    },
+    {
+      "epoch": 0.510252686610514,
+      "grad_norm": 0.4012244939804077,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 4392
+    },
+    {
+      "epoch": 0.5103688643624746,
+      "grad_norm": 0.41358691453933716,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 4393
+    },
+    {
+      "epoch": 0.5104850421144351,
+      "grad_norm": 0.41776931285858154,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 4394
+    },
+    {
+      "epoch": 0.5106012198663956,
+      "grad_norm": 0.4451960325241089,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 4395
+    },
+    {
+      "epoch": 0.5107173976183561,
+      "grad_norm": 0.43334871530532837,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 4396
+    },
+    {
+      "epoch": 0.5108335753703166,
+      "grad_norm": 0.3872377276420593,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 4397
+    },
+    {
+      "epoch": 0.5109497531222771,
+      "grad_norm": 0.3914475440979004,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 4398
+    },
+    {
+      "epoch": 0.5110659308742376,
+      "grad_norm": 0.447561115026474,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 4399
+    },
+    {
+      "epoch": 0.5111821086261981,
+      "grad_norm": 0.45441824197769165,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 4400
+    },
+    {
+      "epoch": 0.5112982863781586,
+      "grad_norm": 0.3847355544567108,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 4401
+    },
+    {
+      "epoch": 0.511414464130119,
+      "grad_norm": 0.41067802906036377,
+      "learning_rate": 0.0001,
+      "loss": 1.6448,
+      "step": 4402
+    },
+    {
+      "epoch": 0.5115306418820795,
+      "grad_norm": 0.4514882564544678,
+      "learning_rate": 0.0001,
+      "loss": 1.792,
+      "step": 4403
+    },
+    {
+      "epoch": 0.5116468196340401,
+      "grad_norm": 0.4275973439216614,
+      "learning_rate": 0.0001,
+      "loss": 1.7367,
+      "step": 4404
+    },
+    {
+      "epoch": 0.5117629973860006,
+      "grad_norm": 0.41547560691833496,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 4405
+    },
+    {
+      "epoch": 0.5118791751379611,
+      "grad_norm": 0.40632376074790955,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 4406
+    },
+    {
+      "epoch": 0.5119953528899216,
+      "grad_norm": 0.4488579034805298,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 4407
+    },
+    {
+      "epoch": 0.5121115306418821,
+      "grad_norm": 0.42902886867523193,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 4408
+    },
+    {
+      "epoch": 0.5122277083938426,
+      "grad_norm": 0.42067989706993103,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 4409
+    },
+    {
+      "epoch": 0.5123438861458031,
+      "grad_norm": 0.4122752845287323,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 4410
+    },
+    {
+      "epoch": 0.5124600638977636,
+      "grad_norm": 0.443730890750885,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 4411
+    },
+    {
+      "epoch": 0.512576241649724,
+      "grad_norm": 0.43061313033103943,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 4412
+    },
+    {
+      "epoch": 0.5126924194016845,
+      "grad_norm": 0.42542704939842224,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 4413
+    },
+    {
+      "epoch": 0.5128085971536451,
+      "grad_norm": 0.41993504762649536,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 4414
+    },
+    {
+      "epoch": 0.5129247749056056,
+      "grad_norm": 0.43750661611557007,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 4415
+    },
+    {
+      "epoch": 0.5130409526575661,
+      "grad_norm": 0.4423817992210388,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 4416
+    },
+    {
+      "epoch": 0.5131571304095266,
+      "grad_norm": 0.44078147411346436,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 4417
+    },
+    {
+      "epoch": 0.5132733081614871,
+      "grad_norm": 0.41512519121170044,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 4418
+    },
+    {
+      "epoch": 0.5133894859134476,
+      "grad_norm": 0.4335169792175293,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 4419
+    },
+    {
+      "epoch": 0.5135056636654081,
+      "grad_norm": 0.39206641912460327,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 4420
+    },
+    {
+      "epoch": 0.5136218414173686,
+      "grad_norm": 0.4225231111049652,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 4421
+    },
+    {
+      "epoch": 0.513738019169329,
+      "grad_norm": 0.4060937762260437,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 4422
+    },
+    {
+      "epoch": 0.5138541969212895,
+      "grad_norm": 0.4321762025356293,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 4423
+    },
+    {
+      "epoch": 0.51397037467325,
+      "grad_norm": 0.4687374234199524,
+      "learning_rate": 0.0001,
+      "loss": 1.7355,
+      "step": 4424
+    },
+    {
+      "epoch": 0.5140865524252106,
+      "grad_norm": 0.4362178146839142,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 4425
+    },
+    {
+      "epoch": 0.5142027301771711,
+      "grad_norm": 0.42870602011680603,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 4426
+    },
+    {
+      "epoch": 0.5143189079291316,
+      "grad_norm": 0.4190438985824585,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 4427
+    },
+    {
+      "epoch": 0.5144350856810921,
+      "grad_norm": 0.43848124146461487,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 4428
+    },
+    {
+      "epoch": 0.5145512634330526,
+      "grad_norm": 4.445634365081787,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 4429
+    },
+    {
+      "epoch": 0.5146674411850131,
+      "grad_norm": 0.43199974298477173,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 4430
+    },
+    {
+      "epoch": 0.5147836189369736,
+      "grad_norm": 0.39670631289482117,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 4431
+    },
+    {
+      "epoch": 0.514899796688934,
+      "grad_norm": 0.4087604880332947,
+      "learning_rate": 0.0001,
+      "loss": 1.6337,
+      "step": 4432
+    },
+    {
+      "epoch": 0.5150159744408945,
+      "grad_norm": 0.4322669804096222,
+      "learning_rate": 0.0001,
+      "loss": 1.7526,
+      "step": 4433
+    },
+    {
+      "epoch": 0.515132152192855,
+      "grad_norm": 0.40673530101776123,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 4434
+    },
+    {
+      "epoch": 0.5152483299448156,
+      "grad_norm": 0.4231918454170227,
+      "learning_rate": 0.0001,
+      "loss": 1.7449,
+      "step": 4435
+    },
+    {
+      "epoch": 0.5153645076967761,
+      "grad_norm": 0.42139682173728943,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 4436
+    },
+    {
+      "epoch": 0.5154806854487366,
+      "grad_norm": 0.4375002384185791,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 4437
+    },
+    {
+      "epoch": 0.5155968632006971,
+      "grad_norm": 0.41837188601493835,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 4438
+    },
+    {
+      "epoch": 0.5157130409526576,
+      "grad_norm": 0.36996695399284363,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 4439
+    },
+    {
+      "epoch": 0.5158292187046181,
+      "grad_norm": 0.428469181060791,
+      "learning_rate": 0.0001,
+      "loss": 1.7409,
+      "step": 4440
+    },
+    {
+      "epoch": 0.5159453964565786,
+      "grad_norm": 0.3917810320854187,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 4441
+    },
+    {
+      "epoch": 0.516061574208539,
+      "grad_norm": 0.41308334469795227,
+      "learning_rate": 0.0001,
+      "loss": 1.845,
+      "step": 4442
+    },
+    {
+      "epoch": 0.5161777519604995,
+      "grad_norm": 0.41742420196533203,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 4443
+    },
+    {
+      "epoch": 0.51629392971246,
+      "grad_norm": 0.42206430435180664,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 4444
+    },
+    {
+      "epoch": 0.5164101074644205,
+      "grad_norm": 0.41111844778060913,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 4445
+    },
+    {
+      "epoch": 0.5165262852163811,
+      "grad_norm": 0.42969468235969543,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 4446
+    },
+    {
+      "epoch": 0.5166424629683416,
+      "grad_norm": 0.4495762288570404,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 4447
+    },
+    {
+      "epoch": 0.5167586407203021,
+      "grad_norm": 0.41785508394241333,
+      "learning_rate": 0.0001,
+      "loss": 1.4878,
+      "step": 4448
+    },
+    {
+      "epoch": 0.5168748184722626,
+      "grad_norm": 0.4101484715938568,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 4449
+    },
+    {
+      "epoch": 0.5169909962242231,
+      "grad_norm": 0.422985315322876,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 4450
+    },
+    {
+      "epoch": 0.5171071739761836,
+      "grad_norm": 0.45259571075439453,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 4451
+    },
+    {
+      "epoch": 0.517223351728144,
+      "grad_norm": 0.3940257728099823,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 4452
+    },
+    {
+      "epoch": 0.5173395294801045,
+      "grad_norm": 0.41753649711608887,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 4453
+    },
+    {
+      "epoch": 0.517455707232065,
+      "grad_norm": 0.4366135895252228,
+      "learning_rate": 0.0001,
+      "loss": 1.7993,
+      "step": 4454
+    },
+    {
+      "epoch": 0.5175718849840255,
+      "grad_norm": 0.43286582827568054,
+      "learning_rate": 0.0001,
+      "loss": 1.7569,
+      "step": 4455
+    },
+    {
+      "epoch": 0.5176880627359861,
+      "grad_norm": 0.4226556420326233,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 4456
+    },
+    {
+      "epoch": 0.5178042404879466,
+      "grad_norm": 0.42732787132263184,
+      "learning_rate": 0.0001,
+      "loss": 1.7325,
+      "step": 4457
+    },
+    {
+      "epoch": 0.5179204182399071,
+      "grad_norm": 0.4511219561100006,
+      "learning_rate": 0.0001,
+      "loss": 1.6626,
+      "step": 4458
+    },
+    {
+      "epoch": 0.5180365959918676,
+      "grad_norm": 0.418562114238739,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 4459
+    },
+    {
+      "epoch": 0.5181527737438281,
+      "grad_norm": 0.4227406978607178,
+      "learning_rate": 0.0001,
+      "loss": 1.6733,
+      "step": 4460
+    },
+    {
+      "epoch": 0.5182689514957886,
+      "grad_norm": 0.4748642146587372,
+      "learning_rate": 0.0001,
+      "loss": 1.8498,
+      "step": 4461
+    },
+    {
+      "epoch": 0.518385129247749,
+      "grad_norm": 0.4098576605319977,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 4462
+    },
+    {
+      "epoch": 0.5185013069997095,
+      "grad_norm": 0.3687531054019928,
+      "learning_rate": 0.0001,
+      "loss": 1.4375,
+      "step": 4463
+    },
+    {
+      "epoch": 0.51861748475167,
+      "grad_norm": 0.43809184432029724,
+      "learning_rate": 0.0001,
+      "loss": 1.6958,
+      "step": 4464
+    },
+    {
+      "epoch": 0.5187336625036305,
+      "grad_norm": 0.4355543851852417,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 4465
+    },
+    {
+      "epoch": 0.518849840255591,
+      "grad_norm": 0.40263766050338745,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 4466
+    },
+    {
+      "epoch": 0.5189660180075516,
+      "grad_norm": 0.40483129024505615,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 4467
+    },
+    {
+      "epoch": 0.5190821957595121,
+      "grad_norm": 0.4438629150390625,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 4468
+    },
+    {
+      "epoch": 0.5191983735114726,
+      "grad_norm": 0.417834997177124,
+      "learning_rate": 0.0001,
+      "loss": 1.5252,
+      "step": 4469
+    },
+    {
+      "epoch": 0.5193145512634331,
+      "grad_norm": 0.4064665138721466,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 4470
+    },
+    {
+      "epoch": 0.5194307290153936,
+      "grad_norm": 0.4300839900970459,
+      "learning_rate": 0.0001,
+      "loss": 1.7744,
+      "step": 4471
+    },
+    {
+      "epoch": 0.519546906767354,
+      "grad_norm": 0.43153315782546997,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 4472
+    },
+    {
+      "epoch": 0.5196630845193145,
+      "grad_norm": 0.40210986137390137,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 4473
+    },
+    {
+      "epoch": 0.519779262271275,
+      "grad_norm": 0.4261165261268616,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 4474
+    },
+    {
+      "epoch": 0.5198954400232355,
+      "grad_norm": 0.40189406275749207,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 4475
+    },
+    {
+      "epoch": 0.520011617775196,
+      "grad_norm": 0.40366873145103455,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 4476
+    },
+    {
+      "epoch": 0.5201277955271566,
+      "grad_norm": 0.39482223987579346,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 4477
+    },
+    {
+      "epoch": 0.5202439732791171,
+      "grad_norm": 0.4213715195655823,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 4478
+    },
+    {
+      "epoch": 0.5203601510310776,
+      "grad_norm": 0.412153959274292,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 4479
+    },
+    {
+      "epoch": 0.5204763287830381,
+      "grad_norm": 0.4511539041996002,
+      "learning_rate": 0.0001,
+      "loss": 1.7117,
+      "step": 4480
+    },
+    {
+      "epoch": 0.5205925065349986,
+      "grad_norm": 0.4023229777812958,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 4481
+    },
+    {
+      "epoch": 0.520708684286959,
+      "grad_norm": 0.410552054643631,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 4482
+    },
+    {
+      "epoch": 0.5208248620389195,
+      "grad_norm": 0.4153655767440796,
+      "learning_rate": 0.0001,
+      "loss": 1.7299,
+      "step": 4483
+    },
+    {
+      "epoch": 0.52094103979088,
+      "grad_norm": 0.4262380599975586,
+      "learning_rate": 0.0001,
+      "loss": 1.7457,
+      "step": 4484
+    },
+    {
+      "epoch": 0.5210572175428405,
+      "grad_norm": 0.38934075832366943,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 4485
+    },
+    {
+      "epoch": 0.521173395294801,
+      "grad_norm": 0.417797327041626,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 4486
+    },
+    {
+      "epoch": 0.5212895730467616,
+      "grad_norm": 0.45602330565452576,
+      "learning_rate": 0.0001,
+      "loss": 1.7248,
+      "step": 4487
+    },
+    {
+      "epoch": 0.5214057507987221,
+      "grad_norm": 0.3888321816921234,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 4488
+    },
+    {
+      "epoch": 0.5215219285506826,
+      "grad_norm": 0.4114174246788025,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 4489
+    },
+    {
+      "epoch": 0.5216381063026431,
+      "grad_norm": 0.43008658289909363,
+      "learning_rate": 0.0001,
+      "loss": 1.6818,
+      "step": 4490
+    },
+    {
+      "epoch": 0.5217542840546036,
+      "grad_norm": 0.40249064564704895,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 4491
+    },
+    {
+      "epoch": 0.521870461806564,
+      "grad_norm": 0.39665487408638,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 4492
+    },
+    {
+      "epoch": 0.5219866395585245,
+      "grad_norm": 0.46716028451919556,
+      "learning_rate": 0.0001,
+      "loss": 1.8122,
+      "step": 4493
+    },
+    {
+      "epoch": 0.522102817310485,
+      "grad_norm": 0.4550071954727173,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 4494
+    },
+    {
+      "epoch": 0.5222189950624455,
+      "grad_norm": 0.40407729148864746,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 4495
+    },
+    {
+      "epoch": 0.522335172814406,
+      "grad_norm": 0.4077194035053253,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 4496
+    },
+    {
+      "epoch": 0.5224513505663665,
+      "grad_norm": 0.4374626576900482,
+      "learning_rate": 0.0001,
+      "loss": 1.8369,
+      "step": 4497
+    },
+    {
+      "epoch": 0.5225675283183271,
+      "grad_norm": 0.44101259112358093,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 4498
+    },
+    {
+      "epoch": 0.5226837060702876,
+      "grad_norm": 0.45485249161720276,
+      "learning_rate": 0.0001,
+      "loss": 1.619,
+      "step": 4499
+    },
+    {
+      "epoch": 0.5227998838222481,
+      "grad_norm": 0.4245244860649109,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 4500
+    },
+    {
+      "epoch": 0.5229160615742086,
+      "grad_norm": 0.4580254554748535,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 4501
+    },
+    {
+      "epoch": 0.523032239326169,
+      "grad_norm": 0.4223475754261017,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 4502
+    },
+    {
+      "epoch": 0.5231484170781295,
+      "grad_norm": 0.42708471417427063,
+      "learning_rate": 0.0001,
+      "loss": 1.8193,
+      "step": 4503
+    },
+    {
+      "epoch": 0.52326459483009,
+      "grad_norm": 0.42466166615486145,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 4504
+    },
+    {
+      "epoch": 0.5233807725820505,
+      "grad_norm": 0.4411070942878723,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 4505
+    },
+    {
+      "epoch": 0.523496950334011,
+      "grad_norm": 0.4360845983028412,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 4506
+    },
+    {
+      "epoch": 0.5236131280859715,
+      "grad_norm": 0.4071287512779236,
+      "learning_rate": 0.0001,
+      "loss": 1.6069,
+      "step": 4507
+    },
+    {
+      "epoch": 0.5237293058379321,
+      "grad_norm": 0.4129199981689453,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 4508
+    },
+    {
+      "epoch": 0.5238454835898926,
+      "grad_norm": 0.40559443831443787,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 4509
+    },
+    {
+      "epoch": 0.5239616613418531,
+      "grad_norm": 0.4351811707019806,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 4510
+    },
+    {
+      "epoch": 0.5240778390938136,
+      "grad_norm": 0.4436653256416321,
+      "learning_rate": 0.0001,
+      "loss": 1.8253,
+      "step": 4511
+    },
+    {
+      "epoch": 0.524194016845774,
+      "grad_norm": 0.4075913727283478,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 4512
+    },
+    {
+      "epoch": 0.5243101945977345,
+      "grad_norm": 0.4155612885951996,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 4513
+    },
+    {
+      "epoch": 0.524426372349695,
+      "grad_norm": 0.4131620526313782,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 4514
+    },
+    {
+      "epoch": 0.5245425501016555,
+      "grad_norm": 0.4026408791542053,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 4515
+    },
+    {
+      "epoch": 0.524658727853616,
+      "grad_norm": 0.42349475622177124,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 4516
+    },
+    {
+      "epoch": 0.5247749056055765,
+      "grad_norm": 0.4265604615211487,
+      "learning_rate": 0.0001,
+      "loss": 1.6916,
+      "step": 4517
+    },
+    {
+      "epoch": 0.524891083357537,
+      "grad_norm": 0.4295273423194885,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 4518
+    },
+    {
+      "epoch": 0.5250072611094976,
+      "grad_norm": 0.42609089612960815,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 4519
+    },
+    {
+      "epoch": 0.5251234388614581,
+      "grad_norm": 0.42496544122695923,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 4520
+    },
+    {
+      "epoch": 0.5252396166134186,
+      "grad_norm": 0.4378863573074341,
+      "learning_rate": 0.0001,
+      "loss": 1.7169,
+      "step": 4521
+    },
+    {
+      "epoch": 0.525355794365379,
+      "grad_norm": 0.43438613414764404,
+      "learning_rate": 0.0001,
+      "loss": 1.7235,
+      "step": 4522
+    },
+    {
+      "epoch": 0.5254719721173395,
+      "grad_norm": 0.4248746931552887,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 4523
+    },
+    {
+      "epoch": 0.5255881498693,
+      "grad_norm": 0.43905171751976013,
+      "learning_rate": 0.0001,
+      "loss": 1.7656,
+      "step": 4524
+    },
+    {
+      "epoch": 0.5257043276212605,
+      "grad_norm": 0.4113710820674896,
+      "learning_rate": 0.0001,
+      "loss": 1.66,
+      "step": 4525
+    },
+    {
+      "epoch": 0.525820505373221,
+      "grad_norm": 0.4203665256500244,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 4526
+    },
+    {
+      "epoch": 0.5259366831251815,
+      "grad_norm": 0.4476137161254883,
+      "learning_rate": 0.0001,
+      "loss": 1.7966,
+      "step": 4527
+    },
+    {
+      "epoch": 0.526052860877142,
+      "grad_norm": 0.400177925825119,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 4528
+    },
+    {
+      "epoch": 0.5261690386291026,
+      "grad_norm": 0.4333209991455078,
+      "learning_rate": 0.0001,
+      "loss": 1.8233,
+      "step": 4529
+    },
+    {
+      "epoch": 0.5262852163810631,
+      "grad_norm": 0.43465256690979004,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 4530
+    },
+    {
+      "epoch": 0.5264013941330236,
+      "grad_norm": 0.4152390658855438,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 4531
+    },
+    {
+      "epoch": 0.526517571884984,
+      "grad_norm": 0.42579424381256104,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 4532
+    },
+    {
+      "epoch": 0.5266337496369445,
+      "grad_norm": 0.4379642903804779,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 4533
+    },
+    {
+      "epoch": 0.526749927388905,
+      "grad_norm": 0.4103189706802368,
+      "learning_rate": 0.0001,
+      "loss": 1.6903,
+      "step": 4534
+    },
+    {
+      "epoch": 0.5268661051408655,
+      "grad_norm": 0.45309609174728394,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 4535
+    },
+    {
+      "epoch": 0.526982282892826,
+      "grad_norm": 0.42796722054481506,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 4536
+    },
+    {
+      "epoch": 0.5270984606447865,
+      "grad_norm": 0.41641005873680115,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 4537
+    },
+    {
+      "epoch": 0.527214638396747,
+      "grad_norm": 0.4048060178756714,
+      "learning_rate": 0.0001,
+      "loss": 1.5817,
+      "step": 4538
+    },
+    {
+      "epoch": 0.5273308161487075,
+      "grad_norm": 0.41168227791786194,
+      "learning_rate": 0.0001,
+      "loss": 1.6224,
+      "step": 4539
+    },
+    {
+      "epoch": 0.5274469939006681,
+      "grad_norm": 0.45252159237861633,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 4540
+    },
+    {
+      "epoch": 0.5275631716526286,
+      "grad_norm": 0.4120829999446869,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 4541
+    },
+    {
+      "epoch": 0.527679349404589,
+      "grad_norm": 0.44721588492393494,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 4542
+    },
+    {
+      "epoch": 0.5277955271565495,
+      "grad_norm": 0.43271827697753906,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 4543
+    },
+    {
+      "epoch": 0.52791170490851,
+      "grad_norm": 0.42740026116371155,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 4544
+    },
+    {
+      "epoch": 0.5280278826604705,
+      "grad_norm": 0.3889728784561157,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 4545
+    },
+    {
+      "epoch": 0.528144060412431,
+      "grad_norm": 0.4437465965747833,
+      "learning_rate": 0.0001,
+      "loss": 1.6956,
+      "step": 4546
+    },
+    {
+      "epoch": 0.5282602381643915,
+      "grad_norm": 0.39472874999046326,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 4547
+    },
+    {
+      "epoch": 0.528376415916352,
+      "grad_norm": 0.4104737937450409,
+      "learning_rate": 0.0001,
+      "loss": 1.8144,
+      "step": 4548
+    },
+    {
+      "epoch": 0.5284925936683125,
+      "grad_norm": 0.4385277330875397,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 4549
+    },
+    {
+      "epoch": 0.5286087714202731,
+      "grad_norm": 0.3903822600841522,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 4550
+    },
+    {
+      "epoch": 0.5287249491722336,
+      "grad_norm": 0.4265904128551483,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 4551
+    },
+    {
+      "epoch": 0.528841126924194,
+      "grad_norm": 0.4125080704689026,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 4552
+    },
+    {
+      "epoch": 0.5289573046761545,
+      "grad_norm": 0.42072439193725586,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 4553
+    },
+    {
+      "epoch": 0.529073482428115,
+      "grad_norm": 0.41168302297592163,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 4554
+    },
+    {
+      "epoch": 0.5291896601800755,
+      "grad_norm": 0.39374563097953796,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 4555
+    },
+    {
+      "epoch": 0.529305837932036,
+      "grad_norm": 0.43736445903778076,
+      "learning_rate": 0.0001,
+      "loss": 1.6175,
+      "step": 4556
+    },
+    {
+      "epoch": 0.5294220156839965,
+      "grad_norm": 0.456046462059021,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 4557
+    },
+    {
+      "epoch": 0.529538193435957,
+      "grad_norm": 0.407843679189682,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 4558
+    },
+    {
+      "epoch": 0.5296543711879175,
+      "grad_norm": 0.4293687045574188,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 4559
+    },
+    {
+      "epoch": 0.529770548939878,
+      "grad_norm": 0.42823317646980286,
+      "learning_rate": 0.0001,
+      "loss": 1.8056,
+      "step": 4560
+    },
+    {
+      "epoch": 0.5298867266918386,
+      "grad_norm": 0.41989070177078247,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 4561
+    },
+    {
+      "epoch": 0.530002904443799,
+      "grad_norm": 0.41696497797966003,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 4562
+    },
+    {
+      "epoch": 0.5301190821957595,
+      "grad_norm": 0.3901718854904175,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 4563
+    },
+    {
+      "epoch": 0.53023525994772,
+      "grad_norm": 0.42461147904396057,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 4564
+    },
+    {
+      "epoch": 0.5303514376996805,
+      "grad_norm": 0.4265443682670593,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 4565
+    },
+    {
+      "epoch": 0.530467615451641,
+      "grad_norm": 0.4418391287326813,
+      "learning_rate": 0.0001,
+      "loss": 1.7592,
+      "step": 4566
+    },
+    {
+      "epoch": 0.5305837932036015,
+      "grad_norm": 0.4429027736186981,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 4567
+    },
+    {
+      "epoch": 0.530699970955562,
+      "grad_norm": 0.4483877420425415,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 4568
+    },
+    {
+      "epoch": 0.5308161487075225,
+      "grad_norm": 0.4467184245586395,
+      "learning_rate": 0.0001,
+      "loss": 1.7746,
+      "step": 4569
+    },
+    {
+      "epoch": 0.530932326459483,
+      "grad_norm": 0.4503554403781891,
+      "learning_rate": 0.0001,
+      "loss": 1.8263,
+      "step": 4570
+    },
+    {
+      "epoch": 0.5310485042114436,
+      "grad_norm": 0.40501412749290466,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 4571
+    },
+    {
+      "epoch": 0.531164681963404,
+      "grad_norm": 0.41153189539909363,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 4572
+    },
+    {
+      "epoch": 0.5312808597153645,
+      "grad_norm": 0.41947031021118164,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 4573
+    },
+    {
+      "epoch": 0.531397037467325,
+      "grad_norm": 0.4129175841808319,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 4574
+    },
+    {
+      "epoch": 0.5315132152192855,
+      "grad_norm": 0.44129297137260437,
+      "learning_rate": 0.0001,
+      "loss": 1.693,
+      "step": 4575
+    },
+    {
+      "epoch": 0.531629392971246,
+      "grad_norm": 0.41922008991241455,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 4576
+    },
+    {
+      "epoch": 0.5317455707232065,
+      "grad_norm": 0.41607466340065,
+      "learning_rate": 0.0001,
+      "loss": 1.6542,
+      "step": 4577
+    },
+    {
+      "epoch": 0.531861748475167,
+      "grad_norm": 0.42458733916282654,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 4578
+    },
+    {
+      "epoch": 0.5319779262271275,
+      "grad_norm": 0.43366140127182007,
+      "learning_rate": 0.0001,
+      "loss": 1.508,
+      "step": 4579
+    },
+    {
+      "epoch": 0.532094103979088,
+      "grad_norm": 0.4278888702392578,
+      "learning_rate": 0.0001,
+      "loss": 1.688,
+      "step": 4580
+    },
+    {
+      "epoch": 0.5322102817310485,
+      "grad_norm": 0.4110510051250458,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 4581
+    },
+    {
+      "epoch": 0.532326459483009,
+      "grad_norm": 0.4136632978916168,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 4582
+    },
+    {
+      "epoch": 0.5324426372349695,
+      "grad_norm": 0.3931570053100586,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 4583
+    },
+    {
+      "epoch": 0.53255881498693,
+      "grad_norm": 0.41531267762184143,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 4584
+    },
+    {
+      "epoch": 0.5326749927388905,
+      "grad_norm": 0.4423351585865021,
+      "learning_rate": 0.0001,
+      "loss": 1.847,
+      "step": 4585
+    },
+    {
+      "epoch": 0.532791170490851,
+      "grad_norm": 0.401030570268631,
+      "learning_rate": 0.0001,
+      "loss": 1.592,
+      "step": 4586
+    },
+    {
+      "epoch": 0.5329073482428115,
+      "grad_norm": 0.42098504304885864,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 4587
+    },
+    {
+      "epoch": 0.533023525994772,
+      "grad_norm": 0.4085943102836609,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 4588
+    },
+    {
+      "epoch": 0.5331397037467325,
+      "grad_norm": 0.39442306756973267,
+      "learning_rate": 0.0001,
+      "loss": 1.4053,
+      "step": 4589
+    },
+    {
+      "epoch": 0.533255881498693,
+      "grad_norm": 0.4020472466945648,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 4590
+    },
+    {
+      "epoch": 0.5333720592506535,
+      "grad_norm": 0.41288527846336365,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 4591
+    },
+    {
+      "epoch": 0.533488237002614,
+      "grad_norm": 0.4358307719230652,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 4592
+    },
+    {
+      "epoch": 0.5336044147545745,
+      "grad_norm": 0.4035166800022125,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 4593
+    },
+    {
+      "epoch": 0.533720592506535,
+      "grad_norm": 0.420097291469574,
+      "learning_rate": 0.0001,
+      "loss": 1.7894,
+      "step": 4594
+    },
+    {
+      "epoch": 0.5338367702584955,
+      "grad_norm": 0.47331398725509644,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 4595
+    },
+    {
+      "epoch": 0.533952948010456,
+      "grad_norm": 0.42985647916793823,
+      "learning_rate": 0.0001,
+      "loss": 1.7268,
+      "step": 4596
+    },
+    {
+      "epoch": 0.5340691257624165,
+      "grad_norm": 0.3984420895576477,
+      "learning_rate": 0.0001,
+      "loss": 1.4863,
+      "step": 4597
+    },
+    {
+      "epoch": 0.534185303514377,
+      "grad_norm": 0.41712093353271484,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 4598
+    },
+    {
+      "epoch": 0.5343014812663375,
+      "grad_norm": 0.42095884680747986,
+      "learning_rate": 0.0001,
+      "loss": 1.7377,
+      "step": 4599
+    },
+    {
+      "epoch": 0.534417659018298,
+      "grad_norm": 0.4140506386756897,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 4600
+    },
+    {
+      "epoch": 0.5345338367702585,
+      "grad_norm": 0.39841312170028687,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 4601
+    },
+    {
+      "epoch": 0.5346500145222189,
+      "grad_norm": 0.46106958389282227,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 4602
+    },
+    {
+      "epoch": 0.5347661922741795,
+      "grad_norm": 0.4085831344127655,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 4603
+    },
+    {
+      "epoch": 0.53488237002614,
+      "grad_norm": 0.49497655034065247,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 4604
+    },
+    {
+      "epoch": 0.5349985477781005,
+      "grad_norm": 0.43607422709465027,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 4605
+    },
+    {
+      "epoch": 0.535114725530061,
+      "grad_norm": 0.4607904255390167,
+      "learning_rate": 0.0001,
+      "loss": 1.6288,
+      "step": 4606
+    },
+    {
+      "epoch": 0.5352309032820215,
+      "grad_norm": 0.4455506205558777,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 4607
+    },
+    {
+      "epoch": 0.535347081033982,
+      "grad_norm": 0.39066752791404724,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 4608
+    },
+    {
+      "epoch": 0.5354632587859425,
+      "grad_norm": 0.4268709421157837,
+      "learning_rate": 0.0001,
+      "loss": 1.8102,
+      "step": 4609
+    },
+    {
+      "epoch": 0.535579436537903,
+      "grad_norm": 0.40803661942481995,
+      "learning_rate": 0.0001,
+      "loss": 1.6831,
+      "step": 4610
+    },
+    {
+      "epoch": 0.5356956142898635,
+      "grad_norm": 0.4266864061355591,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 4611
+    },
+    {
+      "epoch": 0.5358117920418239,
+      "grad_norm": 0.4282558262348175,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 4612
+    },
+    {
+      "epoch": 0.5359279697937845,
+      "grad_norm": 0.4355599880218506,
+      "learning_rate": 0.0001,
+      "loss": 1.7922,
+      "step": 4613
+    },
+    {
+      "epoch": 0.536044147545745,
+      "grad_norm": 0.41130515933036804,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 4614
+    },
+    {
+      "epoch": 0.5361603252977055,
+      "grad_norm": 0.4203796982765198,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 4615
+    },
+    {
+      "epoch": 0.536276503049666,
+      "grad_norm": 0.4113680124282837,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 4616
+    },
+    {
+      "epoch": 0.5363926808016265,
+      "grad_norm": 0.4167897701263428,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 4617
+    },
+    {
+      "epoch": 0.536508858553587,
+      "grad_norm": 0.39767855405807495,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 4618
+    },
+    {
+      "epoch": 0.5366250363055475,
+      "grad_norm": 0.42861250042915344,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 4619
+    },
+    {
+      "epoch": 0.536741214057508,
+      "grad_norm": 0.4207315742969513,
+      "learning_rate": 0.0001,
+      "loss": 1.8116,
+      "step": 4620
+    },
+    {
+      "epoch": 0.5368573918094685,
+      "grad_norm": 0.41089701652526855,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 4621
+    },
+    {
+      "epoch": 0.5369735695614289,
+      "grad_norm": 0.42728808522224426,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 4622
+    },
+    {
+      "epoch": 0.5370897473133894,
+      "grad_norm": 0.4472593069076538,
+      "learning_rate": 0.0001,
+      "loss": 1.7698,
+      "step": 4623
+    },
+    {
+      "epoch": 0.53720592506535,
+      "grad_norm": 0.41069328784942627,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 4624
+    },
+    {
+      "epoch": 0.5373221028173105,
+      "grad_norm": 0.4025200605392456,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 4625
+    },
+    {
+      "epoch": 0.537438280569271,
+      "grad_norm": 0.4271898567676544,
+      "learning_rate": 0.0001,
+      "loss": 1.8612,
+      "step": 4626
+    },
+    {
+      "epoch": 0.5375544583212315,
+      "grad_norm": 0.40859630703926086,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 4627
+    },
+    {
+      "epoch": 0.537670636073192,
+      "grad_norm": 0.390266090631485,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 4628
+    },
+    {
+      "epoch": 0.5377868138251525,
+      "grad_norm": 0.44839224219322205,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 4629
+    },
+    {
+      "epoch": 0.537902991577113,
+      "grad_norm": 0.4331499934196472,
+      "learning_rate": 0.0001,
+      "loss": 1.748,
+      "step": 4630
+    },
+    {
+      "epoch": 0.5380191693290735,
+      "grad_norm": 0.41848304867744446,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 4631
+    },
+    {
+      "epoch": 0.5381353470810339,
+      "grad_norm": 0.4118647873401642,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 4632
+    },
+    {
+      "epoch": 0.5382515248329944,
+      "grad_norm": 0.4422479569911957,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 4633
+    },
+    {
+      "epoch": 0.538367702584955,
+      "grad_norm": 0.4162086844444275,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 4634
+    },
+    {
+      "epoch": 0.5384838803369155,
+      "grad_norm": 0.4036068618297577,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 4635
+    },
+    {
+      "epoch": 0.538600058088876,
+      "grad_norm": 0.4048601984977722,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 4636
+    },
+    {
+      "epoch": 0.5387162358408365,
+      "grad_norm": 0.4456162750720978,
+      "learning_rate": 0.0001,
+      "loss": 1.7916,
+      "step": 4637
+    },
+    {
+      "epoch": 0.538832413592797,
+      "grad_norm": 0.44490668177604675,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 4638
+    },
+    {
+      "epoch": 0.5389485913447575,
+      "grad_norm": 0.42898234724998474,
+      "learning_rate": 0.0001,
+      "loss": 1.6908,
+      "step": 4639
+    },
+    {
+      "epoch": 0.539064769096718,
+      "grad_norm": 0.41526684165000916,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 4640
+    },
+    {
+      "epoch": 0.5391809468486785,
+      "grad_norm": 0.391984224319458,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 4641
+    },
+    {
+      "epoch": 0.5392971246006389,
+      "grad_norm": 0.4175387918949127,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 4642
+    },
+    {
+      "epoch": 0.5394133023525994,
+      "grad_norm": 0.4540769159793854,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 4643
+    },
+    {
+      "epoch": 0.5395294801045599,
+      "grad_norm": 0.43631550669670105,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 4644
+    },
+    {
+      "epoch": 0.5396456578565205,
+      "grad_norm": 0.459330290555954,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 4645
+    },
+    {
+      "epoch": 0.539761835608481,
+      "grad_norm": 0.41860130429267883,
+      "learning_rate": 0.0001,
+      "loss": 1.6275,
+      "step": 4646
+    },
+    {
+      "epoch": 0.5398780133604415,
+      "grad_norm": 0.5068102478981018,
+      "learning_rate": 0.0001,
+      "loss": 1.8484,
+      "step": 4647
+    },
+    {
+      "epoch": 0.539994191112402,
+      "grad_norm": 0.4381267726421356,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5401103688643625,
+      "grad_norm": 0.426881343126297,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 4649
+    },
+    {
+      "epoch": 0.540226546616323,
+      "grad_norm": 0.44243374466896057,
+      "learning_rate": 0.0001,
+      "loss": 1.8272,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5403427243682835,
+      "grad_norm": 0.4320388734340668,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 4651
+    },
+    {
+      "epoch": 0.540458902120244,
+      "grad_norm": 0.418743759393692,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 4652
+    },
+    {
+      "epoch": 0.5405750798722044,
+      "grad_norm": 0.4069058299064636,
+      "learning_rate": 0.0001,
+      "loss": 1.7696,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5406912576241649,
+      "grad_norm": 0.4170093536376953,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 4654
+    },
+    {
+      "epoch": 0.5408074353761255,
+      "grad_norm": 0.4032455384731293,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 4655
+    },
+    {
+      "epoch": 0.540923613128086,
+      "grad_norm": 0.3982067108154297,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 4656
+    },
+    {
+      "epoch": 0.5410397908800465,
+      "grad_norm": 0.4389408528804779,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 4657
+    },
+    {
+      "epoch": 0.541155968632007,
+      "grad_norm": 0.41216394305229187,
+      "learning_rate": 0.0001,
+      "loss": 1.724,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5412721463839675,
+      "grad_norm": 0.4423290491104126,
+      "learning_rate": 0.0001,
+      "loss": 1.8955,
+      "step": 4659
+    },
+    {
+      "epoch": 0.541388324135928,
+      "grad_norm": 0.45288944244384766,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5415045018878885,
+      "grad_norm": 0.43445900082588196,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 4661
+    },
+    {
+      "epoch": 0.541620679639849,
+      "grad_norm": 0.4164069890975952,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5417368573918094,
+      "grad_norm": 0.4147018492221832,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5418530351437699,
+      "grad_norm": 0.4525691568851471,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 4664
+    },
+    {
+      "epoch": 0.5419692128957304,
+      "grad_norm": 0.4286227822303772,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 4665
+    },
+    {
+      "epoch": 0.542085390647691,
+      "grad_norm": 0.4411110281944275,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5422015683996515,
+      "grad_norm": 0.3831879794597626,
+      "learning_rate": 0.0001,
+      "loss": 1.3289,
+      "step": 4667
+    },
+    {
+      "epoch": 0.542317746151612,
+      "grad_norm": 0.42957553267478943,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5424339239035725,
+      "grad_norm": 0.43326959013938904,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 4669
+    },
+    {
+      "epoch": 0.542550101655533,
+      "grad_norm": 0.42409175634384155,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5426662794074935,
+      "grad_norm": 0.42008155584335327,
+      "learning_rate": 0.0001,
+      "loss": 1.6757,
+      "step": 4671
+    },
+    {
+      "epoch": 0.542782457159454,
+      "grad_norm": 0.41546356678009033,
+      "learning_rate": 0.0001,
+      "loss": 1.6692,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5428986349114144,
+      "grad_norm": 0.41974398493766785,
+      "learning_rate": 0.0001,
+      "loss": 1.6507,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5430148126633749,
+      "grad_norm": 0.39772695302963257,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 4674
+    },
+    {
+      "epoch": 0.5431309904153354,
+      "grad_norm": 0.43353089690208435,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 4675
+    },
+    {
+      "epoch": 0.543247168167296,
+      "grad_norm": 0.4182477295398712,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5433633459192565,
+      "grad_norm": 0.4173682928085327,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 4677
+    },
+    {
+      "epoch": 0.543479523671217,
+      "grad_norm": 0.44657060503959656,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5435957014231775,
+      "grad_norm": 0.43546125292778015,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 4679
+    },
+    {
+      "epoch": 0.543711879175138,
+      "grad_norm": 0.45045846700668335,
+      "learning_rate": 0.0001,
+      "loss": 1.6601,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5438280569270985,
+      "grad_norm": 0.4383883774280548,
+      "learning_rate": 0.0001,
+      "loss": 1.5701,
+      "step": 4681
+    },
+    {
+      "epoch": 0.543944234679059,
+      "grad_norm": 0.40852198004722595,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 4682
+    },
+    {
+      "epoch": 0.5440604124310194,
+      "grad_norm": 0.4357270896434784,
+      "learning_rate": 0.0001,
+      "loss": 1.8067,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5441765901829799,
+      "grad_norm": 0.4512038826942444,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 4684
+    },
+    {
+      "epoch": 0.5442927679349404,
+      "grad_norm": 0.41331276297569275,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 4685
+    },
+    {
+      "epoch": 0.544408945686901,
+      "grad_norm": 0.4491327702999115,
+      "learning_rate": 0.0001,
+      "loss": 1.7954,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5445251234388615,
+      "grad_norm": 0.4428486227989197,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 4687
+    },
+    {
+      "epoch": 0.544641301190822,
+      "grad_norm": 0.416236937046051,
+      "learning_rate": 0.0001,
+      "loss": 1.6693,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5447574789427825,
+      "grad_norm": 0.40862324833869934,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 4689
+    },
+    {
+      "epoch": 0.544873656694743,
+      "grad_norm": 0.43137800693511963,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5449898344467035,
+      "grad_norm": 0.4132535755634308,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 4691
+    },
+    {
+      "epoch": 0.545106012198664,
+      "grad_norm": 0.5874444842338562,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 4692
+    },
+    {
+      "epoch": 0.5452221899506244,
+      "grad_norm": 0.379760205745697,
+      "learning_rate": 0.0001,
+      "loss": 1.3516,
+      "step": 4693
+    },
+    {
+      "epoch": 0.5453383677025849,
+      "grad_norm": 0.4385640025138855,
+      "learning_rate": 0.0001,
+      "loss": 1.776,
+      "step": 4694
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.4262996315956116,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 4695
+    },
+    {
+      "epoch": 0.5455707232065059,
+      "grad_norm": 0.4239009916782379,
+      "learning_rate": 0.0001,
+      "loss": 1.6814,
+      "step": 4696
+    },
+    {
+      "epoch": 0.5456869009584665,
+      "grad_norm": 0.40220028162002563,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 4697
+    },
+    {
+      "epoch": 0.545803078710427,
+      "grad_norm": 0.3948020339012146,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 4698
+    },
+    {
+      "epoch": 0.5459192564623875,
+      "grad_norm": 0.3882817029953003,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 4699
+    },
+    {
+      "epoch": 0.546035434214348,
+      "grad_norm": 0.4421759843826294,
+      "learning_rate": 0.0001,
+      "loss": 1.6698,
+      "step": 4700
+    },
+    {
+      "epoch": 0.5461516119663085,
+      "grad_norm": 0.4043938219547272,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 4701
+    },
+    {
+      "epoch": 0.546267789718269,
+      "grad_norm": 0.42511290311813354,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 4702
+    },
+    {
+      "epoch": 0.5463839674702294,
+      "grad_norm": 0.4274735748767853,
+      "learning_rate": 0.0001,
+      "loss": 1.6945,
+      "step": 4703
+    },
+    {
+      "epoch": 0.5465001452221899,
+      "grad_norm": 0.4572598934173584,
+      "learning_rate": 0.0001,
+      "loss": 1.754,
+      "step": 4704
+    },
+    {
+      "epoch": 0.5466163229741504,
+      "grad_norm": 0.41152650117874146,
+      "learning_rate": 0.0001,
+      "loss": 1.3119,
+      "step": 4705
+    },
+    {
+      "epoch": 0.5467325007261109,
+      "grad_norm": 0.4448837637901306,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 4706
+    },
+    {
+      "epoch": 0.5468486784780715,
+      "grad_norm": 0.4235488176345825,
+      "learning_rate": 0.0001,
+      "loss": 1.7473,
+      "step": 4707
+    },
+    {
+      "epoch": 0.546964856230032,
+      "grad_norm": 0.4105146527290344,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 4708
+    },
+    {
+      "epoch": 0.5470810339819925,
+      "grad_norm": 0.4490967094898224,
+      "learning_rate": 0.0001,
+      "loss": 1.7984,
+      "step": 4709
+    },
+    {
+      "epoch": 0.547197211733953,
+      "grad_norm": 0.4733203947544098,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 4710
+    },
+    {
+      "epoch": 0.5473133894859135,
+      "grad_norm": 0.4103139340877533,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 4711
+    },
+    {
+      "epoch": 0.547429567237874,
+      "grad_norm": 0.41262155771255493,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 4712
+    },
+    {
+      "epoch": 0.5475457449898344,
+      "grad_norm": 0.41517654061317444,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 4713
+    },
+    {
+      "epoch": 0.5476619227417949,
+      "grad_norm": 0.43527457118034363,
+      "learning_rate": 0.0001,
+      "loss": 1.7127,
+      "step": 4714
+    },
+    {
+      "epoch": 0.5477781004937554,
+      "grad_norm": 0.3753294348716736,
+      "learning_rate": 0.0001,
+      "loss": 1.4135,
+      "step": 4715
+    },
+    {
+      "epoch": 0.5478942782457159,
+      "grad_norm": 0.4435529112815857,
+      "learning_rate": 0.0001,
+      "loss": 1.7554,
+      "step": 4716
+    },
+    {
+      "epoch": 0.5480104559976764,
+      "grad_norm": 0.39679768681526184,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 4717
+    },
+    {
+      "epoch": 0.548126633749637,
+      "grad_norm": 0.421624094247818,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 4718
+    },
+    {
+      "epoch": 0.5482428115015975,
+      "grad_norm": 0.42037492990493774,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 4719
+    },
+    {
+      "epoch": 0.548358989253558,
+      "grad_norm": 0.44528070092201233,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 4720
+    },
+    {
+      "epoch": 0.5484751670055185,
+      "grad_norm": 0.41840189695358276,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 4721
+    },
+    {
+      "epoch": 0.548591344757479,
+      "grad_norm": 0.3980635404586792,
+      "learning_rate": 0.0001,
+      "loss": 1.7035,
+      "step": 4722
+    },
+    {
+      "epoch": 0.5487075225094394,
+      "grad_norm": 0.42704808712005615,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 4723
+    },
+    {
+      "epoch": 0.5488237002613999,
+      "grad_norm": 0.4258357584476471,
+      "learning_rate": 0.0001,
+      "loss": 1.6625,
+      "step": 4724
+    },
+    {
+      "epoch": 0.5489398780133604,
+      "grad_norm": 0.4512523114681244,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 4725
+    },
+    {
+      "epoch": 0.5490560557653209,
+      "grad_norm": 0.4495634138584137,
+      "learning_rate": 0.0001,
+      "loss": 1.7266,
+      "step": 4726
+    },
+    {
+      "epoch": 0.5491722335172814,
+      "grad_norm": 0.44394296407699585,
+      "learning_rate": 0.0001,
+      "loss": 1.7735,
+      "step": 4727
+    },
+    {
+      "epoch": 0.549288411269242,
+      "grad_norm": 0.4451744258403778,
+      "learning_rate": 0.0001,
+      "loss": 1.8485,
+      "step": 4728
+    },
+    {
+      "epoch": 0.5494045890212025,
+      "grad_norm": 0.42420530319213867,
+      "learning_rate": 0.0001,
+      "loss": 1.7562,
+      "step": 4729
+    },
+    {
+      "epoch": 0.549520766773163,
+      "grad_norm": 0.4318549931049347,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 4730
+    },
+    {
+      "epoch": 0.5496369445251235,
+      "grad_norm": 0.3942951560020447,
+      "learning_rate": 0.0001,
+      "loss": 1.7032,
+      "step": 4731
+    },
+    {
+      "epoch": 0.549753122277084,
+      "grad_norm": 0.42226502299308777,
+      "learning_rate": 0.0001,
+      "loss": 1.723,
+      "step": 4732
+    },
+    {
+      "epoch": 0.5498693000290444,
+      "grad_norm": 0.45085522532463074,
+      "learning_rate": 0.0001,
+      "loss": 1.7787,
+      "step": 4733
+    },
+    {
+      "epoch": 0.5499854777810049,
+      "grad_norm": 0.40263989567756653,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 4734
+    },
+    {
+      "epoch": 0.5501016555329654,
+      "grad_norm": 0.4361323416233063,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 4735
+    },
+    {
+      "epoch": 0.5502178332849259,
+      "grad_norm": 0.4477275311946869,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 4736
+    },
+    {
+      "epoch": 0.5503340110368864,
+      "grad_norm": 0.490419864654541,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 4737
+    },
+    {
+      "epoch": 0.5504501887888469,
+      "grad_norm": 0.4405496418476105,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 4738
+    },
+    {
+      "epoch": 0.5505663665408075,
+      "grad_norm": 0.38646984100341797,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 4739
+    },
+    {
+      "epoch": 0.550682544292768,
+      "grad_norm": 0.40412217378616333,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 4740
+    },
+    {
+      "epoch": 0.5507987220447285,
+      "grad_norm": 0.41205301880836487,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 4741
+    },
+    {
+      "epoch": 0.550914899796689,
+      "grad_norm": 0.4445815086364746,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 4742
+    },
+    {
+      "epoch": 0.5510310775486494,
+      "grad_norm": 0.3821341395378113,
+      "learning_rate": 0.0001,
+      "loss": 1.364,
+      "step": 4743
+    },
+    {
+      "epoch": 0.5511472553006099,
+      "grad_norm": 0.4401394724845886,
+      "learning_rate": 0.0001,
+      "loss": 1.8219,
+      "step": 4744
+    },
+    {
+      "epoch": 0.5512634330525704,
+      "grad_norm": 0.4259290397167206,
+      "learning_rate": 0.0001,
+      "loss": 1.7113,
+      "step": 4745
+    },
+    {
+      "epoch": 0.5513796108045309,
+      "grad_norm": 0.4446122348308563,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 4746
+    },
+    {
+      "epoch": 0.5514957885564914,
+      "grad_norm": 0.4578036367893219,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 4747
+    },
+    {
+      "epoch": 0.5516119663084519,
+      "grad_norm": 0.42073720693588257,
+      "learning_rate": 0.0001,
+      "loss": 1.692,
+      "step": 4748
+    },
+    {
+      "epoch": 0.5517281440604125,
+      "grad_norm": 0.4655468761920929,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 4749
+    },
+    {
+      "epoch": 0.551844321812373,
+      "grad_norm": 0.41886764764785767,
+      "learning_rate": 0.0001,
+      "loss": 1.6146,
+      "step": 4750
+    },
+    {
+      "epoch": 0.5519604995643335,
+      "grad_norm": 0.42605897784233093,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 4751
+    },
+    {
+      "epoch": 0.552076677316294,
+      "grad_norm": 0.43297725915908813,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 4752
+    },
+    {
+      "epoch": 0.5521928550682544,
+      "grad_norm": 0.44777053594589233,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 4753
+    },
+    {
+      "epoch": 0.5523090328202149,
+      "grad_norm": 0.43533116579055786,
+      "learning_rate": 0.0001,
+      "loss": 1.6652,
+      "step": 4754
+    },
+    {
+      "epoch": 0.5524252105721754,
+      "grad_norm": 0.4219701290130615,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5525413883241359,
+      "grad_norm": 0.42743462324142456,
+      "learning_rate": 0.0001,
+      "loss": 1.7591,
+      "step": 4756
+    },
+    {
+      "epoch": 0.5526575660760964,
+      "grad_norm": 0.4393671452999115,
+      "learning_rate": 0.0001,
+      "loss": 1.7385,
+      "step": 4757
+    },
+    {
+      "epoch": 0.5527737438280569,
+      "grad_norm": 0.43937399983406067,
+      "learning_rate": 0.0001,
+      "loss": 1.6714,
+      "step": 4758
+    },
+    {
+      "epoch": 0.5528899215800174,
+      "grad_norm": 0.41498544812202454,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 4759
+    },
+    {
+      "epoch": 0.553006099331978,
+      "grad_norm": 0.3907436430454254,
+      "learning_rate": 0.0001,
+      "loss": 1.4194,
+      "step": 4760
+    },
+    {
+      "epoch": 0.5531222770839385,
+      "grad_norm": 0.4304429292678833,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 4761
+    },
+    {
+      "epoch": 0.553238454835899,
+      "grad_norm": 0.41753122210502625,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5533546325878594,
+      "grad_norm": 0.4009332060813904,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5534708103398199,
+      "grad_norm": 0.41789358854293823,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5535869880917804,
+      "grad_norm": 0.42972517013549805,
+      "learning_rate": 0.0001,
+      "loss": 1.7597,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5537031658437409,
+      "grad_norm": 0.4440890848636627,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5538193435957014,
+      "grad_norm": 0.43731603026390076,
+      "learning_rate": 0.0001,
+      "loss": 1.7988,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5539355213476619,
+      "grad_norm": 0.4374067485332489,
+      "learning_rate": 0.0001,
+      "loss": 1.6098,
+      "step": 4768
+    },
+    {
+      "epoch": 0.5540516990996224,
+      "grad_norm": 0.43468907475471497,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 4769
+    },
+    {
+      "epoch": 0.554167876851583,
+      "grad_norm": 0.44447562098503113,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5542840546035435,
+      "grad_norm": 0.4318912625312805,
+      "learning_rate": 0.0001,
+      "loss": 1.6811,
+      "step": 4771
+    },
+    {
+      "epoch": 0.554400232355504,
+      "grad_norm": 0.4088803827762604,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 4772
+    },
+    {
+      "epoch": 0.5545164101074644,
+      "grad_norm": 0.45096081495285034,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 4773
+    },
+    {
+      "epoch": 0.5546325878594249,
+      "grad_norm": 0.45685335993766785,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5547487656113854,
+      "grad_norm": 0.4274502396583557,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5548649433633459,
+      "grad_norm": 0.3890495300292969,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 4776
+    },
+    {
+      "epoch": 0.5549811211153064,
+      "grad_norm": 0.4626457095146179,
+      "learning_rate": 0.0001,
+      "loss": 1.5708,
+      "step": 4777
+    },
+    {
+      "epoch": 0.5550972988672669,
+      "grad_norm": 0.42439907789230347,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5552134766192274,
+      "grad_norm": 0.45186173915863037,
+      "learning_rate": 0.0001,
+      "loss": 1.7971,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5553296543711879,
+      "grad_norm": 0.4208846688270569,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5554458321231485,
+      "grad_norm": 0.43783605098724365,
+      "learning_rate": 0.0001,
+      "loss": 1.8354,
+      "step": 4781
+    },
+    {
+      "epoch": 0.555562009875109,
+      "grad_norm": 0.4037763178348541,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5556781876270694,
+      "grad_norm": 0.41722458600997925,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5557943653790299,
+      "grad_norm": 0.4464002549648285,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 4784
+    },
+    {
+      "epoch": 0.5559105431309904,
+      "grad_norm": 0.4327322840690613,
+      "learning_rate": 0.0001,
+      "loss": 1.829,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5560267208829509,
+      "grad_norm": 0.4065304100513458,
+      "learning_rate": 0.0001,
+      "loss": 1.6915,
+      "step": 4786
+    },
+    {
+      "epoch": 0.5561428986349114,
+      "grad_norm": 0.4391983449459076,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5562590763868719,
+      "grad_norm": 0.42863357067108154,
+      "learning_rate": 0.0001,
+      "loss": 1.4858,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5563752541388324,
+      "grad_norm": 0.40624338388442993,
+      "learning_rate": 0.0001,
+      "loss": 1.4187,
+      "step": 4789
+    },
+    {
+      "epoch": 0.5564914318907929,
+      "grad_norm": 0.4419528543949127,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5566076096427535,
+      "grad_norm": 0.470945805311203,
+      "learning_rate": 0.0001,
+      "loss": 1.8979,
+      "step": 4791
+    },
+    {
+      "epoch": 0.556723787394714,
+      "grad_norm": 0.4236770272254944,
+      "learning_rate": 0.0001,
+      "loss": 1.8773,
+      "step": 4792
+    },
+    {
+      "epoch": 0.5568399651466744,
+      "grad_norm": 0.426725834608078,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 4793
+    },
+    {
+      "epoch": 0.5569561428986349,
+      "grad_norm": 0.4224294126033783,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5570723206505954,
+      "grad_norm": 0.413329541683197,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5571884984025559,
+      "grad_norm": 0.43398913741111755,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5573046761545164,
+      "grad_norm": 0.4381754696369171,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5574208539064769,
+      "grad_norm": 0.44960010051727295,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 4798
+    },
+    {
+      "epoch": 0.5575370316584374,
+      "grad_norm": 0.4516027271747589,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 4799
+    },
+    {
+      "epoch": 0.5576532094103979,
+      "grad_norm": 0.4343845844268799,
+      "learning_rate": 0.0001,
+      "loss": 1.7357,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5577693871623584,
+      "grad_norm": 0.436753511428833,
+      "learning_rate": 0.0001,
+      "loss": 1.6702,
+      "step": 4801
+    },
+    {
+      "epoch": 0.557885564914319,
+      "grad_norm": 0.4462500512599945,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 4802
+    },
+    {
+      "epoch": 0.5580017426662794,
+      "grad_norm": 0.396566778421402,
+      "learning_rate": 0.0001,
+      "loss": 1.3311,
+      "step": 4803
+    },
+    {
+      "epoch": 0.5581179204182399,
+      "grad_norm": 0.5002578496932983,
+      "learning_rate": 0.0001,
+      "loss": 1.8922,
+      "step": 4804
+    },
+    {
+      "epoch": 0.5582340981702004,
+      "grad_norm": 0.43090498447418213,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5583502759221609,
+      "grad_norm": 0.4082920253276825,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 4806
+    },
+    {
+      "epoch": 0.5584664536741214,
+      "grad_norm": 0.43145760893821716,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 4807
+    },
+    {
+      "epoch": 0.5585826314260819,
+      "grad_norm": 0.3939366638660431,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 4808
+    },
+    {
+      "epoch": 0.5586988091780424,
+      "grad_norm": 0.45775458216667175,
+      "learning_rate": 0.0001,
+      "loss": 1.7803,
+      "step": 4809
+    },
+    {
+      "epoch": 0.5588149869300029,
+      "grad_norm": 0.4209611117839813,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5589311646819634,
+      "grad_norm": 0.4282268285751343,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 4811
+    },
+    {
+      "epoch": 0.559047342433924,
+      "grad_norm": 0.4249371290206909,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 4812
+    },
+    {
+      "epoch": 0.5591635201858844,
+      "grad_norm": 0.4493142068386078,
+      "learning_rate": 0.0001,
+      "loss": 1.5725,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5592796979378449,
+      "grad_norm": 0.43762892484664917,
+      "learning_rate": 0.0001,
+      "loss": 1.79,
+      "step": 4814
+    },
+    {
+      "epoch": 0.5593958756898054,
+      "grad_norm": 0.4867606461048126,
+      "learning_rate": 0.0001,
+      "loss": 1.786,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5595120534417659,
+      "grad_norm": 0.4017328917980194,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5596282311937264,
+      "grad_norm": 0.4285069704055786,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 4817
+    },
+    {
+      "epoch": 0.5597444089456869,
+      "grad_norm": 0.4313235580921173,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 4818
+    },
+    {
+      "epoch": 0.5598605866976474,
+      "grad_norm": 0.3793412744998932,
+      "learning_rate": 0.0001,
+      "loss": 1.404,
+      "step": 4819
+    },
+    {
+      "epoch": 0.5599767644496079,
+      "grad_norm": 0.4507150948047638,
+      "learning_rate": 0.0001,
+      "loss": 1.8942,
+      "step": 4820
+    },
+    {
+      "epoch": 0.5600929422015684,
+      "grad_norm": 0.41890159249305725,
+      "learning_rate": 0.0001,
+      "loss": 1.5985,
+      "step": 4821
+    },
+    {
+      "epoch": 0.5602091199535288,
+      "grad_norm": 0.4463735818862915,
+      "learning_rate": 0.0001,
+      "loss": 1.7219,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5603252977054894,
+      "grad_norm": 0.42564481496810913,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5604414754574499,
+      "grad_norm": 0.4581405222415924,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5605576532094104,
+      "grad_norm": 0.39707350730895996,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5606738309613709,
+      "grad_norm": 0.43881773948669434,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5607900087133314,
+      "grad_norm": 0.4427170157432556,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5609061864652919,
+      "grad_norm": 0.41310209035873413,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 4828
+    },
+    {
+      "epoch": 0.5610223642172524,
+      "grad_norm": 0.4155448079109192,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5611385419692129,
+      "grad_norm": 0.42745399475097656,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5612547197211734,
+      "grad_norm": 0.39938148856163025,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5613708974731338,
+      "grad_norm": 0.4226715862751007,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5614870752250944,
+      "grad_norm": 0.4088357985019684,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5616032529770549,
+      "grad_norm": 0.41541755199432373,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5617194307290154,
+      "grad_norm": 0.3974235951900482,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5618356084809759,
+      "grad_norm": 0.42991939187049866,
+      "learning_rate": 0.0001,
+      "loss": 1.495,
+      "step": 4836
+    },
+    {
+      "epoch": 0.5619517862329364,
+      "grad_norm": 0.43215852975845337,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 4837
+    },
+    {
+      "epoch": 0.5620679639848969,
+      "grad_norm": 0.42073243856430054,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5621841417368574,
+      "grad_norm": 0.40232616662979126,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 4839
+    },
+    {
+      "epoch": 0.5623003194888179,
+      "grad_norm": 0.4098654091358185,
+      "learning_rate": 0.0001,
+      "loss": 1.6134,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5624164972407784,
+      "grad_norm": 0.44964614510536194,
+      "learning_rate": 0.0001,
+      "loss": 1.8357,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5625326749927388,
+      "grad_norm": 0.45787712931632996,
+      "learning_rate": 0.0001,
+      "loss": 1.6967,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5626488527446993,
+      "grad_norm": 0.41997238993644714,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5627650304966599,
+      "grad_norm": 0.4027771055698395,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5628812082486204,
+      "grad_norm": 0.3900171220302582,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 4845
+    },
+    {
+      "epoch": 0.5629973860005809,
+      "grad_norm": 0.4267227053642273,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5631135637525414,
+      "grad_norm": 0.42500099539756775,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5632297415045019,
+      "grad_norm": 0.40846553444862366,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5633459192564624,
+      "grad_norm": 0.43166103959083557,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5634620970084229,
+      "grad_norm": 0.41474416851997375,
+      "learning_rate": 0.0001,
+      "loss": 1.7393,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5635782747603834,
+      "grad_norm": 0.4628075361251831,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 4851
+    },
+    {
+      "epoch": 0.5636944525123438,
+      "grad_norm": 0.48346492648124695,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5638106302643043,
+      "grad_norm": 0.4645657241344452,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 4853
+    },
+    {
+      "epoch": 0.5639268080162649,
+      "grad_norm": 0.4429479241371155,
+      "learning_rate": 0.0001,
+      "loss": 1.7976,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5640429857682254,
+      "grad_norm": 0.4967527389526367,
+      "learning_rate": 0.0001,
+      "loss": 1.8657,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5641591635201859,
+      "grad_norm": 0.45798930525779724,
+      "learning_rate": 0.0001,
+      "loss": 1.6908,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5642753412721464,
+      "grad_norm": 0.4299752116203308,
+      "learning_rate": 0.0001,
+      "loss": 1.7577,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5643915190241069,
+      "grad_norm": 0.4171035885810852,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 4858
+    },
+    {
+      "epoch": 0.5645076967760674,
+      "grad_norm": 0.45656847953796387,
+      "learning_rate": 0.0001,
+      "loss": 1.6746,
+      "step": 4859
+    },
+    {
+      "epoch": 0.5646238745280279,
+      "grad_norm": 0.4058765172958374,
+      "learning_rate": 0.0001,
+      "loss": 1.6135,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5647400522799884,
+      "grad_norm": 0.44879722595214844,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5648562300319488,
+      "grad_norm": 0.4731845259666443,
+      "learning_rate": 0.0001,
+      "loss": 1.713,
+      "step": 4862
+    },
+    {
+      "epoch": 0.5649724077839093,
+      "grad_norm": 0.42394986748695374,
+      "learning_rate": 0.0001,
+      "loss": 1.6365,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5650885855358699,
+      "grad_norm": 0.4450361728668213,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 4864
+    },
+    {
+      "epoch": 0.5652047632878304,
+      "grad_norm": 0.3994229733943939,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5653209410397909,
+      "grad_norm": 0.40167900919914246,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5654371187917514,
+      "grad_norm": 0.409967839717865,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 4867
+    },
+    {
+      "epoch": 0.5655532965437119,
+      "grad_norm": 0.4286467432975769,
+      "learning_rate": 0.0001,
+      "loss": 1.7396,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5656694742956724,
+      "grad_norm": 0.4050009250640869,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5657856520476329,
+      "grad_norm": 0.45172083377838135,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5659018297995934,
+      "grad_norm": 0.45170027017593384,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5660180075515538,
+      "grad_norm": 0.43326306343078613,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5661341853035143,
+      "grad_norm": 0.4260367453098297,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5662503630554748,
+      "grad_norm": 0.4030110836029053,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5663665408074354,
+      "grad_norm": 0.4576740860939026,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5664827185593959,
+      "grad_norm": 0.4269350469112396,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5665988963113564,
+      "grad_norm": 0.3865914046764374,
+      "learning_rate": 0.0001,
+      "loss": 1.4036,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5667150740633169,
+      "grad_norm": 0.4161985218524933,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5668312518152774,
+      "grad_norm": 0.4495078921318054,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5669474295672379,
+      "grad_norm": 0.40627461671829224,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5670636073191984,
+      "grad_norm": 0.3978639543056488,
+      "learning_rate": 0.0001,
+      "loss": 1.5106,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5671797850711588,
+      "grad_norm": 0.4051436185836792,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5672959628231193,
+      "grad_norm": 0.44826844334602356,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5674121405750798,
+      "grad_norm": 0.3976137340068817,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5675283183270404,
+      "grad_norm": 0.39482155442237854,
+      "learning_rate": 0.0001,
+      "loss": 1.4746,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5676444960790009,
+      "grad_norm": 0.43112707138061523,
+      "learning_rate": 0.0001,
+      "loss": 1.697,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5677606738309614,
+      "grad_norm": 0.45522767305374146,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5678768515829219,
+      "grad_norm": 0.425699919462204,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5679930293348824,
+      "grad_norm": 0.4110868573188782,
+      "learning_rate": 0.0001,
+      "loss": 1.735,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5681092070868429,
+      "grad_norm": 0.4198397994041443,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5682253848388034,
+      "grad_norm": 0.41459405422210693,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 4891
+    },
+    {
+      "epoch": 0.5683415625907638,
+      "grad_norm": 0.4587966501712799,
+      "learning_rate": 0.0001,
+      "loss": 1.7476,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5684577403427243,
+      "grad_norm": 0.4369155466556549,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5685739180946848,
+      "grad_norm": 0.4800203740596771,
+      "learning_rate": 0.0001,
+      "loss": 1.7202,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5686900958466453,
+      "grad_norm": 0.4436158835887909,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5688062735986059,
+      "grad_norm": 0.4184040129184723,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 4896
+    },
+    {
+      "epoch": 0.5689224513505664,
+      "grad_norm": 0.3854430913925171,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5690386291025269,
+      "grad_norm": 0.4409714639186859,
+      "learning_rate": 0.0001,
+      "loss": 1.7056,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5691548068544874,
+      "grad_norm": 0.4433159828186035,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5692709846064479,
+      "grad_norm": 0.4146018326282501,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5693871623584084,
+      "grad_norm": 0.45392921566963196,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5695033401103688,
+      "grad_norm": 0.42398884892463684,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5696195178623293,
+      "grad_norm": 0.478370726108551,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5697356956142898,
+      "grad_norm": 0.4152291417121887,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 4904
+    },
+    {
+      "epoch": 0.5698518733662503,
+      "grad_norm": 0.4159678816795349,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 4905
+    },
+    {
+      "epoch": 0.5699680511182109,
+      "grad_norm": 0.40162405371665955,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5700842288701714,
+      "grad_norm": 0.44275495409965515,
+      "learning_rate": 0.0001,
+      "loss": 1.6955,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5702004066221319,
+      "grad_norm": 0.43401822447776794,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5703165843740924,
+      "grad_norm": 0.40605849027633667,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5704327621260529,
+      "grad_norm": 0.4131391644477844,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5705489398780134,
+      "grad_norm": 0.4179399311542511,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5706651176299739,
+      "grad_norm": 0.4298674464225769,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5707812953819343,
+      "grad_norm": 0.4349467158317566,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 4913
+    },
+    {
+      "epoch": 0.5708974731338948,
+      "grad_norm": 0.4512692987918854,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5710136508858553,
+      "grad_norm": 0.45529913902282715,
+      "learning_rate": 0.0001,
+      "loss": 1.6938,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5711298286378158,
+      "grad_norm": 0.4262561798095703,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5712460063897764,
+      "grad_norm": 0.40773990750312805,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 4917
+    },
+    {
+      "epoch": 0.5713621841417369,
+      "grad_norm": 0.43575015664100647,
+      "learning_rate": 0.0001,
+      "loss": 1.7376,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5714783618936974,
+      "grad_norm": 0.4189346730709076,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5715945396456579,
+      "grad_norm": 0.4791758954524994,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5717107173976184,
+      "grad_norm": 0.42903390526771545,
+      "learning_rate": 0.0001,
+      "loss": 1.6155,
+      "step": 4921
+    },
+    {
+      "epoch": 0.5718268951495789,
+      "grad_norm": 0.4235144555568695,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5719430729015393,
+      "grad_norm": 0.4023747742176056,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5720592506534998,
+      "grad_norm": 0.44118964672088623,
+      "learning_rate": 0.0001,
+      "loss": 1.7021,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5721754284054603,
+      "grad_norm": 0.4274747669696808,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5722916061574208,
+      "grad_norm": 0.41285374760627747,
+      "learning_rate": 0.0001,
+      "loss": 1.4082,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5724077839093814,
+      "grad_norm": 0.4146759510040283,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5725239616613419,
+      "grad_norm": 0.41832008957862854,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 4928
+    },
+    {
+      "epoch": 0.5726401394133024,
+      "grad_norm": 0.4418119192123413,
+      "learning_rate": 0.0001,
+      "loss": 1.7025,
+      "step": 4929
+    },
+    {
+      "epoch": 0.5727563171652629,
+      "grad_norm": 0.3945986032485962,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 4930
+    },
+    {
+      "epoch": 0.5728724949172234,
+      "grad_norm": 0.44057559967041016,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5729886726691839,
+      "grad_norm": 0.4348897933959961,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5731048504211443,
+      "grad_norm": 0.431241512298584,
+      "learning_rate": 0.0001,
+      "loss": 1.7089,
+      "step": 4933
+    },
+    {
+      "epoch": 0.5732210281731048,
+      "grad_norm": 0.4173518419265747,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5733372059250653,
+      "grad_norm": 0.4776740074157715,
+      "learning_rate": 0.0001,
+      "loss": 1.6631,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5734533836770258,
+      "grad_norm": 0.449663907289505,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5735695614289863,
+      "grad_norm": 0.4311505854129791,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5736857391809469,
+      "grad_norm": 0.4156731963157654,
+      "learning_rate": 0.0001,
+      "loss": 1.8089,
+      "step": 4938
+    },
+    {
+      "epoch": 0.5738019169329074,
+      "grad_norm": 0.4211094379425049,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5739180946848679,
+      "grad_norm": 0.4334612488746643,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5740342724368284,
+      "grad_norm": 0.43656647205352783,
+      "learning_rate": 0.0001,
+      "loss": 1.7283,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5741504501887889,
+      "grad_norm": 0.3772907555103302,
+      "learning_rate": 0.0001,
+      "loss": 1.4156,
+      "step": 4942
+    },
+    {
+      "epoch": 0.5742666279407493,
+      "grad_norm": 0.4248020350933075,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5743828056927098,
+      "grad_norm": 0.4416458010673523,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5744989834446703,
+      "grad_norm": 0.43282851576805115,
+      "learning_rate": 0.0001,
+      "loss": 1.7414,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5746151611966308,
+      "grad_norm": 0.4210634231567383,
+      "learning_rate": 0.0001,
+      "loss": 1.4285,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5747313389485913,
+      "grad_norm": 0.4776458740234375,
+      "learning_rate": 0.0001,
+      "loss": 1.7215,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5748475167005519,
+      "grad_norm": 0.4351522624492645,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5749636944525124,
+      "grad_norm": 0.4363928437232971,
+      "learning_rate": 0.0001,
+      "loss": 1.6422,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5750798722044729,
+      "grad_norm": 0.4042133390903473,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5751960499564334,
+      "grad_norm": 0.4090398848056793,
+      "learning_rate": 0.0001,
+      "loss": 1.3931,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5753122277083939,
+      "grad_norm": 0.4363382160663605,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5754284054603543,
+      "grad_norm": 0.43779557943344116,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5755445832123148,
+      "grad_norm": 0.4067518413066864,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5756607609642753,
+      "grad_norm": 0.45167723298072815,
+      "learning_rate": 0.0001,
+      "loss": 1.726,
+      "step": 4955
+    },
+    {
+      "epoch": 0.5757769387162358,
+      "grad_norm": 0.4339980185031891,
+      "learning_rate": 0.0001,
+      "loss": 1.7253,
+      "step": 4956
+    },
+    {
+      "epoch": 0.5758931164681963,
+      "grad_norm": 0.41764840483665466,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5760092942201568,
+      "grad_norm": 0.4416648745536804,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 4958
+    },
+    {
+      "epoch": 0.5761254719721174,
+      "grad_norm": 0.42781156301498413,
+      "learning_rate": 0.0001,
+      "loss": 1.6848,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5762416497240779,
+      "grad_norm": 0.44320225715637207,
+      "learning_rate": 0.0001,
+      "loss": 1.686,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5763578274760384,
+      "grad_norm": 0.3853711485862732,
+      "learning_rate": 0.0001,
+      "loss": 1.4755,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5764740052279989,
+      "grad_norm": 0.3990086317062378,
+      "learning_rate": 0.0001,
+      "loss": 1.5151,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5765901829799593,
+      "grad_norm": 0.41801634430885315,
+      "learning_rate": 0.0001,
+      "loss": 1.418,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5767063607319198,
+      "grad_norm": 0.4076370596885681,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5768225384838803,
+      "grad_norm": 0.4171597361564636,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5769387162358408,
+      "grad_norm": 0.4075530171394348,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5770548939878013,
+      "grad_norm": 0.4294959008693695,
+      "learning_rate": 0.0001,
+      "loss": 1.7531,
+      "step": 4967
+    },
+    {
+      "epoch": 0.5771710717397618,
+      "grad_norm": 0.4060986340045929,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 4968
+    },
+    {
+      "epoch": 0.5772872494917224,
+      "grad_norm": 0.43765029311180115,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 4969
+    },
+    {
+      "epoch": 0.5774034272436829,
+      "grad_norm": 0.4338558316230774,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5775196049956434,
+      "grad_norm": 0.42239588499069214,
+      "learning_rate": 0.0001,
+      "loss": 1.64,
+      "step": 4971
+    },
+    {
+      "epoch": 0.5776357827476039,
+      "grad_norm": 0.45466890931129456,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5777519604995643,
+      "grad_norm": 0.43936869502067566,
+      "learning_rate": 0.0001,
+      "loss": 1.6832,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5778681382515248,
+      "grad_norm": 0.4495198130607605,
+      "learning_rate": 0.0001,
+      "loss": 1.6946,
+      "step": 4974
+    },
+    {
+      "epoch": 0.5779843160034853,
+      "grad_norm": 0.41554975509643555,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5781004937554458,
+      "grad_norm": 0.3993794620037079,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5782166715074063,
+      "grad_norm": 0.4077496826648712,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 4977
+    },
+    {
+      "epoch": 0.5783328492593668,
+      "grad_norm": 0.45379477739334106,
+      "learning_rate": 0.0001,
+      "loss": 1.7378,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5784490270113273,
+      "grad_norm": 0.4157100319862366,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 4979
+    },
+    {
+      "epoch": 0.5785652047632879,
+      "grad_norm": 0.4228839874267578,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5786813825152484,
+      "grad_norm": 0.43401455879211426,
+      "learning_rate": 0.0001,
+      "loss": 1.7284,
+      "step": 4981
+    },
+    {
+      "epoch": 0.5787975602672089,
+      "grad_norm": 0.43419334292411804,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 4982
+    },
+    {
+      "epoch": 0.5789137380191693,
+      "grad_norm": 0.4071895182132721,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 4983
+    },
+    {
+      "epoch": 0.5790299157711298,
+      "grad_norm": 0.3975391983985901,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 4984
+    },
+    {
+      "epoch": 0.5791460935230903,
+      "grad_norm": 0.4305202066898346,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5792622712750508,
+      "grad_norm": 0.47176504135131836,
+      "learning_rate": 0.0001,
+      "loss": 1.8323,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5793784490270113,
+      "grad_norm": 0.45612940192222595,
+      "learning_rate": 0.0001,
+      "loss": 1.6724,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5794946267789718,
+      "grad_norm": 0.4044199585914612,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5796108045309323,
+      "grad_norm": 0.42781221866607666,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 4989
+    },
+    {
+      "epoch": 0.5797269822828929,
+      "grad_norm": 0.4110262989997864,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 4990
+    },
+    {
+      "epoch": 0.5798431600348534,
+      "grad_norm": 0.43411797285079956,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 4991
+    },
+    {
+      "epoch": 0.5799593377868139,
+      "grad_norm": 0.4334775507450104,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 4992
+    },
+    {
+      "epoch": 0.5800755155387743,
+      "grad_norm": 0.40531855821609497,
+      "learning_rate": 0.0001,
+      "loss": 1.606,
+      "step": 4993
+    },
+    {
+      "epoch": 0.5801916932907348,
+      "grad_norm": 0.41528868675231934,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 4994
+    },
+    {
+      "epoch": 0.5803078710426953,
+      "grad_norm": 0.44848141074180603,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5804240487946558,
+      "grad_norm": 0.4627237021923065,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 4996
+    },
+    {
+      "epoch": 0.5805402265466163,
+      "grad_norm": 0.4355679154396057,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 4997
+    },
+    {
+      "epoch": 0.5806564042985768,
+      "grad_norm": 0.4059557318687439,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 4998
+    },
+    {
+      "epoch": 0.5807725820505373,
+      "grad_norm": 0.4451749324798584,
+      "learning_rate": 0.0001,
+      "loss": 1.8021,
+      "step": 4999
+    },
+    {
+      "epoch": 0.5808887598024978,
+      "grad_norm": 0.47867774963378906,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5810049375544584,
+      "grad_norm": 0.4432399570941925,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 5001
+    },
+    {
+      "epoch": 0.5811211153064189,
+      "grad_norm": 0.40032443404197693,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 5002
+    },
+    {
+      "epoch": 0.5812372930583793,
+      "grad_norm": 0.4400497376918793,
+      "learning_rate": 0.0001,
+      "loss": 1.7494,
+      "step": 5003
+    },
+    {
+      "epoch": 0.5813534708103398,
+      "grad_norm": 0.410427451133728,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 5004
+    },
+    {
+      "epoch": 0.5814696485623003,
+      "grad_norm": 0.3894704580307007,
+      "learning_rate": 0.0001,
+      "loss": 1.6396,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5815858263142608,
+      "grad_norm": 0.4130994975566864,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 5006
+    },
+    {
+      "epoch": 0.5817020040662213,
+      "grad_norm": 0.4316181242465973,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 5007
+    },
+    {
+      "epoch": 0.5818181818181818,
+      "grad_norm": 0.43336644768714905,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 5008
+    },
+    {
+      "epoch": 0.5819343595701423,
+      "grad_norm": 0.41586264967918396,
+      "learning_rate": 0.0001,
+      "loss": 1.6583,
+      "step": 5009
+    },
+    {
+      "epoch": 0.5820505373221028,
+      "grad_norm": 0.398639053106308,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5821667150740634,
+      "grad_norm": 0.42765846848487854,
+      "learning_rate": 0.0001,
+      "loss": 1.7781,
+      "step": 5011
+    },
+    {
+      "epoch": 0.5822828928260239,
+      "grad_norm": 0.420610249042511,
+      "learning_rate": 0.0001,
+      "loss": 1.6591,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5823990705779843,
+      "grad_norm": 0.43733763694763184,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5825152483299448,
+      "grad_norm": 0.46470460295677185,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 5014
+    },
+    {
+      "epoch": 0.5826314260819053,
+      "grad_norm": 0.43464747071266174,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5827476038338658,
+      "grad_norm": 0.42087873816490173,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 5016
+    },
+    {
+      "epoch": 0.5828637815858263,
+      "grad_norm": 0.4221756160259247,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 5017
+    },
+    {
+      "epoch": 0.5829799593377868,
+      "grad_norm": 0.4236156940460205,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5830961370897473,
+      "grad_norm": 0.39433997869491577,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5832123148417078,
+      "grad_norm": 0.4237521290779114,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5833284925936683,
+      "grad_norm": 0.44210362434387207,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 5021
+    },
+    {
+      "epoch": 0.5834446703456289,
+      "grad_norm": 0.4477832317352295,
+      "learning_rate": 0.0001,
+      "loss": 1.7188,
+      "step": 5022
+    },
+    {
+      "epoch": 0.5835608480975893,
+      "grad_norm": 0.39492443203926086,
+      "learning_rate": 0.0001,
+      "loss": 1.3972,
+      "step": 5023
+    },
+    {
+      "epoch": 0.5836770258495498,
+      "grad_norm": 0.42422086000442505,
+      "learning_rate": 0.0001,
+      "loss": 1.6085,
+      "step": 5024
+    },
+    {
+      "epoch": 0.5837932036015103,
+      "grad_norm": 0.4569999575614929,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5839093813534708,
+      "grad_norm": 0.38789770007133484,
+      "learning_rate": 0.0001,
+      "loss": 1.4186,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5840255591054313,
+      "grad_norm": 0.4335310161113739,
+      "learning_rate": 0.0001,
+      "loss": 1.4464,
+      "step": 5027
+    },
+    {
+      "epoch": 0.5841417368573918,
+      "grad_norm": 0.4649960398674011,
+      "learning_rate": 0.0001,
+      "loss": 1.7433,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5842579146093523,
+      "grad_norm": 0.41210901737213135,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5843740923613128,
+      "grad_norm": 0.4519018232822418,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5844902701132733,
+      "grad_norm": 0.45829200744628906,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 5031
+    },
+    {
+      "epoch": 0.5846064478652339,
+      "grad_norm": 0.4412051737308502,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 5032
+    },
+    {
+      "epoch": 0.5847226256171943,
+      "grad_norm": 0.4298493266105652,
+      "learning_rate": 0.0001,
+      "loss": 1.71,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5848388033691548,
+      "grad_norm": 0.47564446926116943,
+      "learning_rate": 0.0001,
+      "loss": 1.7355,
+      "step": 5034
+    },
+    {
+      "epoch": 0.5849549811211153,
+      "grad_norm": 0.4598512053489685,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5850711588730758,
+      "grad_norm": 0.4496704041957855,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5851873366250363,
+      "grad_norm": 0.4410554766654968,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5853035143769968,
+      "grad_norm": 0.4468580484390259,
+      "learning_rate": 0.0001,
+      "loss": 1.7235,
+      "step": 5038
+    },
+    {
+      "epoch": 0.5854196921289573,
+      "grad_norm": 0.41189512610435486,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 5039
+    },
+    {
+      "epoch": 0.5855358698809178,
+      "grad_norm": 0.38999536633491516,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 5040
+    },
+    {
+      "epoch": 0.5856520476328783,
+      "grad_norm": 0.40551015734672546,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 5041
+    },
+    {
+      "epoch": 0.5857682253848389,
+      "grad_norm": 0.4540075957775116,
+      "learning_rate": 0.0001,
+      "loss": 1.7988,
+      "step": 5042
+    },
+    {
+      "epoch": 0.5858844031367993,
+      "grad_norm": 0.4358082115650177,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 5043
+    },
+    {
+      "epoch": 0.5860005808887598,
+      "grad_norm": 0.42553818225860596,
+      "learning_rate": 0.0001,
+      "loss": 1.5983,
+      "step": 5044
+    },
+    {
+      "epoch": 0.5861167586407203,
+      "grad_norm": 0.4158061444759369,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5862329363926808,
+      "grad_norm": 0.4467496871948242,
+      "learning_rate": 0.0001,
+      "loss": 1.79,
+      "step": 5046
+    },
+    {
+      "epoch": 0.5863491141446413,
+      "grad_norm": 0.4339059293270111,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 5047
+    },
+    {
+      "epoch": 0.5864652918966018,
+      "grad_norm": 0.4463905096054077,
+      "learning_rate": 0.0001,
+      "loss": 1.755,
+      "step": 5048
+    },
+    {
+      "epoch": 0.5865814696485623,
+      "grad_norm": 0.4336758553981781,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 5049
+    },
+    {
+      "epoch": 0.5866976474005228,
+      "grad_norm": 0.3911759555339813,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5868138251524833,
+      "grad_norm": 0.43406540155410767,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 5051
+    },
+    {
+      "epoch": 0.5869300029044437,
+      "grad_norm": 0.40602606534957886,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 5052
+    },
+    {
+      "epoch": 0.5870461806564043,
+      "grad_norm": 0.40257728099823,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 5053
+    },
+    {
+      "epoch": 0.5871623584083648,
+      "grad_norm": 0.42640700936317444,
+      "learning_rate": 0.0001,
+      "loss": 1.6454,
+      "step": 5054
+    },
+    {
+      "epoch": 0.5872785361603253,
+      "grad_norm": 0.4607278108596802,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5873947139122858,
+      "grad_norm": 0.456226110458374,
+      "learning_rate": 0.0001,
+      "loss": 1.7054,
+      "step": 5056
+    },
+    {
+      "epoch": 0.5875108916642463,
+      "grad_norm": 0.44643908739089966,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 5057
+    },
+    {
+      "epoch": 0.5876270694162068,
+      "grad_norm": 0.4707536995410919,
+      "learning_rate": 0.0001,
+      "loss": 1.7986,
+      "step": 5058
+    },
+    {
+      "epoch": 0.5877432471681673,
+      "grad_norm": 0.437148779630661,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 5059
+    },
+    {
+      "epoch": 0.5878594249201278,
+      "grad_norm": 0.4139115810394287,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5879756026720883,
+      "grad_norm": 0.44146662950515747,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 5061
+    },
+    {
+      "epoch": 0.5880917804240487,
+      "grad_norm": 0.41695713996887207,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 5062
+    },
+    {
+      "epoch": 0.5882079581760093,
+      "grad_norm": 0.45756304264068604,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 5063
+    },
+    {
+      "epoch": 0.5883241359279698,
+      "grad_norm": 0.4567687213420868,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 5064
+    },
+    {
+      "epoch": 0.5884403136799303,
+      "grad_norm": 0.43768954277038574,
+      "learning_rate": 0.0001,
+      "loss": 1.6733,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5885564914318908,
+      "grad_norm": 0.4621203541755676,
+      "learning_rate": 0.0001,
+      "loss": 1.7541,
+      "step": 5066
+    },
+    {
+      "epoch": 0.5886726691838513,
+      "grad_norm": 0.420998215675354,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 5067
+    },
+    {
+      "epoch": 0.5887888469358118,
+      "grad_norm": 0.45057764649391174,
+      "learning_rate": 0.0001,
+      "loss": 1.763,
+      "step": 5068
+    },
+    {
+      "epoch": 0.5889050246877723,
+      "grad_norm": 0.47149190306663513,
+      "learning_rate": 0.0001,
+      "loss": 1.8066,
+      "step": 5069
+    },
+    {
+      "epoch": 0.5890212024397328,
+      "grad_norm": 0.4899585545063019,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5891373801916933,
+      "grad_norm": 0.4497217833995819,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 5071
+    },
+    {
+      "epoch": 0.5892535579436537,
+      "grad_norm": 0.47766372561454773,
+      "learning_rate": 0.0001,
+      "loss": 1.7802,
+      "step": 5072
+    },
+    {
+      "epoch": 0.5893697356956142,
+      "grad_norm": 0.4019678831100464,
+      "learning_rate": 0.0001,
+      "loss": 1.413,
+      "step": 5073
+    },
+    {
+      "epoch": 0.5894859134475748,
+      "grad_norm": 0.40686938166618347,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 5074
+    },
+    {
+      "epoch": 0.5896020911995353,
+      "grad_norm": 0.417871356010437,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5897182689514958,
+      "grad_norm": 0.4329603612422943,
+      "learning_rate": 0.0001,
+      "loss": 1.7662,
+      "step": 5076
+    },
+    {
+      "epoch": 0.5898344467034563,
+      "grad_norm": 0.41233643889427185,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 5077
+    },
+    {
+      "epoch": 0.5899506244554168,
+      "grad_norm": 0.4218975603580475,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 5078
+    },
+    {
+      "epoch": 0.5900668022073773,
+      "grad_norm": 0.4207758605480194,
+      "learning_rate": 0.0001,
+      "loss": 1.6849,
+      "step": 5079
+    },
+    {
+      "epoch": 0.5901829799593378,
+      "grad_norm": 0.4465838670730591,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5902991577112983,
+      "grad_norm": 0.43389689922332764,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 5081
+    },
+    {
+      "epoch": 0.5904153354632588,
+      "grad_norm": 0.4294486343860626,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 5082
+    },
+    {
+      "epoch": 0.5905315132152192,
+      "grad_norm": 0.45105913281440735,
+      "learning_rate": 0.0001,
+      "loss": 1.7492,
+      "step": 5083
+    },
+    {
+      "epoch": 0.5906476909671798,
+      "grad_norm": 0.3978818655014038,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 5084
+    },
+    {
+      "epoch": 0.5907638687191403,
+      "grad_norm": 0.44959375262260437,
+      "learning_rate": 0.0001,
+      "loss": 1.7158,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5908800464711008,
+      "grad_norm": 0.40922439098358154,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 5086
+    },
+    {
+      "epoch": 0.5909962242230613,
+      "grad_norm": 0.41393378376960754,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 5087
+    },
+    {
+      "epoch": 0.5911124019750218,
+      "grad_norm": 0.4593432545661926,
+      "learning_rate": 0.0001,
+      "loss": 1.691,
+      "step": 5088
+    },
+    {
+      "epoch": 0.5912285797269823,
+      "grad_norm": 0.4352717995643616,
+      "learning_rate": 0.0001,
+      "loss": 1.5997,
+      "step": 5089
+    },
+    {
+      "epoch": 0.5913447574789428,
+      "grad_norm": 0.4431297779083252,
+      "learning_rate": 0.0001,
+      "loss": 1.7196,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5914609352309033,
+      "grad_norm": 0.4279361069202423,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 5091
+    },
+    {
+      "epoch": 0.5915771129828638,
+      "grad_norm": 0.4590848386287689,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 5092
+    },
+    {
+      "epoch": 0.5916932907348242,
+      "grad_norm": 0.4331355094909668,
+      "learning_rate": 0.0001,
+      "loss": 1.4664,
+      "step": 5093
+    },
+    {
+      "epoch": 0.5918094684867847,
+      "grad_norm": 0.4613255262374878,
+      "learning_rate": 0.0001,
+      "loss": 1.8481,
+      "step": 5094
+    },
+    {
+      "epoch": 0.5919256462387453,
+      "grad_norm": 0.46412867307662964,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5920418239907058,
+      "grad_norm": 0.4639713764190674,
+      "learning_rate": 0.0001,
+      "loss": 1.6082,
+      "step": 5096
+    },
+    {
+      "epoch": 0.5921580017426663,
+      "grad_norm": 0.48971736431121826,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 5097
+    },
+    {
+      "epoch": 0.5922741794946268,
+      "grad_norm": 0.48425406217575073,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 5098
+    },
+    {
+      "epoch": 0.5923903572465873,
+      "grad_norm": 0.45319536328315735,
+      "learning_rate": 0.0001,
+      "loss": 1.7629,
+      "step": 5099
+    },
+    {
+      "epoch": 0.5925065349985478,
+      "grad_norm": 0.436058908700943,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 5100
+    },
+    {
+      "epoch": 0.5926227127505083,
+      "grad_norm": 0.4272605776786804,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 5101
+    },
+    {
+      "epoch": 0.5927388905024688,
+      "grad_norm": 0.4012027978897095,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 5102
+    },
+    {
+      "epoch": 0.5928550682544292,
+      "grad_norm": 0.42485886812210083,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 5103
+    },
+    {
+      "epoch": 0.5929712460063897,
+      "grad_norm": 0.45090624690055847,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5930874237583503,
+      "grad_norm": 0.45741748809814453,
+      "learning_rate": 0.0001,
+      "loss": 1.7975,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5932036015103108,
+      "grad_norm": 0.48514753580093384,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 5106
+    },
+    {
+      "epoch": 0.5933197792622713,
+      "grad_norm": 0.41915470361709595,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 5107
+    },
+    {
+      "epoch": 0.5934359570142318,
+      "grad_norm": 0.3943340480327606,
+      "learning_rate": 0.0001,
+      "loss": 1.4051,
+      "step": 5108
+    },
+    {
+      "epoch": 0.5935521347661923,
+      "grad_norm": 0.4319881796836853,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 5109
+    },
+    {
+      "epoch": 0.5936683125181528,
+      "grad_norm": 0.41222912073135376,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5937844902701133,
+      "grad_norm": 0.4626607894897461,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 5111
+    },
+    {
+      "epoch": 0.5939006680220738,
+      "grad_norm": 0.44612380862236023,
+      "learning_rate": 0.0001,
+      "loss": 1.7116,
+      "step": 5112
+    },
+    {
+      "epoch": 0.5940168457740342,
+      "grad_norm": 0.44148531556129456,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 5113
+    },
+    {
+      "epoch": 0.5941330235259947,
+      "grad_norm": 0.4143647253513336,
+      "learning_rate": 0.0001,
+      "loss": 1.4093,
+      "step": 5114
+    },
+    {
+      "epoch": 0.5942492012779552,
+      "grad_norm": 0.4281814992427826,
+      "learning_rate": 0.0001,
+      "loss": 1.6976,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5943653790299158,
+      "grad_norm": 0.44174307584762573,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 5116
+    },
+    {
+      "epoch": 0.5944815567818763,
+      "grad_norm": 0.4306361675262451,
+      "learning_rate": 0.0001,
+      "loss": 1.6543,
+      "step": 5117
+    },
+    {
+      "epoch": 0.5945977345338368,
+      "grad_norm": 0.46144962310791016,
+      "learning_rate": 0.0001,
+      "loss": 1.8373,
+      "step": 5118
+    },
+    {
+      "epoch": 0.5947139122857973,
+      "grad_norm": 0.43401268124580383,
+      "learning_rate": 0.0001,
+      "loss": 1.6625,
+      "step": 5119
+    },
+    {
+      "epoch": 0.5948300900377578,
+      "grad_norm": 0.4165130853652954,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5949462677897183,
+      "grad_norm": 0.46252015233039856,
+      "learning_rate": 0.0001,
+      "loss": 1.7019,
+      "step": 5121
+    },
+    {
+      "epoch": 0.5950624455416788,
+      "grad_norm": 0.40529441833496094,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 5122
+    },
+    {
+      "epoch": 0.5951786232936392,
+      "grad_norm": 0.44284263253211975,
+      "learning_rate": 0.0001,
+      "loss": 1.7003,
+      "step": 5123
+    },
+    {
+      "epoch": 0.5952948010455997,
+      "grad_norm": 0.4453653395175934,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 5124
+    },
+    {
+      "epoch": 0.5954109787975602,
+      "grad_norm": 0.4209020137786865,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 5125
+    },
+    {
+      "epoch": 0.5955271565495208,
+      "grad_norm": 0.4327007234096527,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 5126
+    },
+    {
+      "epoch": 0.5956433343014813,
+      "grad_norm": 0.4445003271102905,
+      "learning_rate": 0.0001,
+      "loss": 1.7994,
+      "step": 5127
+    },
+    {
+      "epoch": 0.5957595120534418,
+      "grad_norm": 0.40100163221359253,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5958756898054023,
+      "grad_norm": 0.4560073912143707,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 5129
+    },
+    {
+      "epoch": 0.5959918675573628,
+      "grad_norm": 0.42099544405937195,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5961080453093233,
+      "grad_norm": 0.40899744629859924,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 5131
+    },
+    {
+      "epoch": 0.5962242230612838,
+      "grad_norm": 0.3647748827934265,
+      "learning_rate": 0.0001,
+      "loss": 1.3157,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5963404008132442,
+      "grad_norm": 0.43523362278938293,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 5133
+    },
+    {
+      "epoch": 0.5964565785652047,
+      "grad_norm": 0.40236619114875793,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 5134
+    },
+    {
+      "epoch": 0.5965727563171652,
+      "grad_norm": 0.43228352069854736,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5966889340691257,
+      "grad_norm": 0.41964632272720337,
+      "learning_rate": 0.0001,
+      "loss": 1.4214,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5968051118210863,
+      "grad_norm": 0.5177493691444397,
+      "learning_rate": 0.0001,
+      "loss": 1.7107,
+      "step": 5137
+    },
+    {
+      "epoch": 0.5969212895730468,
+      "grad_norm": 0.40368756651878357,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 5138
+    },
+    {
+      "epoch": 0.5970374673250073,
+      "grad_norm": 0.49018216133117676,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 5139
+    },
+    {
+      "epoch": 0.5971536450769678,
+      "grad_norm": 0.4471641480922699,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5972698228289283,
+      "grad_norm": 0.4672026038169861,
+      "learning_rate": 0.0001,
+      "loss": 1.7613,
+      "step": 5141
+    },
+    {
+      "epoch": 0.5973860005808888,
+      "grad_norm": 0.49104636907577515,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 5142
+    },
+    {
+      "epoch": 0.5975021783328492,
+      "grad_norm": 0.4686024487018585,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 5143
+    },
+    {
+      "epoch": 0.5976183560848097,
+      "grad_norm": 0.41973164677619934,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 5144
+    },
+    {
+      "epoch": 0.5977345338367702,
+      "grad_norm": 0.43972551822662354,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5978507115887307,
+      "grad_norm": 0.4550989270210266,
+      "learning_rate": 0.0001,
+      "loss": 1.8349,
+      "step": 5146
+    },
+    {
+      "epoch": 0.5979668893406913,
+      "grad_norm": 0.452958345413208,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 5147
+    },
+    {
+      "epoch": 0.5980830670926518,
+      "grad_norm": 0.4228680431842804,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 5148
+    },
+    {
+      "epoch": 0.5981992448446123,
+      "grad_norm": 0.4337772727012634,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 5149
+    },
+    {
+      "epoch": 0.5983154225965728,
+      "grad_norm": 0.3951295018196106,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5984316003485333,
+      "grad_norm": 0.4213055670261383,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 5151
+    },
+    {
+      "epoch": 0.5985477781004938,
+      "grad_norm": 0.4383963644504547,
+      "learning_rate": 0.0001,
+      "loss": 1.7536,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5986639558524542,
+      "grad_norm": 0.43566179275512695,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 5153
+    },
+    {
+      "epoch": 0.5987801336044147,
+      "grad_norm": 0.4275076985359192,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 5154
+    },
+    {
+      "epoch": 0.5988963113563752,
+      "grad_norm": 0.4178537130355835,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5990124891083357,
+      "grad_norm": 0.4059988558292389,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5991286668602962,
+      "grad_norm": 0.44386255741119385,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 5157
+    },
+    {
+      "epoch": 0.5992448446122568,
+      "grad_norm": 0.430132657289505,
+      "learning_rate": 0.0001,
+      "loss": 1.7439,
+      "step": 5158
+    },
+    {
+      "epoch": 0.5993610223642173,
+      "grad_norm": 0.4542747735977173,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 5159
+    },
+    {
+      "epoch": 0.5994772001161778,
+      "grad_norm": 0.4200115203857422,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5995933778681383,
+      "grad_norm": 0.4746837615966797,
+      "learning_rate": 0.0001,
+      "loss": 1.758,
+      "step": 5161
+    },
+    {
+      "epoch": 0.5997095556200988,
+      "grad_norm": 0.45345309376716614,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 5162
+    },
+    {
+      "epoch": 0.5998257333720592,
+      "grad_norm": 0.4250603914260864,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 5163
+    },
+    {
+      "epoch": 0.5999419111240197,
+      "grad_norm": 0.4513823688030243,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 5164
+    },
+    {
+      "epoch": 0.6000580888759802,
+      "grad_norm": 0.4406159520149231,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 5165
+    },
+    {
+      "epoch": 0.6001742666279407,
+      "grad_norm": 0.45702022314071655,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 5166
+    },
+    {
+      "epoch": 0.6002904443799012,
+      "grad_norm": 0.4588814079761505,
+      "learning_rate": 0.0001,
+      "loss": 1.7822,
+      "step": 5167
+    },
+    {
+      "epoch": 0.6004066221318618,
+      "grad_norm": 0.4281061291694641,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 5168
+    },
+    {
+      "epoch": 0.6005227998838223,
+      "grad_norm": 0.42194947600364685,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 5169
+    },
+    {
+      "epoch": 0.6006389776357828,
+      "grad_norm": 0.43042877316474915,
+      "learning_rate": 0.0001,
+      "loss": 1.6838,
+      "step": 5170
+    },
+    {
+      "epoch": 0.6007551553877433,
+      "grad_norm": 0.40578198432922363,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 5171
+    },
+    {
+      "epoch": 0.6008713331397038,
+      "grad_norm": 0.41954731941223145,
+      "learning_rate": 0.0001,
+      "loss": 1.6543,
+      "step": 5172
+    },
+    {
+      "epoch": 0.6009875108916642,
+      "grad_norm": 0.4439033567905426,
+      "learning_rate": 0.0001,
+      "loss": 1.6539,
+      "step": 5173
+    },
+    {
+      "epoch": 0.6011036886436247,
+      "grad_norm": 0.41968825459480286,
+      "learning_rate": 0.0001,
+      "loss": 1.3706,
+      "step": 5174
+    },
+    {
+      "epoch": 0.6012198663955852,
+      "grad_norm": 0.480570524930954,
+      "learning_rate": 0.0001,
+      "loss": 1.7536,
+      "step": 5175
+    },
+    {
+      "epoch": 0.6013360441475457,
+      "grad_norm": 0.45514610409736633,
+      "learning_rate": 0.0001,
+      "loss": 1.541,
+      "step": 5176
+    },
+    {
+      "epoch": 0.6014522218995062,
+      "grad_norm": 0.4680652618408203,
+      "learning_rate": 0.0001,
+      "loss": 1.7239,
+      "step": 5177
+    },
+    {
+      "epoch": 0.6015683996514667,
+      "grad_norm": 0.47438210248947144,
+      "learning_rate": 0.0001,
+      "loss": 1.6592,
+      "step": 5178
+    },
+    {
+      "epoch": 0.6016845774034273,
+      "grad_norm": 0.41302818059921265,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 5179
+    },
+    {
+      "epoch": 0.6018007551553878,
+      "grad_norm": 0.4282575845718384,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 5180
+    },
+    {
+      "epoch": 0.6019169329073483,
+      "grad_norm": 0.4566362202167511,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 5181
+    },
+    {
+      "epoch": 0.6020331106593088,
+      "grad_norm": 0.42509856820106506,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 5182
+    },
+    {
+      "epoch": 0.6021492884112692,
+      "grad_norm": 0.4111645817756653,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 5183
+    },
+    {
+      "epoch": 0.6022654661632297,
+      "grad_norm": 0.45297473669052124,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 5184
+    },
+    {
+      "epoch": 0.6023816439151902,
+      "grad_norm": 0.4520341157913208,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 5185
+    },
+    {
+      "epoch": 0.6024978216671507,
+      "grad_norm": 0.4603181481361389,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 5186
+    },
+    {
+      "epoch": 0.6026139994191112,
+      "grad_norm": 0.4459368884563446,
+      "learning_rate": 0.0001,
+      "loss": 1.6856,
+      "step": 5187
+    },
+    {
+      "epoch": 0.6027301771710717,
+      "grad_norm": 0.42652809619903564,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 5188
+    },
+    {
+      "epoch": 0.6028463549230323,
+      "grad_norm": 0.41247275471687317,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 5189
+    },
+    {
+      "epoch": 0.6029625326749928,
+      "grad_norm": 0.48231571912765503,
+      "learning_rate": 0.0001,
+      "loss": 1.8699,
+      "step": 5190
+    },
+    {
+      "epoch": 0.6030787104269533,
+      "grad_norm": 0.43310195207595825,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 5191
+    },
+    {
+      "epoch": 0.6031948881789138,
+      "grad_norm": 0.4534681737422943,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 5192
+    },
+    {
+      "epoch": 0.6033110659308742,
+      "grad_norm": 0.41649046540260315,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 5193
+    },
+    {
+      "epoch": 0.6034272436828347,
+      "grad_norm": 0.4298163950443268,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 5194
+    },
+    {
+      "epoch": 0.6035434214347952,
+      "grad_norm": 0.43806689977645874,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 5195
+    },
+    {
+      "epoch": 0.6036595991867557,
+      "grad_norm": 0.4382534325122833,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 5196
+    },
+    {
+      "epoch": 0.6037757769387162,
+      "grad_norm": 0.44435277581214905,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 5197
+    },
+    {
+      "epoch": 0.6038919546906767,
+      "grad_norm": 0.42443951964378357,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 5198
+    },
+    {
+      "epoch": 0.6040081324426372,
+      "grad_norm": 0.43603894114494324,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 5199
+    },
+    {
+      "epoch": 0.6041243101945978,
+      "grad_norm": 0.43126562237739563,
+      "learning_rate": 0.0001,
+      "loss": 1.4945,
+      "step": 5200
+    },
+    {
+      "epoch": 0.6042404879465583,
+      "grad_norm": 0.42838260531425476,
+      "learning_rate": 0.0001,
+      "loss": 1.7171,
+      "step": 5201
+    },
+    {
+      "epoch": 0.6043566656985188,
+      "grad_norm": 0.43312254548072815,
+      "learning_rate": 0.0001,
+      "loss": 1.6684,
+      "step": 5202
+    },
+    {
+      "epoch": 0.6044728434504792,
+      "grad_norm": 0.4641507863998413,
+      "learning_rate": 0.0001,
+      "loss": 1.7897,
+      "step": 5203
+    },
+    {
+      "epoch": 0.6045890212024397,
+      "grad_norm": 0.4522157311439514,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 5204
+    },
+    {
+      "epoch": 0.6047051989544002,
+      "grad_norm": 0.4437788128852844,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 5205
+    },
+    {
+      "epoch": 0.6048213767063607,
+      "grad_norm": 0.40446487069129944,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 5206
+    },
+    {
+      "epoch": 0.6049375544583212,
+      "grad_norm": 0.47073522210121155,
+      "learning_rate": 0.0001,
+      "loss": 1.92,
+      "step": 5207
+    },
+    {
+      "epoch": 0.6050537322102817,
+      "grad_norm": 0.4400515556335449,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 5208
+    },
+    {
+      "epoch": 0.6051699099622422,
+      "grad_norm": 0.41592809557914734,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 5209
+    },
+    {
+      "epoch": 0.6052860877142028,
+      "grad_norm": 0.44804421067237854,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 5210
+    },
+    {
+      "epoch": 0.6054022654661633,
+      "grad_norm": 0.4076915979385376,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 5211
+    },
+    {
+      "epoch": 0.6055184432181238,
+      "grad_norm": 0.42822811007499695,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 5212
+    },
+    {
+      "epoch": 0.6056346209700842,
+      "grad_norm": 0.40919166803359985,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 5213
+    },
+    {
+      "epoch": 0.6057507987220447,
+      "grad_norm": 0.4067077040672302,
+      "learning_rate": 0.0001,
+      "loss": 1.4705,
+      "step": 5214
+    },
+    {
+      "epoch": 0.6058669764740052,
+      "grad_norm": 0.44207772612571716,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 5215
+    },
+    {
+      "epoch": 0.6059831542259657,
+      "grad_norm": 0.4533270001411438,
+      "learning_rate": 0.0001,
+      "loss": 1.7419,
+      "step": 5216
+    },
+    {
+      "epoch": 0.6060993319779262,
+      "grad_norm": 0.4846879839897156,
+      "learning_rate": 0.0001,
+      "loss": 1.8105,
+      "step": 5217
+    },
+    {
+      "epoch": 0.6062155097298867,
+      "grad_norm": 0.48439884185791016,
+      "learning_rate": 0.0001,
+      "loss": 1.8875,
+      "step": 5218
+    },
+    {
+      "epoch": 0.6063316874818472,
+      "grad_norm": 0.418059766292572,
+      "learning_rate": 0.0001,
+      "loss": 1.4664,
+      "step": 5219
+    },
+    {
+      "epoch": 0.6064478652338078,
+      "grad_norm": 0.4088701903820038,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 5220
+    },
+    {
+      "epoch": 0.6065640429857683,
+      "grad_norm": 0.4230385720729828,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 5221
+    },
+    {
+      "epoch": 0.6066802207377288,
+      "grad_norm": 0.4391116797924042,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 5222
+    },
+    {
+      "epoch": 0.6067963984896892,
+      "grad_norm": 0.44503310322761536,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 5223
+    },
+    {
+      "epoch": 0.6069125762416497,
+      "grad_norm": 0.4621829390525818,
+      "learning_rate": 0.0001,
+      "loss": 1.743,
+      "step": 5224
+    },
+    {
+      "epoch": 0.6070287539936102,
+      "grad_norm": 0.4562551975250244,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 5225
+    },
+    {
+      "epoch": 0.6071449317455707,
+      "grad_norm": 0.4189031720161438,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 5226
+    },
+    {
+      "epoch": 0.6072611094975312,
+      "grad_norm": 0.45642954111099243,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 5227
+    },
+    {
+      "epoch": 0.6073772872494917,
+      "grad_norm": 0.4381450116634369,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 5228
+    },
+    {
+      "epoch": 0.6074934650014522,
+      "grad_norm": 0.4203230142593384,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 5229
+    },
+    {
+      "epoch": 0.6076096427534127,
+      "grad_norm": 0.41312652826309204,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 5230
+    },
+    {
+      "epoch": 0.6077258205053733,
+      "grad_norm": 0.46580955386161804,
+      "learning_rate": 0.0001,
+      "loss": 1.7905,
+      "step": 5231
+    },
+    {
+      "epoch": 0.6078419982573338,
+      "grad_norm": 0.42524558305740356,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 5232
+    },
+    {
+      "epoch": 0.6079581760092942,
+      "grad_norm": 0.43027788400650024,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 5233
+    },
+    {
+      "epoch": 0.6080743537612547,
+      "grad_norm": 0.4252438545227051,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 5234
+    },
+    {
+      "epoch": 0.6081905315132152,
+      "grad_norm": 0.4131147861480713,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 5235
+    },
+    {
+      "epoch": 0.6083067092651757,
+      "grad_norm": 0.48660963773727417,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 5236
+    },
+    {
+      "epoch": 0.6084228870171362,
+      "grad_norm": 0.43840891122817993,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 5237
+    },
+    {
+      "epoch": 0.6085390647690967,
+      "grad_norm": 0.48248180747032166,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 5238
+    },
+    {
+      "epoch": 0.6086552425210572,
+      "grad_norm": 0.45849937200546265,
+      "learning_rate": 0.0001,
+      "loss": 1.7877,
+      "step": 5239
+    },
+    {
+      "epoch": 0.6087714202730177,
+      "grad_norm": 0.43766871094703674,
+      "learning_rate": 0.0001,
+      "loss": 1.741,
+      "step": 5240
+    },
+    {
+      "epoch": 0.6088875980249783,
+      "grad_norm": 0.41059789061546326,
+      "learning_rate": 0.0001,
+      "loss": 1.7146,
+      "step": 5241
+    },
+    {
+      "epoch": 0.6090037757769388,
+      "grad_norm": 0.40786266326904297,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 5242
+    },
+    {
+      "epoch": 0.6091199535288992,
+      "grad_norm": 0.4507090747356415,
+      "learning_rate": 0.0001,
+      "loss": 1.6575,
+      "step": 5243
+    },
+    {
+      "epoch": 0.6092361312808597,
+      "grad_norm": 0.45171281695365906,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 5244
+    },
+    {
+      "epoch": 0.6093523090328202,
+      "grad_norm": 0.4520759582519531,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 5245
+    },
+    {
+      "epoch": 0.6094684867847807,
+      "grad_norm": 0.4407515823841095,
+      "learning_rate": 0.0001,
+      "loss": 1.4963,
+      "step": 5246
+    },
+    {
+      "epoch": 0.6095846645367412,
+      "grad_norm": 0.49019014835357666,
+      "learning_rate": 0.0001,
+      "loss": 1.7029,
+      "step": 5247
+    },
+    {
+      "epoch": 0.6097008422887017,
+      "grad_norm": 0.4346254765987396,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 5248
+    },
+    {
+      "epoch": 0.6098170200406622,
+      "grad_norm": 0.43381422758102417,
+      "learning_rate": 0.0001,
+      "loss": 1.6572,
+      "step": 5249
+    },
+    {
+      "epoch": 0.6099331977926227,
+      "grad_norm": 0.43752437829971313,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 5250
+    },
+    {
+      "epoch": 0.6100493755445832,
+      "grad_norm": 0.4531850814819336,
+      "learning_rate": 0.0001,
+      "loss": 1.7279,
+      "step": 5251
+    },
+    {
+      "epoch": 0.6101655532965438,
+      "grad_norm": 0.4308491051197052,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 5252
+    },
+    {
+      "epoch": 0.6102817310485043,
+      "grad_norm": 0.45477020740509033,
+      "learning_rate": 0.0001,
+      "loss": 1.7004,
+      "step": 5253
+    },
+    {
+      "epoch": 0.6103979088004647,
+      "grad_norm": 0.42559128999710083,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 5254
+    },
+    {
+      "epoch": 0.6105140865524252,
+      "grad_norm": 0.45015910267829895,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 5255
+    },
+    {
+      "epoch": 0.6106302643043857,
+      "grad_norm": 0.44628676772117615,
+      "learning_rate": 0.0001,
+      "loss": 1.7035,
+      "step": 5256
+    },
+    {
+      "epoch": 0.6107464420563462,
+      "grad_norm": 0.4328848421573639,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 5257
+    },
+    {
+      "epoch": 0.6108626198083067,
+      "grad_norm": 0.42768430709838867,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 5258
+    },
+    {
+      "epoch": 0.6109787975602672,
+      "grad_norm": 0.4138183891773224,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 5259
+    },
+    {
+      "epoch": 0.6110949753122277,
+      "grad_norm": 0.40088340640068054,
+      "learning_rate": 0.0001,
+      "loss": 1.2582,
+      "step": 5260
+    },
+    {
+      "epoch": 0.6112111530641882,
+      "grad_norm": 0.47067928314208984,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 5261
+    },
+    {
+      "epoch": 0.6113273308161488,
+      "grad_norm": 0.4110230803489685,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 5262
+    },
+    {
+      "epoch": 0.6114435085681093,
+      "grad_norm": 0.4310549795627594,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 5263
+    },
+    {
+      "epoch": 0.6115596863200697,
+      "grad_norm": 0.4675642251968384,
+      "learning_rate": 0.0001,
+      "loss": 1.7689,
+      "step": 5264
+    },
+    {
+      "epoch": 0.6116758640720302,
+      "grad_norm": 0.4433719217777252,
+      "learning_rate": 0.0001,
+      "loss": 1.6569,
+      "step": 5265
+    },
+    {
+      "epoch": 0.6117920418239907,
+      "grad_norm": 0.4305296242237091,
+      "learning_rate": 0.0001,
+      "loss": 1.5723,
+      "step": 5266
+    },
+    {
+      "epoch": 0.6119082195759512,
+      "grad_norm": 0.42353424429893494,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 5267
+    },
+    {
+      "epoch": 0.6120243973279117,
+      "grad_norm": 0.43402886390686035,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 5268
+    },
+    {
+      "epoch": 0.6121405750798722,
+      "grad_norm": 0.41518083214759827,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 5269
+    },
+    {
+      "epoch": 0.6122567528318327,
+      "grad_norm": 0.42695778608322144,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 5270
+    },
+    {
+      "epoch": 0.6123729305837932,
+      "grad_norm": 0.41428840160369873,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 5271
+    },
+    {
+      "epoch": 0.6124891083357537,
+      "grad_norm": 0.42342090606689453,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 5272
+    },
+    {
+      "epoch": 0.6126052860877143,
+      "grad_norm": 0.479927122592926,
+      "learning_rate": 0.0001,
+      "loss": 1.838,
+      "step": 5273
+    },
+    {
+      "epoch": 0.6127214638396747,
+      "grad_norm": 0.4496026039123535,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 5274
+    },
+    {
+      "epoch": 0.6128376415916352,
+      "grad_norm": 0.4338063895702362,
+      "learning_rate": 0.0001,
+      "loss": 1.6543,
+      "step": 5275
+    },
+    {
+      "epoch": 0.6129538193435957,
+      "grad_norm": 0.4147947132587433,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 5276
+    },
+    {
+      "epoch": 0.6130699970955562,
+      "grad_norm": 0.4108210802078247,
+      "learning_rate": 0.0001,
+      "loss": 1.5585,
+      "step": 5277
+    },
+    {
+      "epoch": 0.6131861748475167,
+      "grad_norm": 0.40176138281822205,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 5278
+    },
+    {
+      "epoch": 0.6133023525994772,
+      "grad_norm": 0.44862088561058044,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 5279
+    },
+    {
+      "epoch": 0.6134185303514377,
+      "grad_norm": 0.4072551429271698,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 5280
+    },
+    {
+      "epoch": 0.6135347081033982,
+      "grad_norm": 0.47543540596961975,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 5281
+    },
+    {
+      "epoch": 0.6136508858553587,
+      "grad_norm": 0.4115746021270752,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 5282
+    },
+    {
+      "epoch": 0.6137670636073193,
+      "grad_norm": 0.42444461584091187,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 5283
+    },
+    {
+      "epoch": 0.6138832413592797,
+      "grad_norm": 0.4719880223274231,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 5284
+    },
+    {
+      "epoch": 0.6139994191112402,
+      "grad_norm": 0.4645368158817291,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 5285
+    },
+    {
+      "epoch": 0.6141155968632007,
+      "grad_norm": 0.4200696349143982,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 5286
+    },
+    {
+      "epoch": 0.6142317746151612,
+      "grad_norm": 0.439284086227417,
+      "learning_rate": 0.0001,
+      "loss": 1.5819,
+      "step": 5287
+    },
+    {
+      "epoch": 0.6143479523671217,
+      "grad_norm": 0.42954742908477783,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 5288
+    },
+    {
+      "epoch": 0.6144641301190822,
+      "grad_norm": 0.4379776120185852,
+      "learning_rate": 0.0001,
+      "loss": 1.7162,
+      "step": 5289
+    },
+    {
+      "epoch": 0.6145803078710427,
+      "grad_norm": 0.4193497896194458,
+      "learning_rate": 0.0001,
+      "loss": 1.6855,
+      "step": 5290
+    },
+    {
+      "epoch": 0.6146964856230032,
+      "grad_norm": 0.43286967277526855,
+      "learning_rate": 0.0001,
+      "loss": 1.7413,
+      "step": 5291
+    },
+    {
+      "epoch": 0.6148126633749637,
+      "grad_norm": 0.43035081028938293,
+      "learning_rate": 0.0001,
+      "loss": 1.5693,
+      "step": 5292
+    },
+    {
+      "epoch": 0.6149288411269241,
+      "grad_norm": 0.43507617712020874,
+      "learning_rate": 0.0001,
+      "loss": 1.8054,
+      "step": 5293
+    },
+    {
+      "epoch": 0.6150450188788847,
+      "grad_norm": 0.3959396779537201,
+      "learning_rate": 0.0001,
+      "loss": 1.4206,
+      "step": 5294
+    },
+    {
+      "epoch": 0.6151611966308452,
+      "grad_norm": 0.45225653052330017,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 5295
+    },
+    {
+      "epoch": 0.6152773743828057,
+      "grad_norm": 0.40767356753349304,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 5296
+    },
+    {
+      "epoch": 0.6153935521347662,
+      "grad_norm": 0.4382137954235077,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 5297
+    },
+    {
+      "epoch": 0.6155097298867267,
+      "grad_norm": 0.44092151522636414,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 5298
+    },
+    {
+      "epoch": 0.6156259076386872,
+      "grad_norm": 0.46326959133148193,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 5299
+    },
+    {
+      "epoch": 0.6157420853906477,
+      "grad_norm": 0.4669477343559265,
+      "learning_rate": 0.0001,
+      "loss": 1.7398,
+      "step": 5300
+    },
+    {
+      "epoch": 0.6158582631426082,
+      "grad_norm": 0.42557385563850403,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 5301
+    },
+    {
+      "epoch": 0.6159744408945687,
+      "grad_norm": 0.4492860436439514,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 5302
+    },
+    {
+      "epoch": 0.6160906186465291,
+      "grad_norm": 0.43879732489585876,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 5303
+    },
+    {
+      "epoch": 0.6162067963984897,
+      "grad_norm": 0.43372800946235657,
+      "learning_rate": 0.0001,
+      "loss": 1.6623,
+      "step": 5304
+    },
+    {
+      "epoch": 0.6163229741504502,
+      "grad_norm": 0.43956896662712097,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 5305
+    },
+    {
+      "epoch": 0.6164391519024107,
+      "grad_norm": 0.44320589303970337,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 5306
+    },
+    {
+      "epoch": 0.6165553296543712,
+      "grad_norm": 0.44610658288002014,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 5307
+    },
+    {
+      "epoch": 0.6166715074063317,
+      "grad_norm": 0.4208918511867523,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 5308
+    },
+    {
+      "epoch": 0.6167876851582922,
+      "grad_norm": 0.43007153272628784,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 5309
+    },
+    {
+      "epoch": 0.6169038629102527,
+      "grad_norm": 0.4086022973060608,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 5310
+    },
+    {
+      "epoch": 0.6170200406622132,
+      "grad_norm": 0.4730951189994812,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 5311
+    },
+    {
+      "epoch": 0.6171362184141737,
+      "grad_norm": 0.41164976358413696,
+      "learning_rate": 0.0001,
+      "loss": 1.3929,
+      "step": 5312
+    },
+    {
+      "epoch": 0.6172523961661341,
+      "grad_norm": 0.43444958329200745,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 5313
+    },
+    {
+      "epoch": 0.6173685739180946,
+      "grad_norm": 0.4134158790111542,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 5314
+    },
+    {
+      "epoch": 0.6174847516700552,
+      "grad_norm": 0.42741191387176514,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 5315
+    },
+    {
+      "epoch": 0.6176009294220157,
+      "grad_norm": 0.42607542872428894,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 5316
+    },
+    {
+      "epoch": 0.6177171071739762,
+      "grad_norm": 0.450953871011734,
+      "learning_rate": 0.0001,
+      "loss": 1.7549,
+      "step": 5317
+    },
+    {
+      "epoch": 0.6178332849259367,
+      "grad_norm": 0.4400138556957245,
+      "learning_rate": 0.0001,
+      "loss": 1.6905,
+      "step": 5318
+    },
+    {
+      "epoch": 0.6179494626778972,
+      "grad_norm": 0.4217263460159302,
+      "learning_rate": 0.0001,
+      "loss": 1.6168,
+      "step": 5319
+    },
+    {
+      "epoch": 0.6180656404298577,
+      "grad_norm": 0.4471561908721924,
+      "learning_rate": 0.0001,
+      "loss": 1.5789,
+      "step": 5320
+    },
+    {
+      "epoch": 0.6181818181818182,
+      "grad_norm": 0.45624133944511414,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 5321
+    },
+    {
+      "epoch": 0.6182979959337787,
+      "grad_norm": 0.4240606129169464,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 5322
+    },
+    {
+      "epoch": 0.6184141736857391,
+      "grad_norm": 0.41618114709854126,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 5323
+    },
+    {
+      "epoch": 0.6185303514376996,
+      "grad_norm": 0.4757783114910126,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 5324
+    },
+    {
+      "epoch": 0.6186465291896602,
+      "grad_norm": 0.45356112718582153,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 5325
+    },
+    {
+      "epoch": 0.6187627069416207,
+      "grad_norm": 0.42001453042030334,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 5326
+    },
+    {
+      "epoch": 0.6188788846935812,
+      "grad_norm": 0.4233144521713257,
+      "learning_rate": 0.0001,
+      "loss": 1.8047,
+      "step": 5327
+    },
+    {
+      "epoch": 0.6189950624455417,
+      "grad_norm": 0.451232373714447,
+      "learning_rate": 0.0001,
+      "loss": 1.7163,
+      "step": 5328
+    },
+    {
+      "epoch": 0.6191112401975022,
+      "grad_norm": 0.4210103750228882,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 5329
+    },
+    {
+      "epoch": 0.6192274179494627,
+      "grad_norm": 0.43561357259750366,
+      "learning_rate": 0.0001,
+      "loss": 1.5469,
+      "step": 5330
+    },
+    {
+      "epoch": 0.6193435957014232,
+      "grad_norm": 0.43155673146247864,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 5331
+    },
+    {
+      "epoch": 0.6194597734533837,
+      "grad_norm": 0.45339235663414,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 5332
+    },
+    {
+      "epoch": 0.6195759512053441,
+      "grad_norm": 0.5200868248939514,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 5333
+    },
+    {
+      "epoch": 0.6196921289573046,
+      "grad_norm": 0.45999056100845337,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 5334
+    },
+    {
+      "epoch": 0.6198083067092651,
+      "grad_norm": 0.46080732345581055,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 5335
+    },
+    {
+      "epoch": 0.6199244844612257,
+      "grad_norm": 0.4588463008403778,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 5336
+    },
+    {
+      "epoch": 0.6200406622131862,
+      "grad_norm": 0.4329404830932617,
+      "learning_rate": 0.0001,
+      "loss": 1.7793,
+      "step": 5337
+    },
+    {
+      "epoch": 0.6201568399651467,
+      "grad_norm": 0.42099177837371826,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 5338
+    },
+    {
+      "epoch": 0.6202730177171072,
+      "grad_norm": 0.4456626772880554,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 5339
+    },
+    {
+      "epoch": 0.6203891954690677,
+      "grad_norm": 0.45711061358451843,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 5340
+    },
+    {
+      "epoch": 0.6205053732210282,
+      "grad_norm": 0.4267137944698334,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 5341
+    },
+    {
+      "epoch": 0.6206215509729887,
+      "grad_norm": 0.438827782869339,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 5342
+    },
+    {
+      "epoch": 0.6207377287249491,
+      "grad_norm": 0.43408849835395813,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 5343
+    },
+    {
+      "epoch": 0.6208539064769096,
+      "grad_norm": 0.5134710669517517,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 5344
+    },
+    {
+      "epoch": 0.6209700842288701,
+      "grad_norm": 0.47032633423805237,
+      "learning_rate": 0.0001,
+      "loss": 1.6972,
+      "step": 5345
+    },
+    {
+      "epoch": 0.6210862619808307,
+      "grad_norm": 0.47240790724754333,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 5346
+    },
+    {
+      "epoch": 0.6212024397327912,
+      "grad_norm": 0.46300870180130005,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 5347
+    },
+    {
+      "epoch": 0.6213186174847517,
+      "grad_norm": 0.4472779333591461,
+      "learning_rate": 0.0001,
+      "loss": 1.5877,
+      "step": 5348
+    },
+    {
+      "epoch": 0.6214347952367122,
+      "grad_norm": 0.45526379346847534,
+      "learning_rate": 0.0001,
+      "loss": 1.676,
+      "step": 5349
+    },
+    {
+      "epoch": 0.6215509729886727,
+      "grad_norm": 0.4606863260269165,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 5350
+    },
+    {
+      "epoch": 0.6216671507406332,
+      "grad_norm": 0.4486534297466278,
+      "learning_rate": 0.0001,
+      "loss": 1.6654,
+      "step": 5351
+    },
+    {
+      "epoch": 0.6217833284925937,
+      "grad_norm": 0.5138721466064453,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 5352
+    },
+    {
+      "epoch": 0.6218995062445541,
+      "grad_norm": 0.43069109320640564,
+      "learning_rate": 0.0001,
+      "loss": 1.5228,
+      "step": 5353
+    },
+    {
+      "epoch": 0.6220156839965146,
+      "grad_norm": 0.4056498408317566,
+      "learning_rate": 0.0001,
+      "loss": 1.5213,
+      "step": 5354
+    },
+    {
+      "epoch": 0.6221318617484751,
+      "grad_norm": 0.4548470079898834,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 5355
+    },
+    {
+      "epoch": 0.6222480395004356,
+      "grad_norm": 0.4468959867954254,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 5356
+    },
+    {
+      "epoch": 0.6223642172523962,
+      "grad_norm": 0.4434147775173187,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 5357
+    },
+    {
+      "epoch": 0.6224803950043567,
+      "grad_norm": 0.4365995526313782,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 5358
+    },
+    {
+      "epoch": 0.6225965727563172,
+      "grad_norm": 0.42630013823509216,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 5359
+    },
+    {
+      "epoch": 0.6227127505082777,
+      "grad_norm": 0.4268414378166199,
+      "learning_rate": 0.0001,
+      "loss": 1.727,
+      "step": 5360
+    },
+    {
+      "epoch": 0.6228289282602382,
+      "grad_norm": 0.4438273012638092,
+      "learning_rate": 0.0001,
+      "loss": 1.7561,
+      "step": 5361
+    },
+    {
+      "epoch": 0.6229451060121987,
+      "grad_norm": 0.4021027088165283,
+      "learning_rate": 0.0001,
+      "loss": 1.3959,
+      "step": 5362
+    },
+    {
+      "epoch": 0.6230612837641591,
+      "grad_norm": 0.40190452337265015,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 5363
+    },
+    {
+      "epoch": 0.6231774615161196,
+      "grad_norm": 0.45445725321769714,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 5364
+    },
+    {
+      "epoch": 0.6232936392680801,
+      "grad_norm": 0.41883429884910583,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 5365
+    },
+    {
+      "epoch": 0.6234098170200406,
+      "grad_norm": 0.4443962574005127,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 5366
+    },
+    {
+      "epoch": 0.6235259947720012,
+      "grad_norm": 0.4436537027359009,
+      "learning_rate": 0.0001,
+      "loss": 1.7087,
+      "step": 5367
+    },
+    {
+      "epoch": 0.6236421725239617,
+      "grad_norm": 0.4414960741996765,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 5368
+    },
+    {
+      "epoch": 0.6237583502759222,
+      "grad_norm": 0.43961644172668457,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 5369
+    },
+    {
+      "epoch": 0.6238745280278827,
+      "grad_norm": 0.42468035221099854,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 5370
+    },
+    {
+      "epoch": 0.6239907057798432,
+      "grad_norm": 0.40939033031463623,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 5371
+    },
+    {
+      "epoch": 0.6241068835318037,
+      "grad_norm": 0.43494606018066406,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 5372
+    },
+    {
+      "epoch": 0.6242230612837641,
+      "grad_norm": 0.4493979513645172,
+      "learning_rate": 0.0001,
+      "loss": 1.7368,
+      "step": 5373
+    },
+    {
+      "epoch": 0.6243392390357246,
+      "grad_norm": 0.3963182270526886,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 5374
+    },
+    {
+      "epoch": 0.6244554167876851,
+      "grad_norm": 0.4189557731151581,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 5375
+    },
+    {
+      "epoch": 0.6245715945396456,
+      "grad_norm": 0.449720174074173,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 5376
+    },
+    {
+      "epoch": 0.6246877722916061,
+      "grad_norm": 0.4135220944881439,
+      "learning_rate": 0.0001,
+      "loss": 1.6108,
+      "step": 5377
+    },
+    {
+      "epoch": 0.6248039500435667,
+      "grad_norm": 0.47096794843673706,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 5378
+    },
+    {
+      "epoch": 0.6249201277955272,
+      "grad_norm": 0.4510813057422638,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 5379
+    },
+    {
+      "epoch": 0.6250363055474877,
+      "grad_norm": 0.43997013568878174,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 5380
+    },
+    {
+      "epoch": 0.6251524832994482,
+      "grad_norm": 0.4731866419315338,
+      "learning_rate": 0.0001,
+      "loss": 1.7788,
+      "step": 5381
+    },
+    {
+      "epoch": 0.6252686610514087,
+      "grad_norm": 0.44134780764579773,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 5382
+    },
+    {
+      "epoch": 0.6253848388033691,
+      "grad_norm": 0.5102000832557678,
+      "learning_rate": 0.0001,
+      "loss": 1.7645,
+      "step": 5383
+    },
+    {
+      "epoch": 0.6255010165553296,
+      "grad_norm": 0.43674057722091675,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 5384
+    },
+    {
+      "epoch": 0.6256171943072901,
+      "grad_norm": 0.47883304953575134,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 5385
+    },
+    {
+      "epoch": 0.6257333720592506,
+      "grad_norm": 0.43253079056739807,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 5386
+    },
+    {
+      "epoch": 0.6258495498112111,
+      "grad_norm": 0.41759157180786133,
+      "learning_rate": 0.0001,
+      "loss": 1.6168,
+      "step": 5387
+    },
+    {
+      "epoch": 0.6259657275631717,
+      "grad_norm": 0.421507865190506,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 5388
+    },
+    {
+      "epoch": 0.6260819053151322,
+      "grad_norm": 0.4465677738189697,
+      "learning_rate": 0.0001,
+      "loss": 1.8354,
+      "step": 5389
+    },
+    {
+      "epoch": 0.6261980830670927,
+      "grad_norm": 0.4306974709033966,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 5390
+    },
+    {
+      "epoch": 0.6263142608190532,
+      "grad_norm": 0.4195205271244049,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 5391
+    },
+    {
+      "epoch": 0.6264304385710137,
+      "grad_norm": 0.44183290004730225,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 5392
+    },
+    {
+      "epoch": 0.6265466163229741,
+      "grad_norm": 0.44616198539733887,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 5393
+    },
+    {
+      "epoch": 0.6266627940749346,
+      "grad_norm": 0.44162172079086304,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 5394
+    },
+    {
+      "epoch": 0.6267789718268951,
+      "grad_norm": 0.46100786328315735,
+      "learning_rate": 0.0001,
+      "loss": 1.8678,
+      "step": 5395
+    },
+    {
+      "epoch": 0.6268951495788556,
+      "grad_norm": 0.4163057208061218,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 5396
+    },
+    {
+      "epoch": 0.6270113273308161,
+      "grad_norm": 0.4567055404186249,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 5397
+    },
+    {
+      "epoch": 0.6271275050827766,
+      "grad_norm": 0.4351733326911926,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 5398
+    },
+    {
+      "epoch": 0.6272436828347372,
+      "grad_norm": 0.4696790277957916,
+      "learning_rate": 0.0001,
+      "loss": 1.773,
+      "step": 5399
+    },
+    {
+      "epoch": 0.6273598605866977,
+      "grad_norm": 0.4156184792518616,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 5400
+    },
+    {
+      "epoch": 0.6274760383386582,
+      "grad_norm": 0.450946182012558,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 5401
+    },
+    {
+      "epoch": 0.6275922160906187,
+      "grad_norm": 0.48097074031829834,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 5402
+    },
+    {
+      "epoch": 0.6277083938425791,
+      "grad_norm": 0.4515106976032257,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 5403
+    },
+    {
+      "epoch": 0.6278245715945396,
+      "grad_norm": 0.4210405647754669,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 5404
+    },
+    {
+      "epoch": 0.6279407493465001,
+      "grad_norm": 0.42682531476020813,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 5405
+    },
+    {
+      "epoch": 0.6280569270984606,
+      "grad_norm": 0.4539082944393158,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 5406
+    },
+    {
+      "epoch": 0.6281731048504211,
+      "grad_norm": 0.43941760063171387,
+      "learning_rate": 0.0001,
+      "loss": 1.5897,
+      "step": 5407
+    },
+    {
+      "epoch": 0.6282892826023816,
+      "grad_norm": 0.4232398271560669,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 5408
+    },
+    {
+      "epoch": 0.6284054603543422,
+      "grad_norm": 0.4403201639652252,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 5409
+    },
+    {
+      "epoch": 0.6285216381063027,
+      "grad_norm": 0.45847928524017334,
+      "learning_rate": 0.0001,
+      "loss": 1.7491,
+      "step": 5410
+    },
+    {
+      "epoch": 0.6286378158582632,
+      "grad_norm": 0.4543238878250122,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 5411
+    },
+    {
+      "epoch": 0.6287539936102237,
+      "grad_norm": 0.4506990611553192,
+      "learning_rate": 0.0001,
+      "loss": 1.7011,
+      "step": 5412
+    },
+    {
+      "epoch": 0.6288701713621841,
+      "grad_norm": 0.38618630170822144,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 5413
+    },
+    {
+      "epoch": 0.6289863491141446,
+      "grad_norm": 0.48193633556365967,
+      "learning_rate": 0.0001,
+      "loss": 1.8388,
+      "step": 5414
+    },
+    {
+      "epoch": 0.6291025268661051,
+      "grad_norm": 0.45501387119293213,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 5415
+    },
+    {
+      "epoch": 0.6292187046180656,
+      "grad_norm": 0.43010449409484863,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 5416
+    },
+    {
+      "epoch": 0.6293348823700261,
+      "grad_norm": 0.42179417610168457,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 5417
+    },
+    {
+      "epoch": 0.6294510601219866,
+      "grad_norm": 0.4156195819377899,
+      "learning_rate": 0.0001,
+      "loss": 1.3938,
+      "step": 5418
+    },
+    {
+      "epoch": 0.6295672378739472,
+      "grad_norm": 0.4323188364505768,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 5419
+    },
+    {
+      "epoch": 0.6296834156259077,
+      "grad_norm": 0.4095346927642822,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 5420
+    },
+    {
+      "epoch": 0.6297995933778682,
+      "grad_norm": 0.4441397190093994,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 5421
+    },
+    {
+      "epoch": 0.6299157711298287,
+      "grad_norm": 0.4608078896999359,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 5422
+    },
+    {
+      "epoch": 0.6300319488817892,
+      "grad_norm": 0.4394521415233612,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 5423
+    },
+    {
+      "epoch": 0.6301481266337496,
+      "grad_norm": 0.4374762773513794,
+      "learning_rate": 0.0001,
+      "loss": 1.6325,
+      "step": 5424
+    },
+    {
+      "epoch": 0.6302643043857101,
+      "grad_norm": 0.49020811915397644,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 5425
+    },
+    {
+      "epoch": 0.6303804821376706,
+      "grad_norm": 0.45695215463638306,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 5426
+    },
+    {
+      "epoch": 0.6304966598896311,
+      "grad_norm": 0.4443933367729187,
+      "learning_rate": 0.0001,
+      "loss": 1.6863,
+      "step": 5427
+    },
+    {
+      "epoch": 0.6306128376415916,
+      "grad_norm": 0.47112321853637695,
+      "learning_rate": 0.0001,
+      "loss": 1.8038,
+      "step": 5428
+    },
+    {
+      "epoch": 0.6307290153935521,
+      "grad_norm": 0.43735185265541077,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 5429
+    },
+    {
+      "epoch": 0.6308451931455127,
+      "grad_norm": 0.42006251215934753,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 5430
+    },
+    {
+      "epoch": 0.6309613708974732,
+      "grad_norm": 0.4126074016094208,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 5431
+    },
+    {
+      "epoch": 0.6310775486494337,
+      "grad_norm": 0.4392840564250946,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 5432
+    },
+    {
+      "epoch": 0.6311937264013942,
+      "grad_norm": 0.4268054962158203,
+      "learning_rate": 0.0001,
+      "loss": 1.6143,
+      "step": 5433
+    },
+    {
+      "epoch": 0.6313099041533546,
+      "grad_norm": 0.4118890166282654,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 5434
+    },
+    {
+      "epoch": 0.6314260819053151,
+      "grad_norm": 0.4566260874271393,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6315422596572756,
+      "grad_norm": 0.4307934045791626,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 5436
+    },
+    {
+      "epoch": 0.6316584374092361,
+      "grad_norm": 0.4302327632904053,
+      "learning_rate": 0.0001,
+      "loss": 1.7334,
+      "step": 5437
+    },
+    {
+      "epoch": 0.6317746151611966,
+      "grad_norm": 0.4081536531448364,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 5438
+    },
+    {
+      "epoch": 0.6318907929131571,
+      "grad_norm": 0.43301841616630554,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 5439
+    },
+    {
+      "epoch": 0.6320069706651177,
+      "grad_norm": 0.4393984377384186,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6321231484170782,
+      "grad_norm": 0.4503820240497589,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 5441
+    },
+    {
+      "epoch": 0.6322393261690387,
+      "grad_norm": 0.4837632179260254,
+      "learning_rate": 0.0001,
+      "loss": 1.7753,
+      "step": 5442
+    },
+    {
+      "epoch": 0.6323555039209992,
+      "grad_norm": 0.4605303108692169,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 5443
+    },
+    {
+      "epoch": 0.6324716816729596,
+      "grad_norm": 0.4571518003940582,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 5444
+    },
+    {
+      "epoch": 0.6325878594249201,
+      "grad_norm": 0.4223042130470276,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6327040371768806,
+      "grad_norm": 0.4364268183708191,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 5446
+    },
+    {
+      "epoch": 0.6328202149288411,
+      "grad_norm": 0.43388858437538147,
+      "learning_rate": 0.0001,
+      "loss": 1.6069,
+      "step": 5447
+    },
+    {
+      "epoch": 0.6329363926808016,
+      "grad_norm": 0.4458199143409729,
+      "learning_rate": 0.0001,
+      "loss": 1.7514,
+      "step": 5448
+    },
+    {
+      "epoch": 0.6330525704327621,
+      "grad_norm": 0.43075671792030334,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 5449
+    },
+    {
+      "epoch": 0.6331687481847226,
+      "grad_norm": 0.42168936133384705,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6332849259366832,
+      "grad_norm": 0.41019555926322937,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 5451
+    },
+    {
+      "epoch": 0.6334011036886437,
+      "grad_norm": 0.4191059470176697,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 5452
+    },
+    {
+      "epoch": 0.6335172814406042,
+      "grad_norm": 0.4607682526111603,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 5453
+    },
+    {
+      "epoch": 0.6336334591925646,
+      "grad_norm": 0.4602360725402832,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 5454
+    },
+    {
+      "epoch": 0.6337496369445251,
+      "grad_norm": 0.4206385910511017,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 5455
+    },
+    {
+      "epoch": 0.6338658146964856,
+      "grad_norm": 0.427676796913147,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 5456
+    },
+    {
+      "epoch": 0.6339819924484461,
+      "grad_norm": 0.47075173258781433,
+      "learning_rate": 0.0001,
+      "loss": 1.7829,
+      "step": 5457
+    },
+    {
+      "epoch": 0.6340981702004066,
+      "grad_norm": 0.45573490858078003,
+      "learning_rate": 0.0001,
+      "loss": 1.6889,
+      "step": 5458
+    },
+    {
+      "epoch": 0.6342143479523671,
+      "grad_norm": 0.43033167719841003,
+      "learning_rate": 0.0001,
+      "loss": 1.8298,
+      "step": 5459
+    },
+    {
+      "epoch": 0.6343305257043276,
+      "grad_norm": 0.429253488779068,
+      "learning_rate": 0.0001,
+      "loss": 1.5989,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6344467034562882,
+      "grad_norm": 0.4369482696056366,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 5461
+    },
+    {
+      "epoch": 0.6345628812082487,
+      "grad_norm": 0.4227111041545868,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 5462
+    },
+    {
+      "epoch": 0.6346790589602092,
+      "grad_norm": 0.42353329062461853,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 5463
+    },
+    {
+      "epoch": 0.6347952367121696,
+      "grad_norm": 0.44350820779800415,
+      "learning_rate": 0.0001,
+      "loss": 1.6899,
+      "step": 5464
+    },
+    {
+      "epoch": 0.6349114144641301,
+      "grad_norm": 0.42905986309051514,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 5465
+    },
+    {
+      "epoch": 0.6350275922160906,
+      "grad_norm": 0.4360484480857849,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 5466
+    },
+    {
+      "epoch": 0.6351437699680511,
+      "grad_norm": 0.4489285349845886,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 5467
+    },
+    {
+      "epoch": 0.6352599477200116,
+      "grad_norm": 0.43288835883140564,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 5468
+    },
+    {
+      "epoch": 0.6353761254719721,
+      "grad_norm": 0.4269554316997528,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 5469
+    },
+    {
+      "epoch": 0.6354923032239326,
+      "grad_norm": 0.43894824385643005,
+      "learning_rate": 0.0001,
+      "loss": 1.6423,
+      "step": 5470
+    },
+    {
+      "epoch": 0.6356084809758931,
+      "grad_norm": 0.42554864287376404,
+      "learning_rate": 0.0001,
+      "loss": 1.382,
+      "step": 5471
+    },
+    {
+      "epoch": 0.6357246587278537,
+      "grad_norm": 0.43490105867385864,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 5472
+    },
+    {
+      "epoch": 0.6358408364798142,
+      "grad_norm": 0.41975313425064087,
+      "learning_rate": 0.0001,
+      "loss": 1.5971,
+      "step": 5473
+    },
+    {
+      "epoch": 0.6359570142317746,
+      "grad_norm": 0.43810316920280457,
+      "learning_rate": 0.0001,
+      "loss": 1.5934,
+      "step": 5474
+    },
+    {
+      "epoch": 0.6360731919837351,
+      "grad_norm": 0.4118810296058655,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 5475
+    },
+    {
+      "epoch": 0.6361893697356956,
+      "grad_norm": 0.4668201804161072,
+      "learning_rate": 0.0001,
+      "loss": 1.7682,
+      "step": 5476
+    },
+    {
+      "epoch": 0.6363055474876561,
+      "grad_norm": 0.43905314803123474,
+      "learning_rate": 0.0001,
+      "loss": 1.6805,
+      "step": 5477
+    },
+    {
+      "epoch": 0.6364217252396166,
+      "grad_norm": 0.43517014384269714,
+      "learning_rate": 0.0001,
+      "loss": 1.6264,
+      "step": 5478
+    },
+    {
+      "epoch": 0.6365379029915771,
+      "grad_norm": 0.4257369339466095,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 5479
+    },
+    {
+      "epoch": 0.6366540807435376,
+      "grad_norm": 0.4415474534034729,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 5480
+    },
+    {
+      "epoch": 0.6367702584954981,
+      "grad_norm": 0.48000675439834595,
+      "learning_rate": 0.0001,
+      "loss": 1.8075,
+      "step": 5481
+    },
+    {
+      "epoch": 0.6368864362474587,
+      "grad_norm": 0.433585524559021,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 5482
+    },
+    {
+      "epoch": 0.6370026139994192,
+      "grad_norm": 0.46495142579078674,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 5483
+    },
+    {
+      "epoch": 0.6371187917513796,
+      "grad_norm": 0.4620204567909241,
+      "learning_rate": 0.0001,
+      "loss": 1.7422,
+      "step": 5484
+    },
+    {
+      "epoch": 0.6372349695033401,
+      "grad_norm": 0.457280695438385,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 5485
+    },
+    {
+      "epoch": 0.6373511472553006,
+      "grad_norm": 0.40286630392074585,
+      "learning_rate": 0.0001,
+      "loss": 1.5871,
+      "step": 5486
+    },
+    {
+      "epoch": 0.6374673250072611,
+      "grad_norm": 0.41824090480804443,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 5487
+    },
+    {
+      "epoch": 0.6375835027592216,
+      "grad_norm": 0.4374508261680603,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 5488
+    },
+    {
+      "epoch": 0.6376996805111821,
+      "grad_norm": 0.44496670365333557,
+      "learning_rate": 0.0001,
+      "loss": 1.6769,
+      "step": 5489
+    },
+    {
+      "epoch": 0.6378158582631426,
+      "grad_norm": 0.49528950452804565,
+      "learning_rate": 0.0001,
+      "loss": 1.9251,
+      "step": 5490
+    },
+    {
+      "epoch": 0.6379320360151031,
+      "grad_norm": 0.4341500401496887,
+      "learning_rate": 0.0001,
+      "loss": 1.6974,
+      "step": 5491
+    },
+    {
+      "epoch": 0.6380482137670636,
+      "grad_norm": 0.4738743305206299,
+      "learning_rate": 0.0001,
+      "loss": 1.8318,
+      "step": 5492
+    },
+    {
+      "epoch": 0.6381643915190242,
+      "grad_norm": 0.4284060001373291,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 5493
+    },
+    {
+      "epoch": 0.6382805692709846,
+      "grad_norm": 0.4526841938495636,
+      "learning_rate": 0.0001,
+      "loss": 1.7989,
+      "step": 5494
+    },
+    {
+      "epoch": 0.6383967470229451,
+      "grad_norm": 0.47774431109428406,
+      "learning_rate": 0.0001,
+      "loss": 1.8372,
+      "step": 5495
+    },
+    {
+      "epoch": 0.6385129247749056,
+      "grad_norm": 0.41493383049964905,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 5496
+    },
+    {
+      "epoch": 0.6386291025268661,
+      "grad_norm": 0.3971651494503021,
+      "learning_rate": 0.0001,
+      "loss": 1.3297,
+      "step": 5497
+    },
+    {
+      "epoch": 0.6387452802788266,
+      "grad_norm": 0.4250967502593994,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 5498
+    },
+    {
+      "epoch": 0.6388614580307871,
+      "grad_norm": 0.4181077778339386,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 5499
+    },
+    {
+      "epoch": 0.6389776357827476,
+      "grad_norm": 0.45691534876823425,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 5500
+    },
+    {
+      "epoch": 0.6390938135347081,
+      "grad_norm": 0.43789535760879517,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 5501
+    },
+    {
+      "epoch": 0.6392099912866686,
+      "grad_norm": 0.42184701561927795,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 5502
+    },
+    {
+      "epoch": 0.6393261690386292,
+      "grad_norm": 0.4412265717983246,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 5503
+    },
+    {
+      "epoch": 0.6394423467905896,
+      "grad_norm": 0.4428151845932007,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 5504
+    },
+    {
+      "epoch": 0.6395585245425501,
+      "grad_norm": 0.4467543959617615,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 5505
+    },
+    {
+      "epoch": 0.6396747022945106,
+      "grad_norm": 0.40767189860343933,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 5506
+    },
+    {
+      "epoch": 0.6397908800464711,
+      "grad_norm": 0.4650069773197174,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 5507
+    },
+    {
+      "epoch": 0.6399070577984316,
+      "grad_norm": 0.4141228199005127,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 5508
+    },
+    {
+      "epoch": 0.6400232355503921,
+      "grad_norm": 0.4306548833847046,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 5509
+    },
+    {
+      "epoch": 0.6401394133023526,
+      "grad_norm": 0.45290401577949524,
+      "learning_rate": 0.0001,
+      "loss": 1.6879,
+      "step": 5510
+    },
+    {
+      "epoch": 0.6402555910543131,
+      "grad_norm": 0.43989840149879456,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 5511
+    },
+    {
+      "epoch": 0.6403717688062736,
+      "grad_norm": 0.46134138107299805,
+      "learning_rate": 0.0001,
+      "loss": 1.6924,
+      "step": 5512
+    },
+    {
+      "epoch": 0.640487946558234,
+      "grad_norm": 0.4793362617492676,
+      "learning_rate": 0.0001,
+      "loss": 1.7376,
+      "step": 5513
+    },
+    {
+      "epoch": 0.6406041243101946,
+      "grad_norm": 0.4687068462371826,
+      "learning_rate": 0.0001,
+      "loss": 1.3945,
+      "step": 5514
+    },
+    {
+      "epoch": 0.6407203020621551,
+      "grad_norm": 0.4695538878440857,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 5515
+    },
+    {
+      "epoch": 0.6408364798141156,
+      "grad_norm": 0.4554864168167114,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 5516
+    },
+    {
+      "epoch": 0.6409526575660761,
+      "grad_norm": 0.45408567786216736,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 5517
+    },
+    {
+      "epoch": 0.6410688353180366,
+      "grad_norm": 0.4672776162624359,
+      "learning_rate": 0.0001,
+      "loss": 1.7557,
+      "step": 5518
+    },
+    {
+      "epoch": 0.6411850130699971,
+      "grad_norm": 0.43636855483055115,
+      "learning_rate": 0.0001,
+      "loss": 1.6705,
+      "step": 5519
+    },
+    {
+      "epoch": 0.6413011908219576,
+      "grad_norm": 0.4704853892326355,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 5520
+    },
+    {
+      "epoch": 0.6414173685739181,
+      "grad_norm": 0.45459243655204773,
+      "learning_rate": 0.0001,
+      "loss": 1.5321,
+      "step": 5521
+    },
+    {
+      "epoch": 0.6415335463258786,
+      "grad_norm": 0.4612571895122528,
+      "learning_rate": 0.0001,
+      "loss": 1.6599,
+      "step": 5522
+    },
+    {
+      "epoch": 0.641649724077839,
+      "grad_norm": 0.44641435146331787,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 5523
+    },
+    {
+      "epoch": 0.6417659018297996,
+      "grad_norm": 0.4449585974216461,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 5524
+    },
+    {
+      "epoch": 0.6418820795817601,
+      "grad_norm": 0.49576014280319214,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 5525
+    },
+    {
+      "epoch": 0.6419982573337206,
+      "grad_norm": 0.4535317122936249,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 5526
+    },
+    {
+      "epoch": 0.6421144350856811,
+      "grad_norm": 0.43689054250717163,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 5527
+    },
+    {
+      "epoch": 0.6422306128376416,
+      "grad_norm": 0.4465075135231018,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 5528
+    },
+    {
+      "epoch": 0.6423467905896021,
+      "grad_norm": 0.4437648355960846,
+      "learning_rate": 0.0001,
+      "loss": 1.6005,
+      "step": 5529
+    },
+    {
+      "epoch": 0.6424629683415626,
+      "grad_norm": 0.43459346890449524,
+      "learning_rate": 0.0001,
+      "loss": 1.5339,
+      "step": 5530
+    },
+    {
+      "epoch": 0.6425791460935231,
+      "grad_norm": 0.46662724018096924,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 5531
+    },
+    {
+      "epoch": 0.6426953238454836,
+      "grad_norm": 0.4122602939605713,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 5532
+    },
+    {
+      "epoch": 0.642811501597444,
+      "grad_norm": 0.4282759726047516,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 5533
+    },
+    {
+      "epoch": 0.6429276793494045,
+      "grad_norm": 0.4355936348438263,
+      "learning_rate": 0.0001,
+      "loss": 1.6888,
+      "step": 5534
+    },
+    {
+      "epoch": 0.6430438571013651,
+      "grad_norm": 0.4413192570209503,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 5535
+    },
+    {
+      "epoch": 0.6431600348533256,
+      "grad_norm": 0.4456358253955841,
+      "learning_rate": 0.0001,
+      "loss": 1.5277,
+      "step": 5536
+    },
+    {
+      "epoch": 0.6432762126052861,
+      "grad_norm": 0.4487074017524719,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 5537
+    },
+    {
+      "epoch": 0.6433923903572466,
+      "grad_norm": 0.4546639323234558,
+      "learning_rate": 0.0001,
+      "loss": 1.7977,
+      "step": 5538
+    },
+    {
+      "epoch": 0.6435085681092071,
+      "grad_norm": 0.47125598788261414,
+      "learning_rate": 0.0001,
+      "loss": 1.6502,
+      "step": 5539
+    },
+    {
+      "epoch": 0.6436247458611676,
+      "grad_norm": 0.4681374132633209,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 5540
+    },
+    {
+      "epoch": 0.6437409236131281,
+      "grad_norm": 0.41679567098617554,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 5541
+    },
+    {
+      "epoch": 0.6438571013650886,
+      "grad_norm": 0.4643932282924652,
+      "learning_rate": 0.0001,
+      "loss": 1.4515,
+      "step": 5542
+    },
+    {
+      "epoch": 0.643973279117049,
+      "grad_norm": 0.42899298667907715,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 5543
+    },
+    {
+      "epoch": 0.6440894568690095,
+      "grad_norm": 0.44600605964660645,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 5544
+    },
+    {
+      "epoch": 0.6442056346209701,
+      "grad_norm": 0.4713478088378906,
+      "learning_rate": 0.0001,
+      "loss": 1.6812,
+      "step": 5545
+    },
+    {
+      "epoch": 0.6443218123729306,
+      "grad_norm": 0.48742958903312683,
+      "learning_rate": 0.0001,
+      "loss": 1.5977,
+      "step": 5546
+    },
+    {
+      "epoch": 0.6444379901248911,
+      "grad_norm": 0.4531098008155823,
+      "learning_rate": 0.0001,
+      "loss": 1.7098,
+      "step": 5547
+    },
+    {
+      "epoch": 0.6445541678768516,
+      "grad_norm": 0.4089454710483551,
+      "learning_rate": 0.0001,
+      "loss": 1.4264,
+      "step": 5548
+    },
+    {
+      "epoch": 0.6446703456288121,
+      "grad_norm": 0.4439248740673065,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 5549
+    },
+    {
+      "epoch": 0.6447865233807726,
+      "grad_norm": 0.4419444799423218,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 5550
+    },
+    {
+      "epoch": 0.6449027011327331,
+      "grad_norm": 0.43608152866363525,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 5551
+    },
+    {
+      "epoch": 0.6450188788846936,
+      "grad_norm": 0.4767807722091675,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 5552
+    },
+    {
+      "epoch": 0.645135056636654,
+      "grad_norm": 0.4356260597705841,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 5553
+    },
+    {
+      "epoch": 0.6452512343886145,
+      "grad_norm": 0.42101210355758667,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 5554
+    },
+    {
+      "epoch": 0.645367412140575,
+      "grad_norm": 0.4493582248687744,
+      "learning_rate": 0.0001,
+      "loss": 1.3819,
+      "step": 5555
+    },
+    {
+      "epoch": 0.6454835898925356,
+      "grad_norm": 0.4049502909183502,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 5556
+    },
+    {
+      "epoch": 0.6455997676444961,
+      "grad_norm": 0.42648744583129883,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 5557
+    },
+    {
+      "epoch": 0.6457159453964566,
+      "grad_norm": 0.45338109135627747,
+      "learning_rate": 0.0001,
+      "loss": 1.7413,
+      "step": 5558
+    },
+    {
+      "epoch": 0.6458321231484171,
+      "grad_norm": 0.4680498540401459,
+      "learning_rate": 0.0001,
+      "loss": 1.7397,
+      "step": 5559
+    },
+    {
+      "epoch": 0.6459483009003776,
+      "grad_norm": 0.48937198519706726,
+      "learning_rate": 0.0001,
+      "loss": 1.7993,
+      "step": 5560
+    },
+    {
+      "epoch": 0.6460644786523381,
+      "grad_norm": 0.41471385955810547,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 5561
+    },
+    {
+      "epoch": 0.6461806564042986,
+      "grad_norm": 0.4447079002857208,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 5562
+    },
+    {
+      "epoch": 0.646296834156259,
+      "grad_norm": 0.42607253789901733,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 5563
+    },
+    {
+      "epoch": 0.6464130119082195,
+      "grad_norm": 0.4518924653530121,
+      "learning_rate": 0.0001,
+      "loss": 1.7441,
+      "step": 5564
+    },
+    {
+      "epoch": 0.64652918966018,
+      "grad_norm": 0.4289340376853943,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 5565
+    },
+    {
+      "epoch": 0.6466453674121406,
+      "grad_norm": 0.45490723848342896,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 5566
+    },
+    {
+      "epoch": 0.6467615451641011,
+      "grad_norm": 0.4541492462158203,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 5567
+    },
+    {
+      "epoch": 0.6468777229160616,
+      "grad_norm": 0.42619723081588745,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 5568
+    },
+    {
+      "epoch": 0.6469939006680221,
+      "grad_norm": 0.43783965706825256,
+      "learning_rate": 0.0001,
+      "loss": 1.7282,
+      "step": 5569
+    },
+    {
+      "epoch": 0.6471100784199826,
+      "grad_norm": 0.4257659912109375,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 5570
+    },
+    {
+      "epoch": 0.6472262561719431,
+      "grad_norm": 0.4469713270664215,
+      "learning_rate": 0.0001,
+      "loss": 1.6628,
+      "step": 5571
+    },
+    {
+      "epoch": 0.6473424339239036,
+      "grad_norm": 0.4462829530239105,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 5572
+    },
+    {
+      "epoch": 0.647458611675864,
+      "grad_norm": 0.4520929455757141,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 5573
+    },
+    {
+      "epoch": 0.6475747894278245,
+      "grad_norm": 0.47630277276039124,
+      "learning_rate": 0.0001,
+      "loss": 1.6382,
+      "step": 5574
+    },
+    {
+      "epoch": 0.647690967179785,
+      "grad_norm": 0.435032457113266,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 5575
+    },
+    {
+      "epoch": 0.6478071449317455,
+      "grad_norm": 0.46478384733200073,
+      "learning_rate": 0.0001,
+      "loss": 1.6891,
+      "step": 5576
+    },
+    {
+      "epoch": 0.6479233226837061,
+      "grad_norm": 0.4479098618030548,
+      "learning_rate": 0.0001,
+      "loss": 1.6942,
+      "step": 5577
+    },
+    {
+      "epoch": 0.6480395004356666,
+      "grad_norm": 0.4304809868335724,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 5578
+    },
+    {
+      "epoch": 0.6481556781876271,
+      "grad_norm": 0.43313372135162354,
+      "learning_rate": 0.0001,
+      "loss": 1.6466,
+      "step": 5579
+    },
+    {
+      "epoch": 0.6482718559395876,
+      "grad_norm": 0.4252367615699768,
+      "learning_rate": 0.0001,
+      "loss": 1.6563,
+      "step": 5580
+    },
+    {
+      "epoch": 0.6483880336915481,
+      "grad_norm": 0.42418304085731506,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 5581
+    },
+    {
+      "epoch": 0.6485042114435086,
+      "grad_norm": 0.45233771204948425,
+      "learning_rate": 0.0001,
+      "loss": 1.779,
+      "step": 5582
+    },
+    {
+      "epoch": 0.648620389195469,
+      "grad_norm": 0.43534642457962036,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 5583
+    },
+    {
+      "epoch": 0.6487365669474295,
+      "grad_norm": 0.40678098797798157,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 5584
+    },
+    {
+      "epoch": 0.64885274469939,
+      "grad_norm": 0.4426053762435913,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 5585
+    },
+    {
+      "epoch": 0.6489689224513505,
+      "grad_norm": 0.43607258796691895,
+      "learning_rate": 0.0001,
+      "loss": 1.6646,
+      "step": 5586
+    },
+    {
+      "epoch": 0.6490851002033111,
+      "grad_norm": 0.43579816818237305,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 5587
+    },
+    {
+      "epoch": 0.6492012779552716,
+      "grad_norm": 0.43486765027046204,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 5588
+    },
+    {
+      "epoch": 0.6493174557072321,
+      "grad_norm": 0.42424148321151733,
+      "learning_rate": 0.0001,
+      "loss": 1.5596,
+      "step": 5589
+    },
+    {
+      "epoch": 0.6494336334591926,
+      "grad_norm": 0.42990225553512573,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 5590
+    },
+    {
+      "epoch": 0.6495498112111531,
+      "grad_norm": 0.43480220437049866,
+      "learning_rate": 0.0001,
+      "loss": 1.7013,
+      "step": 5591
+    },
+    {
+      "epoch": 0.6496659889631136,
+      "grad_norm": 0.4241524040699005,
+      "learning_rate": 0.0001,
+      "loss": 1.6738,
+      "step": 5592
+    },
+    {
+      "epoch": 0.649782166715074,
+      "grad_norm": 0.4739820957183838,
+      "learning_rate": 0.0001,
+      "loss": 1.7373,
+      "step": 5593
+    },
+    {
+      "epoch": 0.6498983444670345,
+      "grad_norm": 0.44619935750961304,
+      "learning_rate": 0.0001,
+      "loss": 1.5907,
+      "step": 5594
+    },
+    {
+      "epoch": 0.650014522218995,
+      "grad_norm": 0.444965660572052,
+      "learning_rate": 0.0001,
+      "loss": 1.721,
+      "step": 5595
+    },
+    {
+      "epoch": 0.6501306999709555,
+      "grad_norm": 0.4169558882713318,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 5596
+    },
+    {
+      "epoch": 0.6502468777229161,
+      "grad_norm": 0.44122716784477234,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 5597
+    },
+    {
+      "epoch": 0.6503630554748766,
+      "grad_norm": 0.4570583403110504,
+      "learning_rate": 0.0001,
+      "loss": 1.7827,
+      "step": 5598
+    },
+    {
+      "epoch": 0.6504792332268371,
+      "grad_norm": 0.4605511426925659,
+      "learning_rate": 0.0001,
+      "loss": 1.6557,
+      "step": 5599
+    },
+    {
+      "epoch": 0.6505954109787976,
+      "grad_norm": 0.44260162115097046,
+      "learning_rate": 0.0001,
+      "loss": 1.6296,
+      "step": 5600
+    },
+    {
+      "epoch": 0.6507115887307581,
+      "grad_norm": 0.4196053445339203,
+      "learning_rate": 0.0001,
+      "loss": 1.4482,
+      "step": 5601
+    },
+    {
+      "epoch": 0.6508277664827186,
+      "grad_norm": 0.44127750396728516,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 5602
+    },
+    {
+      "epoch": 0.650943944234679,
+      "grad_norm": 0.3989742398262024,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 5603
+    },
+    {
+      "epoch": 0.6510601219866395,
+      "grad_norm": 0.4253181517124176,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 5604
+    },
+    {
+      "epoch": 0.6511762997386,
+      "grad_norm": 0.43197429180145264,
+      "learning_rate": 0.0001,
+      "loss": 1.5778,
+      "step": 5605
+    },
+    {
+      "epoch": 0.6512924774905605,
+      "grad_norm": 0.41109520196914673,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 5606
+    },
+    {
+      "epoch": 0.651408655242521,
+      "grad_norm": 0.41764628887176514,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 5607
+    },
+    {
+      "epoch": 0.6515248329944816,
+      "grad_norm": 0.4168471395969391,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 5608
+    },
+    {
+      "epoch": 0.6516410107464421,
+      "grad_norm": 0.47764724493026733,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 5609
+    },
+    {
+      "epoch": 0.6517571884984026,
+      "grad_norm": 0.432256281375885,
+      "learning_rate": 0.0001,
+      "loss": 1.6464,
+      "step": 5610
+    },
+    {
+      "epoch": 0.6518733662503631,
+      "grad_norm": 0.4086950421333313,
+      "learning_rate": 0.0001,
+      "loss": 1.3479,
+      "step": 5611
+    },
+    {
+      "epoch": 0.6519895440023236,
+      "grad_norm": 0.4752063453197479,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 5612
+    },
+    {
+      "epoch": 0.652105721754284,
+      "grad_norm": 0.41802459955215454,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 5613
+    },
+    {
+      "epoch": 0.6522218995062445,
+      "grad_norm": 0.43375664949417114,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 5614
+    },
+    {
+      "epoch": 0.652338077258205,
+      "grad_norm": 0.4103739559650421,
+      "learning_rate": 0.0001,
+      "loss": 1.3726,
+      "step": 5615
+    },
+    {
+      "epoch": 0.6524542550101655,
+      "grad_norm": 0.4138474464416504,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 5616
+    },
+    {
+      "epoch": 0.652570432762126,
+      "grad_norm": 0.4464971423149109,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 5617
+    },
+    {
+      "epoch": 0.6526866105140866,
+      "grad_norm": 0.4133894741535187,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 5618
+    },
+    {
+      "epoch": 0.6528027882660471,
+      "grad_norm": 0.4327857792377472,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 5619
+    },
+    {
+      "epoch": 0.6529189660180076,
+      "grad_norm": 0.4483093321323395,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 5620
+    },
+    {
+      "epoch": 0.6530351437699681,
+      "grad_norm": 0.43492817878723145,
+      "learning_rate": 0.0001,
+      "loss": 1.6154,
+      "step": 5621
+    },
+    {
+      "epoch": 0.6531513215219286,
+      "grad_norm": 0.4094054400920868,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 5622
+    },
+    {
+      "epoch": 0.653267499273889,
+      "grad_norm": 0.4260571599006653,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 5623
+    },
+    {
+      "epoch": 0.6533836770258495,
+      "grad_norm": 0.46979352831840515,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 5624
+    },
+    {
+      "epoch": 0.65349985477781,
+      "grad_norm": 0.4641943871974945,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 5625
+    },
+    {
+      "epoch": 0.6536160325297705,
+      "grad_norm": 0.4234953224658966,
+      "learning_rate": 0.0001,
+      "loss": 1.4507,
+      "step": 5626
+    },
+    {
+      "epoch": 0.653732210281731,
+      "grad_norm": 0.4371975064277649,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 5627
+    },
+    {
+      "epoch": 0.6538483880336915,
+      "grad_norm": 0.46449965238571167,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 5628
+    },
+    {
+      "epoch": 0.6539645657856521,
+      "grad_norm": 0.4338880479335785,
+      "learning_rate": 0.0001,
+      "loss": 1.613,
+      "step": 5629
+    },
+    {
+      "epoch": 0.6540807435376126,
+      "grad_norm": 0.4431876540184021,
+      "learning_rate": 0.0001,
+      "loss": 1.628,
+      "step": 5630
+    },
+    {
+      "epoch": 0.6541969212895731,
+      "grad_norm": 0.4568195641040802,
+      "learning_rate": 0.0001,
+      "loss": 1.7121,
+      "step": 5631
+    },
+    {
+      "epoch": 0.6543130990415336,
+      "grad_norm": 0.42718222737312317,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 5632
+    },
+    {
+      "epoch": 0.654429276793494,
+      "grad_norm": 0.4288146197795868,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 5633
+    },
+    {
+      "epoch": 0.6545454545454545,
+      "grad_norm": 0.4955886900424957,
+      "learning_rate": 0.0001,
+      "loss": 1.8059,
+      "step": 5634
+    },
+    {
+      "epoch": 0.654661632297415,
+      "grad_norm": 0.46111878752708435,
+      "learning_rate": 0.0001,
+      "loss": 1.6836,
+      "step": 5635
+    },
+    {
+      "epoch": 0.6547778100493755,
+      "grad_norm": 0.46014752984046936,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 5636
+    },
+    {
+      "epoch": 0.654893987801336,
+      "grad_norm": 0.4325878322124481,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 5637
+    },
+    {
+      "epoch": 0.6550101655532965,
+      "grad_norm": 0.4380306899547577,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 5638
+    },
+    {
+      "epoch": 0.6551263433052571,
+      "grad_norm": 0.37767666578292847,
+      "learning_rate": 0.0001,
+      "loss": 1.1668,
+      "step": 5639
+    },
+    {
+      "epoch": 0.6552425210572176,
+      "grad_norm": 0.4658316969871521,
+      "learning_rate": 0.0001,
+      "loss": 1.6848,
+      "step": 5640
+    },
+    {
+      "epoch": 0.6553586988091781,
+      "grad_norm": 0.4257132411003113,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 5641
+    },
+    {
+      "epoch": 0.6554748765611386,
+      "grad_norm": 0.43129682540893555,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 5642
+    },
+    {
+      "epoch": 0.655591054313099,
+      "grad_norm": 0.4526180326938629,
+      "learning_rate": 0.0001,
+      "loss": 1.7933,
+      "step": 5643
+    },
+    {
+      "epoch": 0.6557072320650595,
+      "grad_norm": 0.5669481754302979,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 5644
+    },
+    {
+      "epoch": 0.65582340981702,
+      "grad_norm": 0.4606925845146179,
+      "learning_rate": 0.0001,
+      "loss": 1.6747,
+      "step": 5645
+    },
+    {
+      "epoch": 0.6559395875689805,
+      "grad_norm": 0.4286077916622162,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 5646
+    },
+    {
+      "epoch": 0.656055765320941,
+      "grad_norm": 0.422355979681015,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 5647
+    },
+    {
+      "epoch": 0.6561719430729015,
+      "grad_norm": 0.4737590551376343,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 5648
+    },
+    {
+      "epoch": 0.656288120824862,
+      "grad_norm": 0.47408369183540344,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 5649
+    },
+    {
+      "epoch": 0.6564042985768226,
+      "grad_norm": 0.4657224416732788,
+      "learning_rate": 0.0001,
+      "loss": 1.714,
+      "step": 5650
+    },
+    {
+      "epoch": 0.6565204763287831,
+      "grad_norm": 0.4612825810909271,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 5651
+    },
+    {
+      "epoch": 0.6566366540807436,
+      "grad_norm": 0.4196070730686188,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 5652
+    },
+    {
+      "epoch": 0.656752831832704,
+      "grad_norm": 0.43713754415512085,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 5653
+    },
+    {
+      "epoch": 0.6568690095846645,
+      "grad_norm": 0.4284767508506775,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 5654
+    },
+    {
+      "epoch": 0.656985187336625,
+      "grad_norm": 0.4555676281452179,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 5655
+    },
+    {
+      "epoch": 0.6571013650885855,
+      "grad_norm": 0.43046674132347107,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 5656
+    },
+    {
+      "epoch": 0.657217542840546,
+      "grad_norm": 0.4417819678783417,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 5657
+    },
+    {
+      "epoch": 0.6573337205925065,
+      "grad_norm": 0.4391897916793823,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 5658
+    },
+    {
+      "epoch": 0.657449898344467,
+      "grad_norm": 0.4507284164428711,
+      "learning_rate": 0.0001,
+      "loss": 1.6983,
+      "step": 5659
+    },
+    {
+      "epoch": 0.6575660760964276,
+      "grad_norm": 0.43838751316070557,
+      "learning_rate": 0.0001,
+      "loss": 1.6327,
+      "step": 5660
+    },
+    {
+      "epoch": 0.6576822538483881,
+      "grad_norm": 0.4852067232131958,
+      "learning_rate": 0.0001,
+      "loss": 1.7812,
+      "step": 5661
+    },
+    {
+      "epoch": 0.6577984316003486,
+      "grad_norm": 0.416293740272522,
+      "learning_rate": 0.0001,
+      "loss": 1.5342,
+      "step": 5662
+    },
+    {
+      "epoch": 0.657914609352309,
+      "grad_norm": 0.45646652579307556,
+      "learning_rate": 0.0001,
+      "loss": 1.5847,
+      "step": 5663
+    },
+    {
+      "epoch": 0.6580307871042695,
+      "grad_norm": 0.43007123470306396,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 5664
+    },
+    {
+      "epoch": 0.65814696485623,
+      "grad_norm": 0.4502851963043213,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 5665
+    },
+    {
+      "epoch": 0.6582631426081905,
+      "grad_norm": 0.44901666045188904,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 5666
+    },
+    {
+      "epoch": 0.658379320360151,
+      "grad_norm": 0.4369884431362152,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 5667
+    },
+    {
+      "epoch": 0.6584954981121115,
+      "grad_norm": 0.4983995854854584,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 5668
+    },
+    {
+      "epoch": 0.658611675864072,
+      "grad_norm": 0.5106979608535767,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 5669
+    },
+    {
+      "epoch": 0.6587278536160325,
+      "grad_norm": 0.4607015550136566,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 5670
+    },
+    {
+      "epoch": 0.6588440313679931,
+      "grad_norm": 0.44511786103248596,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 5671
+    },
+    {
+      "epoch": 0.6589602091199536,
+      "grad_norm": 0.44376087188720703,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 5672
+    },
+    {
+      "epoch": 0.659076386871914,
+      "grad_norm": 0.463748574256897,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 5673
+    },
+    {
+      "epoch": 0.6591925646238745,
+      "grad_norm": 0.4355449676513672,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 5674
+    },
+    {
+      "epoch": 0.659308742375835,
+      "grad_norm": 0.4218558967113495,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 5675
+    },
+    {
+      "epoch": 0.6594249201277955,
+      "grad_norm": 0.4563169479370117,
+      "learning_rate": 0.0001,
+      "loss": 1.7053,
+      "step": 5676
+    },
+    {
+      "epoch": 0.659541097879756,
+      "grad_norm": 0.4865387976169586,
+      "learning_rate": 0.0001,
+      "loss": 1.975,
+      "step": 5677
+    },
+    {
+      "epoch": 0.6596572756317165,
+      "grad_norm": 0.4624812602996826,
+      "learning_rate": 0.0001,
+      "loss": 1.7371,
+      "step": 5678
+    },
+    {
+      "epoch": 0.659773453383677,
+      "grad_norm": 0.4570534825325012,
+      "learning_rate": 0.0001,
+      "loss": 1.7626,
+      "step": 5679
+    },
+    {
+      "epoch": 0.6598896311356375,
+      "grad_norm": 0.42516836524009705,
+      "learning_rate": 0.0001,
+      "loss": 1.5708,
+      "step": 5680
+    },
+    {
+      "epoch": 0.6600058088875981,
+      "grad_norm": 0.4255434572696686,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 5681
+    },
+    {
+      "epoch": 0.6601219866395586,
+      "grad_norm": 0.4116609990596771,
+      "learning_rate": 0.0001,
+      "loss": 1.4613,
+      "step": 5682
+    },
+    {
+      "epoch": 0.6602381643915191,
+      "grad_norm": 0.45455899834632874,
+      "learning_rate": 0.0001,
+      "loss": 1.7634,
+      "step": 5683
+    },
+    {
+      "epoch": 0.6603543421434795,
+      "grad_norm": 0.4295278787612915,
+      "learning_rate": 0.0001,
+      "loss": 1.7607,
+      "step": 5684
+    },
+    {
+      "epoch": 0.66047051989544,
+      "grad_norm": 0.43869319558143616,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 5685
+    },
+    {
+      "epoch": 0.6605866976474005,
+      "grad_norm": 0.4462525546550751,
+      "learning_rate": 0.0001,
+      "loss": 1.7752,
+      "step": 5686
+    },
+    {
+      "epoch": 0.660702875399361,
+      "grad_norm": 0.4246017038822174,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 5687
+    },
+    {
+      "epoch": 0.6608190531513215,
+      "grad_norm": 0.43097880482673645,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 5688
+    },
+    {
+      "epoch": 0.660935230903282,
+      "grad_norm": 0.45352938771247864,
+      "learning_rate": 0.0001,
+      "loss": 1.6469,
+      "step": 5689
+    },
+    {
+      "epoch": 0.6610514086552425,
+      "grad_norm": 0.46061140298843384,
+      "learning_rate": 0.0001,
+      "loss": 1.5656,
+      "step": 5690
+    },
+    {
+      "epoch": 0.661167586407203,
+      "grad_norm": 0.4334444999694824,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 5691
+    },
+    {
+      "epoch": 0.6612837641591636,
+      "grad_norm": 0.43552273511886597,
+      "learning_rate": 0.0001,
+      "loss": 1.7532,
+      "step": 5692
+    },
+    {
+      "epoch": 0.6613999419111241,
+      "grad_norm": 0.4349510669708252,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 5693
+    },
+    {
+      "epoch": 0.6615161196630845,
+      "grad_norm": 0.45464491844177246,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 5694
+    },
+    {
+      "epoch": 0.661632297415045,
+      "grad_norm": 0.44270920753479004,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 5695
+    },
+    {
+      "epoch": 0.6617484751670055,
+      "grad_norm": 0.4432513117790222,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 5696
+    },
+    {
+      "epoch": 0.661864652918966,
+      "grad_norm": 0.4541747272014618,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 5697
+    },
+    {
+      "epoch": 0.6619808306709265,
+      "grad_norm": 0.43986377120018005,
+      "learning_rate": 0.0001,
+      "loss": 1.6965,
+      "step": 5698
+    },
+    {
+      "epoch": 0.662097008422887,
+      "grad_norm": 0.463550329208374,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 5699
+    },
+    {
+      "epoch": 0.6622131861748475,
+      "grad_norm": 0.42243123054504395,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 5700
+    },
+    {
+      "epoch": 0.662329363926808,
+      "grad_norm": 0.4550102651119232,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 5701
+    },
+    {
+      "epoch": 0.6624455416787686,
+      "grad_norm": 0.48717159032821655,
+      "learning_rate": 0.0001,
+      "loss": 1.8157,
+      "step": 5702
+    },
+    {
+      "epoch": 0.6625617194307291,
+      "grad_norm": 0.43235111236572266,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 5703
+    },
+    {
+      "epoch": 0.6626778971826895,
+      "grad_norm": 0.39643558859825134,
+      "learning_rate": 0.0001,
+      "loss": 1.4418,
+      "step": 5704
+    },
+    {
+      "epoch": 0.66279407493465,
+      "grad_norm": 0.445264607667923,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6629102526866105,
+      "grad_norm": 0.45641717314720154,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 5706
+    },
+    {
+      "epoch": 0.663026430438571,
+      "grad_norm": 0.4507564902305603,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 5707
+    },
+    {
+      "epoch": 0.6631426081905315,
+      "grad_norm": 0.4877043068408966,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 5708
+    },
+    {
+      "epoch": 0.663258785942492,
+      "grad_norm": 0.42610713839530945,
+      "learning_rate": 0.0001,
+      "loss": 1.6014,
+      "step": 5709
+    },
+    {
+      "epoch": 0.6633749636944525,
+      "grad_norm": 0.4246615767478943,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 5710
+    },
+    {
+      "epoch": 0.663491141446413,
+      "grad_norm": 0.46146342158317566,
+      "learning_rate": 0.0001,
+      "loss": 1.7475,
+      "step": 5711
+    },
+    {
+      "epoch": 0.6636073191983735,
+      "grad_norm": 0.45187118649482727,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 5712
+    },
+    {
+      "epoch": 0.6637234969503341,
+      "grad_norm": 0.4408842921257019,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 5713
+    },
+    {
+      "epoch": 0.6638396747022945,
+      "grad_norm": 0.4070923626422882,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 5714
+    },
+    {
+      "epoch": 0.663955852454255,
+      "grad_norm": 0.4318831264972687,
+      "learning_rate": 0.0001,
+      "loss": 1.5662,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6640720302062155,
+      "grad_norm": 0.48511743545532227,
+      "learning_rate": 0.0001,
+      "loss": 1.7734,
+      "step": 5716
+    },
+    {
+      "epoch": 0.664188207958176,
+      "grad_norm": 0.41092416644096375,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 5717
+    },
+    {
+      "epoch": 0.6643043857101365,
+      "grad_norm": 0.4554840922355652,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 5718
+    },
+    {
+      "epoch": 0.664420563462097,
+      "grad_norm": 0.42100241780281067,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 5719
+    },
+    {
+      "epoch": 0.6645367412140575,
+      "grad_norm": 0.421871542930603,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 5720
+    },
+    {
+      "epoch": 0.664652918966018,
+      "grad_norm": 0.46472758054733276,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 5721
+    },
+    {
+      "epoch": 0.6647690967179785,
+      "grad_norm": 0.4486168920993805,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 5722
+    },
+    {
+      "epoch": 0.6648852744699391,
+      "grad_norm": 0.44002678990364075,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6650014522218995,
+      "grad_norm": 0.427792489528656,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 5724
+    },
+    {
+      "epoch": 0.66511762997386,
+      "grad_norm": 0.44811704754829407,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6652338077258205,
+      "grad_norm": 0.4508500099182129,
+      "learning_rate": 0.0001,
+      "loss": 1.7428,
+      "step": 5726
+    },
+    {
+      "epoch": 0.665349985477781,
+      "grad_norm": 0.4322895407676697,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 5727
+    },
+    {
+      "epoch": 0.6654661632297415,
+      "grad_norm": 0.4665115177631378,
+      "learning_rate": 0.0001,
+      "loss": 1.7279,
+      "step": 5728
+    },
+    {
+      "epoch": 0.665582340981702,
+      "grad_norm": 0.45049217343330383,
+      "learning_rate": 0.0001,
+      "loss": 1.6839,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6656985187336625,
+      "grad_norm": 0.4667704701423645,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 5730
+    },
+    {
+      "epoch": 0.665814696485623,
+      "grad_norm": 0.43236711621284485,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6659308742375835,
+      "grad_norm": 0.4469371438026428,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 5732
+    },
+    {
+      "epoch": 0.666047051989544,
+      "grad_norm": 0.4933030307292938,
+      "learning_rate": 0.0001,
+      "loss": 1.7527,
+      "step": 5733
+    },
+    {
+      "epoch": 0.6661632297415045,
+      "grad_norm": 0.466074138879776,
+      "learning_rate": 0.0001,
+      "loss": 1.658,
+      "step": 5734
+    },
+    {
+      "epoch": 0.666279407493465,
+      "grad_norm": 0.418959379196167,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6663955852454255,
+      "grad_norm": 0.4188354015350342,
+      "learning_rate": 0.0001,
+      "loss": 1.4785,
+      "step": 5736
+    },
+    {
+      "epoch": 0.666511762997386,
+      "grad_norm": 0.46846193075180054,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 5737
+    },
+    {
+      "epoch": 0.6666279407493465,
+      "grad_norm": 0.41429415345191956,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 5738
+    },
+    {
+      "epoch": 0.666744118501307,
+      "grad_norm": 0.4290158748626709,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 5739
+    },
+    {
+      "epoch": 0.6668602962532675,
+      "grad_norm": 0.45462852716445923,
+      "learning_rate": 0.0001,
+      "loss": 1.7486,
+      "step": 5740
+    },
+    {
+      "epoch": 0.666976474005228,
+      "grad_norm": 0.4142838716506958,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 5741
+    },
+    {
+      "epoch": 0.6670926517571885,
+      "grad_norm": 0.4205038249492645,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 5742
+    },
+    {
+      "epoch": 0.667208829509149,
+      "grad_norm": 0.45246177911758423,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6673250072611095,
+      "grad_norm": 0.4968937039375305,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 5744
+    },
+    {
+      "epoch": 0.66744118501307,
+      "grad_norm": 0.44130516052246094,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 5745
+    },
+    {
+      "epoch": 0.6675573627650305,
+      "grad_norm": 0.4321398138999939,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 5746
+    },
+    {
+      "epoch": 0.667673540516991,
+      "grad_norm": 0.43933340907096863,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6677897182689515,
+      "grad_norm": 0.4424324333667755,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 5748
+    },
+    {
+      "epoch": 0.667905896020912,
+      "grad_norm": 0.44031408429145813,
+      "learning_rate": 0.0001,
+      "loss": 1.6682,
+      "step": 5749
+    },
+    {
+      "epoch": 0.6680220737728725,
+      "grad_norm": 0.4373178482055664,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 5750
+    },
+    {
+      "epoch": 0.668138251524833,
+      "grad_norm": 0.517252504825592,
+      "learning_rate": 0.0001,
+      "loss": 1.8808,
+      "step": 5751
+    },
+    {
+      "epoch": 0.6682544292767935,
+      "grad_norm": 0.44176408648490906,
+      "learning_rate": 0.0001,
+      "loss": 1.6645,
+      "step": 5752
+    },
+    {
+      "epoch": 0.668370607028754,
+      "grad_norm": 0.4362224340438843,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 5753
+    },
+    {
+      "epoch": 0.6684867847807144,
+      "grad_norm": 0.44093477725982666,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 5754
+    },
+    {
+      "epoch": 0.668602962532675,
+      "grad_norm": 0.42824220657348633,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6687191402846355,
+      "grad_norm": 0.46376579999923706,
+      "learning_rate": 0.0001,
+      "loss": 1.7126,
+      "step": 5756
+    },
+    {
+      "epoch": 0.668835318036596,
+      "grad_norm": 0.47102320194244385,
+      "learning_rate": 0.0001,
+      "loss": 1.8464,
+      "step": 5757
+    },
+    {
+      "epoch": 0.6689514957885565,
+      "grad_norm": 0.4689522385597229,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 5758
+    },
+    {
+      "epoch": 0.669067673540517,
+      "grad_norm": 0.43695563077926636,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 5759
+    },
+    {
+      "epoch": 0.6691838512924775,
+      "grad_norm": 0.48102685809135437,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 5760
+    },
+    {
+      "epoch": 0.669300029044438,
+      "grad_norm": 0.4178447127342224,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 5761
+    },
+    {
+      "epoch": 0.6694162067963985,
+      "grad_norm": 0.43344780802726746,
+      "learning_rate": 0.0001,
+      "loss": 1.5005,
+      "step": 5762
+    },
+    {
+      "epoch": 0.669532384548359,
+      "grad_norm": 0.4385789632797241,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 5763
+    },
+    {
+      "epoch": 0.6696485623003194,
+      "grad_norm": 0.49828559160232544,
+      "learning_rate": 0.0001,
+      "loss": 1.8065,
+      "step": 5764
+    },
+    {
+      "epoch": 0.66976474005228,
+      "grad_norm": 0.45839375257492065,
+      "learning_rate": 0.0001,
+      "loss": 1.7687,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6698809178042405,
+      "grad_norm": 0.45659172534942627,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 5766
+    },
+    {
+      "epoch": 0.669997095556201,
+      "grad_norm": 0.43861255049705505,
+      "learning_rate": 0.0001,
+      "loss": 1.3606,
+      "step": 5767
+    },
+    {
+      "epoch": 0.6701132733081615,
+      "grad_norm": 0.47588106989860535,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 5768
+    },
+    {
+      "epoch": 0.670229451060122,
+      "grad_norm": 0.43281859159469604,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 5769
+    },
+    {
+      "epoch": 0.6703456288120825,
+      "grad_norm": 0.42776212096214294,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 5770
+    },
+    {
+      "epoch": 0.670461806564043,
+      "grad_norm": 0.48418179154396057,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 5771
+    },
+    {
+      "epoch": 0.6705779843160035,
+      "grad_norm": 0.47568055987358093,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 5772
+    },
+    {
+      "epoch": 0.670694162067964,
+      "grad_norm": 0.46286889910697937,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 5773
+    },
+    {
+      "epoch": 0.6708103398199244,
+      "grad_norm": 0.42690396308898926,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 5774
+    },
+    {
+      "epoch": 0.670926517571885,
+      "grad_norm": 0.5125055313110352,
+      "learning_rate": 0.0001,
+      "loss": 1.8218,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6710426953238455,
+      "grad_norm": 0.4569031894207001,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 5776
+    },
+    {
+      "epoch": 0.671158873075806,
+      "grad_norm": 0.3980884253978729,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6712750508277665,
+      "grad_norm": 0.450810968875885,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 5778
+    },
+    {
+      "epoch": 0.671391228579727,
+      "grad_norm": 0.47312337160110474,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6715074063316875,
+      "grad_norm": 0.42309361696243286,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 5780
+    },
+    {
+      "epoch": 0.671623584083648,
+      "grad_norm": 0.4467894434928894,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 5781
+    },
+    {
+      "epoch": 0.6717397618356085,
+      "grad_norm": 0.41931766271591187,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 5782
+    },
+    {
+      "epoch": 0.671855939587569,
+      "grad_norm": 0.49468186497688293,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 5783
+    },
+    {
+      "epoch": 0.6719721173395294,
+      "grad_norm": 0.5051024556159973,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 5784
+    },
+    {
+      "epoch": 0.6720882950914899,
+      "grad_norm": 0.4598732888698578,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6722044728434505,
+      "grad_norm": 0.4720596671104431,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 5786
+    },
+    {
+      "epoch": 0.672320650595411,
+      "grad_norm": 0.40756291151046753,
+      "learning_rate": 0.0001,
+      "loss": 1.4437,
+      "step": 5787
+    },
+    {
+      "epoch": 0.6724368283473715,
+      "grad_norm": 0.44663846492767334,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 5788
+    },
+    {
+      "epoch": 0.672553006099332,
+      "grad_norm": 0.44735023379325867,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6726691838512925,
+      "grad_norm": 0.4546589255332947,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 5790
+    },
+    {
+      "epoch": 0.672785361603253,
+      "grad_norm": 0.44948139786720276,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6729015393552135,
+      "grad_norm": 0.45747241377830505,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 5792
+    },
+    {
+      "epoch": 0.673017717107174,
+      "grad_norm": 0.46398770809173584,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 5793
+    },
+    {
+      "epoch": 0.6731338948591344,
+      "grad_norm": 0.4531368017196655,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6732500726110949,
+      "grad_norm": 0.43872615694999695,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 5795
+    },
+    {
+      "epoch": 0.6733662503630555,
+      "grad_norm": 0.4827875792980194,
+      "learning_rate": 0.0001,
+      "loss": 1.7964,
+      "step": 5796
+    },
+    {
+      "epoch": 0.673482428115016,
+      "grad_norm": 0.46272164583206177,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6735986058669765,
+      "grad_norm": 0.43684569001197815,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 5798
+    },
+    {
+      "epoch": 0.673714783618937,
+      "grad_norm": 0.43560728430747986,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6738309613708975,
+      "grad_norm": 0.43896791338920593,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 5800
+    },
+    {
+      "epoch": 0.673947139122858,
+      "grad_norm": 0.4776453375816345,
+      "learning_rate": 0.0001,
+      "loss": 1.64,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6740633168748185,
+      "grad_norm": 0.44419828057289124,
+      "learning_rate": 0.0001,
+      "loss": 1.6425,
+      "step": 5802
+    },
+    {
+      "epoch": 0.674179494626779,
+      "grad_norm": 0.44315245747566223,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6742956723787394,
+      "grad_norm": 0.4680185914039612,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 5804
+    },
+    {
+      "epoch": 0.6744118501306999,
+      "grad_norm": 0.4504956007003784,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6745280278826604,
+      "grad_norm": 0.4400990307331085,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 5806
+    },
+    {
+      "epoch": 0.674644205634621,
+      "grad_norm": 0.5055308938026428,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 5807
+    },
+    {
+      "epoch": 0.6747603833865815,
+      "grad_norm": 0.4236086905002594,
+      "learning_rate": 0.0001,
+      "loss": 1.6601,
+      "step": 5808
+    },
+    {
+      "epoch": 0.674876561138542,
+      "grad_norm": 0.4559587836265564,
+      "learning_rate": 0.0001,
+      "loss": 1.7122,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6749927388905025,
+      "grad_norm": 0.42813077569007874,
+      "learning_rate": 0.0001,
+      "loss": 1.6111,
+      "step": 5810
+    },
+    {
+      "epoch": 0.675108916642463,
+      "grad_norm": 0.4201710820198059,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6752250943944235,
+      "grad_norm": 0.4242464005947113,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 5812
+    },
+    {
+      "epoch": 0.675341272146384,
+      "grad_norm": 0.45076656341552734,
+      "learning_rate": 0.0001,
+      "loss": 1.698,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6754574498983444,
+      "grad_norm": 0.4501783549785614,
+      "learning_rate": 0.0001,
+      "loss": 1.7337,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6755736276503049,
+      "grad_norm": 0.4316518306732178,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6756898054022654,
+      "grad_norm": 0.4382632076740265,
+      "learning_rate": 0.0001,
+      "loss": 1.8078,
+      "step": 5816
+    },
+    {
+      "epoch": 0.675805983154226,
+      "grad_norm": 0.42367810010910034,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6759221609061865,
+      "grad_norm": 0.434133917093277,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 5818
+    },
+    {
+      "epoch": 0.676038338658147,
+      "grad_norm": 0.4283505976200104,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6761545164101075,
+      "grad_norm": 0.4194677770137787,
+      "learning_rate": 0.0001,
+      "loss": 1.6286,
+      "step": 5820
+    },
+    {
+      "epoch": 0.676270694162068,
+      "grad_norm": 0.43860924243927,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 5821
+    },
+    {
+      "epoch": 0.6763868719140285,
+      "grad_norm": 0.43896040320396423,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 5822
+    },
+    {
+      "epoch": 0.676503049665989,
+      "grad_norm": 0.41808873414993286,
+      "learning_rate": 0.0001,
+      "loss": 1.3372,
+      "step": 5823
+    },
+    {
+      "epoch": 0.6766192274179494,
+      "grad_norm": 0.5032152533531189,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 5824
+    },
+    {
+      "epoch": 0.6767354051699099,
+      "grad_norm": 0.48192158341407776,
+      "learning_rate": 0.0001,
+      "loss": 1.7263,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6768515829218704,
+      "grad_norm": 0.4845306873321533,
+      "learning_rate": 0.0001,
+      "loss": 1.7389,
+      "step": 5826
+    },
+    {
+      "epoch": 0.6769677606738309,
+      "grad_norm": 0.46503588557243347,
+      "learning_rate": 0.0001,
+      "loss": 1.7579,
+      "step": 5827
+    },
+    {
+      "epoch": 0.6770839384257915,
+      "grad_norm": 0.42725270986557007,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 5828
+    },
+    {
+      "epoch": 0.677200116177752,
+      "grad_norm": 0.43792206048965454,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 5829
+    },
+    {
+      "epoch": 0.6773162939297125,
+      "grad_norm": 0.42509621381759644,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 5830
+    },
+    {
+      "epoch": 0.677432471681673,
+      "grad_norm": 0.4887797236442566,
+      "learning_rate": 0.0001,
+      "loss": 1.7904,
+      "step": 5831
+    },
+    {
+      "epoch": 0.6775486494336335,
+      "grad_norm": 0.4223867356777191,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 5832
+    },
+    {
+      "epoch": 0.677664827185594,
+      "grad_norm": 0.44735440611839294,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 5833
+    },
+    {
+      "epoch": 0.6777810049375544,
+      "grad_norm": 0.4342286288738251,
+      "learning_rate": 0.0001,
+      "loss": 1.4604,
+      "step": 5834
+    },
+    {
+      "epoch": 0.6778971826895149,
+      "grad_norm": 0.47264301776885986,
+      "learning_rate": 0.0001,
+      "loss": 1.5672,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6780133604414754,
+      "grad_norm": 0.45144325494766235,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 5836
+    },
+    {
+      "epoch": 0.6781295381934359,
+      "grad_norm": 0.45070862770080566,
+      "learning_rate": 0.0001,
+      "loss": 1.7091,
+      "step": 5837
+    },
+    {
+      "epoch": 0.6782457159453965,
+      "grad_norm": 0.44926077127456665,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 5838
+    },
+    {
+      "epoch": 0.678361893697357,
+      "grad_norm": 0.42580538988113403,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 5839
+    },
+    {
+      "epoch": 0.6784780714493175,
+      "grad_norm": 0.4390771985054016,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 5840
+    },
+    {
+      "epoch": 0.678594249201278,
+      "grad_norm": 0.43495962023735046,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 5841
+    },
+    {
+      "epoch": 0.6787104269532385,
+      "grad_norm": 0.42545855045318604,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 5842
+    },
+    {
+      "epoch": 0.678826604705199,
+      "grad_norm": 0.40862634778022766,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 5843
+    },
+    {
+      "epoch": 0.6789427824571594,
+      "grad_norm": 0.4625224173069,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 5844
+    },
+    {
+      "epoch": 0.6790589602091199,
+      "grad_norm": 0.4572260081768036,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6791751379610804,
+      "grad_norm": 0.45550736784935,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 5846
+    },
+    {
+      "epoch": 0.6792913157130409,
+      "grad_norm": 0.49932238459587097,
+      "learning_rate": 0.0001,
+      "loss": 1.6559,
+      "step": 5847
+    },
+    {
+      "epoch": 0.6794074934650014,
+      "grad_norm": 0.43598851561546326,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 5848
+    },
+    {
+      "epoch": 0.679523671216962,
+      "grad_norm": 0.4407789707183838,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 5849
+    },
+    {
+      "epoch": 0.6796398489689225,
+      "grad_norm": 0.5003304481506348,
+      "learning_rate": 0.0001,
+      "loss": 1.9146,
+      "step": 5850
+    },
+    {
+      "epoch": 0.679756026720883,
+      "grad_norm": 0.45455658435821533,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6798722044728435,
+      "grad_norm": 0.429861456155777,
+      "learning_rate": 0.0001,
+      "loss": 1.7047,
+      "step": 5852
+    },
+    {
+      "epoch": 0.679988382224804,
+      "grad_norm": 0.41853785514831543,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 5853
+    },
+    {
+      "epoch": 0.6801045599767644,
+      "grad_norm": 0.4636412560939789,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6802207377287249,
+      "grad_norm": 0.4763379395008087,
+      "learning_rate": 0.0001,
+      "loss": 1.9099,
+      "step": 5855
+    },
+    {
+      "epoch": 0.6803369154806854,
+      "grad_norm": 0.40933549404144287,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 5856
+    },
+    {
+      "epoch": 0.6804530932326459,
+      "grad_norm": 0.43275657296180725,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6805692709846064,
+      "grad_norm": 0.4391117990016937,
+      "learning_rate": 0.0001,
+      "loss": 1.6513,
+      "step": 5858
+    },
+    {
+      "epoch": 0.680685448736567,
+      "grad_norm": 0.456015408039093,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6808016264885275,
+      "grad_norm": 0.4378184378147125,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 5860
+    },
+    {
+      "epoch": 0.680917804240488,
+      "grad_norm": 0.4603559672832489,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6810339819924485,
+      "grad_norm": 0.46735942363739014,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 5862
+    },
+    {
+      "epoch": 0.681150159744409,
+      "grad_norm": 0.42926350235939026,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 5863
+    },
+    {
+      "epoch": 0.6812663374963694,
+      "grad_norm": 0.4338090717792511,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6813825152483299,
+      "grad_norm": 0.4733595550060272,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6814986930002904,
+      "grad_norm": 0.43445518612861633,
+      "learning_rate": 0.0001,
+      "loss": 1.6658,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6816148707522509,
+      "grad_norm": 0.4225948750972748,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6817310485042114,
+      "grad_norm": 0.4454067051410675,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6818472262561719,
+      "grad_norm": 0.4258386194705963,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6819634040081325,
+      "grad_norm": 0.5034085512161255,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 5870
+    },
+    {
+      "epoch": 0.682079581760093,
+      "grad_norm": 0.48741772770881653,
+      "learning_rate": 0.0001,
+      "loss": 1.8736,
+      "step": 5871
+    },
+    {
+      "epoch": 0.6821957595120535,
+      "grad_norm": 0.4586668908596039,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 5872
+    },
+    {
+      "epoch": 0.682311937264014,
+      "grad_norm": 0.445441335439682,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6824281150159744,
+      "grad_norm": 0.42984551191329956,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6825442927679349,
+      "grad_norm": 0.42550551891326904,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6826604705198954,
+      "grad_norm": 0.43332982063293457,
+      "learning_rate": 0.0001,
+      "loss": 1.4163,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6827766482718559,
+      "grad_norm": 0.46797114610671997,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 5877
+    },
+    {
+      "epoch": 0.6828928260238164,
+      "grad_norm": 0.45049670338630676,
+      "learning_rate": 0.0001,
+      "loss": 1.7157,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6830090037757769,
+      "grad_norm": 0.4513972997665405,
+      "learning_rate": 0.0001,
+      "loss": 1.5783,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6831251815277375,
+      "grad_norm": 0.46550363302230835,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 5880
+    },
+    {
+      "epoch": 0.683241359279698,
+      "grad_norm": 0.4839429259300232,
+      "learning_rate": 0.0001,
+      "loss": 1.7258,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6833575370316585,
+      "grad_norm": 0.4608222544193268,
+      "learning_rate": 0.0001,
+      "loss": 1.5762,
+      "step": 5882
+    },
+    {
+      "epoch": 0.683473714783619,
+      "grad_norm": 0.4843682050704956,
+      "learning_rate": 0.0001,
+      "loss": 1.7465,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6835898925355794,
+      "grad_norm": 0.44494950771331787,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6837060702875399,
+      "grad_norm": 0.4825565814971924,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6838222480395004,
+      "grad_norm": 0.4432181715965271,
+      "learning_rate": 0.0001,
+      "loss": 1.7055,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6839384257914609,
+      "grad_norm": 0.43164384365081787,
+      "learning_rate": 0.0001,
+      "loss": 1.6208,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6840546035434214,
+      "grad_norm": 0.4706867039203644,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6841707812953819,
+      "grad_norm": 0.4217129647731781,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6842869590473424,
+      "grad_norm": 0.4253685772418976,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 5890
+    },
+    {
+      "epoch": 0.684403136799303,
+      "grad_norm": 0.431268572807312,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6845193145512635,
+      "grad_norm": 0.44175827503204346,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 5892
+    },
+    {
+      "epoch": 0.684635492303224,
+      "grad_norm": 0.41143807768821716,
+      "learning_rate": 0.0001,
+      "loss": 1.5274,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6847516700551844,
+      "grad_norm": 0.4632609784603119,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6848678478071449,
+      "grad_norm": 0.45641466975212097,
+      "learning_rate": 0.0001,
+      "loss": 1.7331,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6849840255591054,
+      "grad_norm": 0.4799252152442932,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6851002033110659,
+      "grad_norm": 0.42571592330932617,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6852163810630264,
+      "grad_norm": 0.42581233382225037,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6853325588149869,
+      "grad_norm": 0.4454191327095032,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 5899
+    },
+    {
+      "epoch": 0.6854487365669474,
+      "grad_norm": 0.4443952143192291,
+      "learning_rate": 0.0001,
+      "loss": 1.3731,
+      "step": 5900
+    },
+    {
+      "epoch": 0.685564914318908,
+      "grad_norm": 0.4548780918121338,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6856810920708685,
+      "grad_norm": 0.4646033048629761,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 5902
+    },
+    {
+      "epoch": 0.685797269822829,
+      "grad_norm": 0.4402289390563965,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6859134475747894,
+      "grad_norm": 0.462422251701355,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6860296253267499,
+      "grad_norm": 0.4708074629306793,
+      "learning_rate": 0.0001,
+      "loss": 1.7784,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6861458030787104,
+      "grad_norm": 0.4279741942882538,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6862619808306709,
+      "grad_norm": 0.45395079255104065,
+      "learning_rate": 0.0001,
+      "loss": 1.7005,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6863781585826314,
+      "grad_norm": 0.46580180525779724,
+      "learning_rate": 0.0001,
+      "loss": 1.7221,
+      "step": 5908
+    },
+    {
+      "epoch": 0.6864943363345919,
+      "grad_norm": 0.44675102829933167,
+      "learning_rate": 0.0001,
+      "loss": 1.6527,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6866105140865524,
+      "grad_norm": 0.41608941555023193,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6867266918385129,
+      "grad_norm": 0.4041667878627777,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 5911
+    },
+    {
+      "epoch": 0.6868428695904735,
+      "grad_norm": 0.46621033549308777,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 5912
+    },
+    {
+      "epoch": 0.686959047342434,
+      "grad_norm": 0.42089179158210754,
+      "learning_rate": 0.0001,
+      "loss": 1.3727,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6870752250943944,
+      "grad_norm": 0.43382903933525085,
+      "learning_rate": 0.0001,
+      "loss": 1.3902,
+      "step": 5914
+    },
+    {
+      "epoch": 0.6871914028463549,
+      "grad_norm": 0.45685744285583496,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6873075805983154,
+      "grad_norm": 0.4664427936077118,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6874237583502759,
+      "grad_norm": 0.43341436982154846,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6875399361022364,
+      "grad_norm": 0.4530865252017975,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 5918
+    },
+    {
+      "epoch": 0.6876561138541969,
+      "grad_norm": 0.4333666265010834,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 5919
+    },
+    {
+      "epoch": 0.6877722916061574,
+      "grad_norm": 0.4299314618110657,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6878884693581179,
+      "grad_norm": 0.43617698550224304,
+      "learning_rate": 0.0001,
+      "loss": 1.7213,
+      "step": 5921
+    },
+    {
+      "epoch": 0.6880046471100785,
+      "grad_norm": 0.4485790431499481,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 5922
+    },
+    {
+      "epoch": 0.688120824862039,
+      "grad_norm": 0.44341012835502625,
+      "learning_rate": 0.0001,
+      "loss": 1.4585,
+      "step": 5923
+    },
+    {
+      "epoch": 0.6882370026139994,
+      "grad_norm": 0.4278321862220764,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 5924
+    },
+    {
+      "epoch": 0.6883531803659599,
+      "grad_norm": 0.4742536246776581,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6884693581179204,
+      "grad_norm": 0.43478327989578247,
+      "learning_rate": 0.0001,
+      "loss": 1.4958,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6885855358698809,
+      "grad_norm": 0.4236944317817688,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 5927
+    },
+    {
+      "epoch": 0.6887017136218414,
+      "grad_norm": 0.4370484948158264,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 5928
+    },
+    {
+      "epoch": 0.6888178913738019,
+      "grad_norm": 0.45525872707366943,
+      "learning_rate": 0.0001,
+      "loss": 1.6832,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6889340691257624,
+      "grad_norm": 0.4394777715206146,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6890502468777229,
+      "grad_norm": 0.4571029841899872,
+      "learning_rate": 0.0001,
+      "loss": 1.7596,
+      "step": 5931
+    },
+    {
+      "epoch": 0.6891664246296834,
+      "grad_norm": 0.4554816782474518,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 5932
+    },
+    {
+      "epoch": 0.689282602381644,
+      "grad_norm": 0.43854042887687683,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 5933
+    },
+    {
+      "epoch": 0.6893987801336044,
+      "grad_norm": 0.4328182339668274,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6895149578855649,
+      "grad_norm": 0.44765332341194153,
+      "learning_rate": 0.0001,
+      "loss": 1.6783,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6896311356375254,
+      "grad_norm": 0.43624910712242126,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 5936
+    },
+    {
+      "epoch": 0.6897473133894859,
+      "grad_norm": 0.46665430068969727,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 5937
+    },
+    {
+      "epoch": 0.6898634911414464,
+      "grad_norm": 0.45889371633529663,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6899796688934069,
+      "grad_norm": 0.4659807085990906,
+      "learning_rate": 0.0001,
+      "loss": 1.878,
+      "step": 5939
+    },
+    {
+      "epoch": 0.6900958466453674,
+      "grad_norm": 0.45350781083106995,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 5940
+    },
+    {
+      "epoch": 0.6902120243973279,
+      "grad_norm": 0.44755759835243225,
+      "learning_rate": 0.0001,
+      "loss": 1.6358,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6903282021492884,
+      "grad_norm": 0.39925143122673035,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 5942
+    },
+    {
+      "epoch": 0.690444379901249,
+      "grad_norm": 0.4676305055618286,
+      "learning_rate": 0.0001,
+      "loss": 1.7827,
+      "step": 5943
+    },
+    {
+      "epoch": 0.6905605576532095,
+      "grad_norm": 0.42078080773353577,
+      "learning_rate": 0.0001,
+      "loss": 1.3661,
+      "step": 5944
+    },
+    {
+      "epoch": 0.6906767354051699,
+      "grad_norm": 0.4294394254684448,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6907929131571304,
+      "grad_norm": 0.4754204750061035,
+      "learning_rate": 0.0001,
+      "loss": 1.4769,
+      "step": 5946
+    },
+    {
+      "epoch": 0.6909090909090909,
+      "grad_norm": 0.4373209476470947,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 5947
+    },
+    {
+      "epoch": 0.6910252686610514,
+      "grad_norm": 0.4468858242034912,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 5948
+    },
+    {
+      "epoch": 0.6911414464130119,
+      "grad_norm": 0.41958677768707275,
+      "learning_rate": 0.0001,
+      "loss": 1.6706,
+      "step": 5949
+    },
+    {
+      "epoch": 0.6912576241649724,
+      "grad_norm": 0.4472271203994751,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6913738019169329,
+      "grad_norm": 0.4623844623565674,
+      "learning_rate": 0.0001,
+      "loss": 1.6781,
+      "step": 5951
+    },
+    {
+      "epoch": 0.6914899796688934,
+      "grad_norm": 0.4381425082683563,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 5952
+    },
+    {
+      "epoch": 0.691606157420854,
+      "grad_norm": 0.4867214858531952,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 5953
+    },
+    {
+      "epoch": 0.6917223351728145,
+      "grad_norm": 0.45404472947120667,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 5954
+    },
+    {
+      "epoch": 0.6918385129247749,
+      "grad_norm": 0.45104196667671204,
+      "learning_rate": 0.0001,
+      "loss": 1.759,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6919546906767354,
+      "grad_norm": 0.43430057168006897,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 5956
+    },
+    {
+      "epoch": 0.6920708684286959,
+      "grad_norm": 0.4231088161468506,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 5957
+    },
+    {
+      "epoch": 0.6921870461806564,
+      "grad_norm": 0.43777701258659363,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 5958
+    },
+    {
+      "epoch": 0.6923032239326169,
+      "grad_norm": 0.4415545165538788,
+      "learning_rate": 0.0001,
+      "loss": 1.6971,
+      "step": 5959
+    },
+    {
+      "epoch": 0.6924194016845774,
+      "grad_norm": 0.4693138599395752,
+      "learning_rate": 0.0001,
+      "loss": 1.827,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6925355794365379,
+      "grad_norm": 0.458947092294693,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 5961
+    },
+    {
+      "epoch": 0.6926517571884984,
+      "grad_norm": 0.47753122448921204,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 5962
+    },
+    {
+      "epoch": 0.6927679349404589,
+      "grad_norm": 0.44865846633911133,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 5963
+    },
+    {
+      "epoch": 0.6928841126924195,
+      "grad_norm": 0.46769386529922485,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 5964
+    },
+    {
+      "epoch": 0.6930002904443799,
+      "grad_norm": 0.46594002842903137,
+      "learning_rate": 0.0001,
+      "loss": 1.5307,
+      "step": 5965
+    },
+    {
+      "epoch": 0.6931164681963404,
+      "grad_norm": 0.44469648599624634,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 5966
+    },
+    {
+      "epoch": 0.6932326459483009,
+      "grad_norm": 0.44019195437431335,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 5967
+    },
+    {
+      "epoch": 0.6933488237002614,
+      "grad_norm": 0.4828648567199707,
+      "learning_rate": 0.0001,
+      "loss": 1.734,
+      "step": 5968
+    },
+    {
+      "epoch": 0.6934650014522219,
+      "grad_norm": 0.4316853880882263,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 5969
+    },
+    {
+      "epoch": 0.6935811792041824,
+      "grad_norm": 0.45684826374053955,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6936973569561429,
+      "grad_norm": 0.4338705539703369,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 5971
+    },
+    {
+      "epoch": 0.6938135347081034,
+      "grad_norm": 0.46950671076774597,
+      "learning_rate": 0.0001,
+      "loss": 1.6252,
+      "step": 5972
+    },
+    {
+      "epoch": 0.6939297124600639,
+      "grad_norm": 0.4339956045150757,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 5973
+    },
+    {
+      "epoch": 0.6940458902120245,
+      "grad_norm": 0.5043813586235046,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 5974
+    },
+    {
+      "epoch": 0.6941620679639849,
+      "grad_norm": 0.4209142029285431,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6942782457159454,
+      "grad_norm": 0.4390365779399872,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 5976
+    },
+    {
+      "epoch": 0.6943944234679059,
+      "grad_norm": 0.4435301423072815,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 5977
+    },
+    {
+      "epoch": 0.6945106012198664,
+      "grad_norm": 0.4381083548069,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 5978
+    },
+    {
+      "epoch": 0.6946267789718269,
+      "grad_norm": 0.4319905638694763,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 5979
+    },
+    {
+      "epoch": 0.6947429567237874,
+      "grad_norm": 0.42140626907348633,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6948591344757479,
+      "grad_norm": 0.46980568766593933,
+      "learning_rate": 0.0001,
+      "loss": 1.8191,
+      "step": 5981
+    },
+    {
+      "epoch": 0.6949753122277084,
+      "grad_norm": 0.44185250997543335,
+      "learning_rate": 0.0001,
+      "loss": 1.6686,
+      "step": 5982
+    },
+    {
+      "epoch": 0.6950914899796689,
+      "grad_norm": 0.4495440423488617,
+      "learning_rate": 0.0001,
+      "loss": 1.7767,
+      "step": 5983
+    },
+    {
+      "epoch": 0.6952076677316293,
+      "grad_norm": 0.43711593747138977,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 5984
+    },
+    {
+      "epoch": 0.6953238454835899,
+      "grad_norm": 0.44120272994041443,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6954400232355504,
+      "grad_norm": 0.46513938903808594,
+      "learning_rate": 0.0001,
+      "loss": 1.7124,
+      "step": 5986
+    },
+    {
+      "epoch": 0.6955562009875109,
+      "grad_norm": 0.43529126048088074,
+      "learning_rate": 0.0001,
+      "loss": 1.548,
+      "step": 5987
+    },
+    {
+      "epoch": 0.6956723787394714,
+      "grad_norm": 0.4164985120296478,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 5988
+    },
+    {
+      "epoch": 0.6957885564914319,
+      "grad_norm": 0.4299083352088928,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 5989
+    },
+    {
+      "epoch": 0.6959047342433924,
+      "grad_norm": 0.4332889914512634,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6960209119953529,
+      "grad_norm": 0.4063509404659271,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 5991
+    },
+    {
+      "epoch": 0.6961370897473134,
+      "grad_norm": 0.4663069248199463,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 5992
+    },
+    {
+      "epoch": 0.6962532674992739,
+      "grad_norm": 0.46274805068969727,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 5993
+    },
+    {
+      "epoch": 0.6963694452512343,
+      "grad_norm": 0.45924481749534607,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 5994
+    },
+    {
+      "epoch": 0.6964856230031949,
+      "grad_norm": 0.4352482557296753,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6966018007551554,
+      "grad_norm": 0.4291403591632843,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 5996
+    },
+    {
+      "epoch": 0.6967179785071159,
+      "grad_norm": 0.5188395977020264,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 5997
+    },
+    {
+      "epoch": 0.6968341562590764,
+      "grad_norm": 0.4158150851726532,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 5998
+    },
+    {
+      "epoch": 0.6969503340110369,
+      "grad_norm": 0.461252897977829,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 5999
+    },
+    {
+      "epoch": 0.6970665117629974,
+      "grad_norm": 0.40928980708122253,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6971826895149579,
+      "grad_norm": 0.4179105758666992,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 6001
+    },
+    {
+      "epoch": 0.6972988672669184,
+      "grad_norm": 0.4326860010623932,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 6002
+    },
+    {
+      "epoch": 0.6974150450188789,
+      "grad_norm": 0.42176535725593567,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 6003
+    },
+    {
+      "epoch": 0.6975312227708393,
+      "grad_norm": 0.4395149350166321,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 6004
+    },
+    {
+      "epoch": 0.6976474005227998,
+      "grad_norm": 0.43788623809814453,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6977635782747604,
+      "grad_norm": 0.4315040111541748,
+      "learning_rate": 0.0001,
+      "loss": 1.6495,
+      "step": 6006
+    },
+    {
+      "epoch": 0.6978797560267209,
+      "grad_norm": 0.43025481700897217,
+      "learning_rate": 0.0001,
+      "loss": 1.6953,
+      "step": 6007
+    },
+    {
+      "epoch": 0.6979959337786814,
+      "grad_norm": 0.4490424692630768,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 6008
+    },
+    {
+      "epoch": 0.6981121115306419,
+      "grad_norm": 0.4578996002674103,
+      "learning_rate": 0.0001,
+      "loss": 1.7515,
+      "step": 6009
+    },
+    {
+      "epoch": 0.6982282892826024,
+      "grad_norm": 0.42355626821517944,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6983444670345629,
+      "grad_norm": 0.4342043399810791,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 6011
+    },
+    {
+      "epoch": 0.6984606447865234,
+      "grad_norm": 0.41684943437576294,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 6012
+    },
+    {
+      "epoch": 0.6985768225384839,
+      "grad_norm": 0.4408121705055237,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 6013
+    },
+    {
+      "epoch": 0.6986930002904443,
+      "grad_norm": 0.43644100427627563,
+      "learning_rate": 0.0001,
+      "loss": 1.4511,
+      "step": 6014
+    },
+    {
+      "epoch": 0.6988091780424048,
+      "grad_norm": 0.4824478328227997,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6989253557943654,
+      "grad_norm": 0.4514261782169342,
+      "learning_rate": 0.0001,
+      "loss": 1.6404,
+      "step": 6016
+    },
+    {
+      "epoch": 0.6990415335463259,
+      "grad_norm": 0.4663759768009186,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 6017
+    },
+    {
+      "epoch": 0.6991577112982864,
+      "grad_norm": 0.45603466033935547,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 6018
+    },
+    {
+      "epoch": 0.6992738890502469,
+      "grad_norm": 0.44485101103782654,
+      "learning_rate": 0.0001,
+      "loss": 1.614,
+      "step": 6019
+    },
+    {
+      "epoch": 0.6993900668022074,
+      "grad_norm": 0.4779071509838104,
+      "learning_rate": 0.0001,
+      "loss": 1.7452,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6995062445541679,
+      "grad_norm": 0.4669143259525299,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 6021
+    },
+    {
+      "epoch": 0.6996224223061284,
+      "grad_norm": 0.4332158863544464,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 6022
+    },
+    {
+      "epoch": 0.6997386000580889,
+      "grad_norm": 0.4396936595439911,
+      "learning_rate": 0.0001,
+      "loss": 1.7863,
+      "step": 6023
+    },
+    {
+      "epoch": 0.6998547778100493,
+      "grad_norm": 0.43175163865089417,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 6024
+    },
+    {
+      "epoch": 0.6999709555620098,
+      "grad_norm": 0.4532018005847931,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 6025
+    },
+    {
+      "epoch": 0.7000871333139703,
+      "grad_norm": 0.4363815188407898,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 6026
+    },
+    {
+      "epoch": 0.7002033110659309,
+      "grad_norm": 0.4686264097690582,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 6027
+    },
+    {
+      "epoch": 0.7003194888178914,
+      "grad_norm": 0.4810371696949005,
+      "learning_rate": 0.0001,
+      "loss": 1.5524,
+      "step": 6028
+    },
+    {
+      "epoch": 0.7004356665698519,
+      "grad_norm": 0.4395377039909363,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 6029
+    },
+    {
+      "epoch": 0.7005518443218124,
+      "grad_norm": 0.43942928314208984,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 6030
+    },
+    {
+      "epoch": 0.7006680220737729,
+      "grad_norm": 0.43220555782318115,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 6031
+    },
+    {
+      "epoch": 0.7007841998257334,
+      "grad_norm": 0.44977232813835144,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 6032
+    },
+    {
+      "epoch": 0.7009003775776939,
+      "grad_norm": 0.4994456171989441,
+      "learning_rate": 0.0001,
+      "loss": 1.729,
+      "step": 6033
+    },
+    {
+      "epoch": 0.7010165553296543,
+      "grad_norm": 0.44498246908187866,
+      "learning_rate": 0.0001,
+      "loss": 1.6094,
+      "step": 6034
+    },
+    {
+      "epoch": 0.7011327330816148,
+      "grad_norm": 0.4909377992153168,
+      "learning_rate": 0.0001,
+      "loss": 1.7201,
+      "step": 6035
+    },
+    {
+      "epoch": 0.7012489108335753,
+      "grad_norm": 0.4470892548561096,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 6036
+    },
+    {
+      "epoch": 0.7013650885855359,
+      "grad_norm": 0.47130686044692993,
+      "learning_rate": 0.0001,
+      "loss": 1.7369,
+      "step": 6037
+    },
+    {
+      "epoch": 0.7014812663374964,
+      "grad_norm": 0.4447838068008423,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 6038
+    },
+    {
+      "epoch": 0.7015974440894569,
+      "grad_norm": 0.48111826181411743,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 6039
+    },
+    {
+      "epoch": 0.7017136218414174,
+      "grad_norm": 0.4512770175933838,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 6040
+    },
+    {
+      "epoch": 0.7018297995933779,
+      "grad_norm": 0.4442228078842163,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 6041
+    },
+    {
+      "epoch": 0.7019459773453384,
+      "grad_norm": 0.49722543358802795,
+      "learning_rate": 0.0001,
+      "loss": 1.7205,
+      "step": 6042
+    },
+    {
+      "epoch": 0.7020621550972989,
+      "grad_norm": 0.4556099474430084,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 6043
+    },
+    {
+      "epoch": 0.7021783328492593,
+      "grad_norm": 0.465631365776062,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 6044
+    },
+    {
+      "epoch": 0.7022945106012198,
+      "grad_norm": 0.4637199938297272,
+      "learning_rate": 0.0001,
+      "loss": 1.7575,
+      "step": 6045
+    },
+    {
+      "epoch": 0.7024106883531803,
+      "grad_norm": 0.5277969241142273,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 6046
+    },
+    {
+      "epoch": 0.7025268661051408,
+      "grad_norm": 0.43430525064468384,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 6047
+    },
+    {
+      "epoch": 0.7026430438571014,
+      "grad_norm": 0.4051744043827057,
+      "learning_rate": 0.0001,
+      "loss": 1.357,
+      "step": 6048
+    },
+    {
+      "epoch": 0.7027592216090619,
+      "grad_norm": 0.4759519696235657,
+      "learning_rate": 0.0001,
+      "loss": 1.6906,
+      "step": 6049
+    },
+    {
+      "epoch": 0.7028753993610224,
+      "grad_norm": 0.4706842005252838,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 6050
+    },
+    {
+      "epoch": 0.7029915771129829,
+      "grad_norm": 0.46256595849990845,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 6051
+    },
+    {
+      "epoch": 0.7031077548649434,
+      "grad_norm": 0.43394848704338074,
+      "learning_rate": 0.0001,
+      "loss": 1.6269,
+      "step": 6052
+    },
+    {
+      "epoch": 0.7032239326169039,
+      "grad_norm": 0.44326069951057434,
+      "learning_rate": 0.0001,
+      "loss": 1.5096,
+      "step": 6053
+    },
+    {
+      "epoch": 0.7033401103688643,
+      "grad_norm": 0.47456949949264526,
+      "learning_rate": 0.0001,
+      "loss": 1.7654,
+      "step": 6054
+    },
+    {
+      "epoch": 0.7034562881208248,
+      "grad_norm": 0.42411813139915466,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 6055
+    },
+    {
+      "epoch": 0.7035724658727853,
+      "grad_norm": 0.43997639417648315,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 6056
+    },
+    {
+      "epoch": 0.7036886436247458,
+      "grad_norm": 0.3923532962799072,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 6057
+    },
+    {
+      "epoch": 0.7038048213767064,
+      "grad_norm": 0.4113970994949341,
+      "learning_rate": 0.0001,
+      "loss": 1.4705,
+      "step": 6058
+    },
+    {
+      "epoch": 0.7039209991286669,
+      "grad_norm": 0.42837658524513245,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 6059
+    },
+    {
+      "epoch": 0.7040371768806274,
+      "grad_norm": 0.44520092010498047,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 6060
+    },
+    {
+      "epoch": 0.7041533546325879,
+      "grad_norm": 0.43117523193359375,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 6061
+    },
+    {
+      "epoch": 0.7042695323845484,
+      "grad_norm": 0.4564533233642578,
+      "learning_rate": 0.0001,
+      "loss": 1.6178,
+      "step": 6062
+    },
+    {
+      "epoch": 0.7043857101365089,
+      "grad_norm": 0.4633900225162506,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 6063
+    },
+    {
+      "epoch": 0.7045018878884693,
+      "grad_norm": 0.439037561416626,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 6064
+    },
+    {
+      "epoch": 0.7046180656404298,
+      "grad_norm": 0.42920568585395813,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 6065
+    },
+    {
+      "epoch": 0.7047342433923903,
+      "grad_norm": 0.44558966159820557,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 6066
+    },
+    {
+      "epoch": 0.7048504211443508,
+      "grad_norm": 0.4488900899887085,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 6067
+    },
+    {
+      "epoch": 0.7049665988963113,
+      "grad_norm": 0.4773916006088257,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 6068
+    },
+    {
+      "epoch": 0.7050827766482719,
+      "grad_norm": 0.4205687344074249,
+      "learning_rate": 0.0001,
+      "loss": 1.5134,
+      "step": 6069
+    },
+    {
+      "epoch": 0.7051989544002324,
+      "grad_norm": 0.4228312373161316,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 6070
+    },
+    {
+      "epoch": 0.7053151321521929,
+      "grad_norm": 0.42489486932754517,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 6071
+    },
+    {
+      "epoch": 0.7054313099041534,
+      "grad_norm": 0.432058185338974,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 6072
+    },
+    {
+      "epoch": 0.7055474876561139,
+      "grad_norm": 0.4359389543533325,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 6073
+    },
+    {
+      "epoch": 0.7056636654080743,
+      "grad_norm": 0.46697720885276794,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 6074
+    },
+    {
+      "epoch": 0.7057798431600348,
+      "grad_norm": 0.4296872317790985,
+      "learning_rate": 0.0001,
+      "loss": 1.6521,
+      "step": 6075
+    },
+    {
+      "epoch": 0.7058960209119953,
+      "grad_norm": 0.4327179491519928,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 6076
+    },
+    {
+      "epoch": 0.7060121986639558,
+      "grad_norm": 0.49642127752304077,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 6077
+    },
+    {
+      "epoch": 0.7061283764159163,
+      "grad_norm": 0.44212982058525085,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 6078
+    },
+    {
+      "epoch": 0.7062445541678769,
+      "grad_norm": 0.45311763882637024,
+      "learning_rate": 0.0001,
+      "loss": 1.6611,
+      "step": 6079
+    },
+    {
+      "epoch": 0.7063607319198374,
+      "grad_norm": 0.4574792981147766,
+      "learning_rate": 0.0001,
+      "loss": 1.5746,
+      "step": 6080
+    },
+    {
+      "epoch": 0.7064769096717979,
+      "grad_norm": 0.4645473062992096,
+      "learning_rate": 0.0001,
+      "loss": 1.5848,
+      "step": 6081
+    },
+    {
+      "epoch": 0.7065930874237584,
+      "grad_norm": 0.4442659318447113,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 6082
+    },
+    {
+      "epoch": 0.7067092651757189,
+      "grad_norm": 0.41226938366889954,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 6083
+    },
+    {
+      "epoch": 0.7068254429276793,
+      "grad_norm": 0.4444557726383209,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 6084
+    },
+    {
+      "epoch": 0.7069416206796398,
+      "grad_norm": 0.4641963839530945,
+      "learning_rate": 0.0001,
+      "loss": 1.4054,
+      "step": 6085
+    },
+    {
+      "epoch": 0.7070577984316003,
+      "grad_norm": 0.44319379329681396,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 6086
+    },
+    {
+      "epoch": 0.7071739761835608,
+      "grad_norm": 0.4771478474140167,
+      "learning_rate": 0.0001,
+      "loss": 1.6775,
+      "step": 6087
+    },
+    {
+      "epoch": 0.7072901539355213,
+      "grad_norm": 0.45707952976226807,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 6088
+    },
+    {
+      "epoch": 0.7074063316874818,
+      "grad_norm": 0.49724724888801575,
+      "learning_rate": 0.0001,
+      "loss": 1.6874,
+      "step": 6089
+    },
+    {
+      "epoch": 0.7075225094394424,
+      "grad_norm": 0.44829389452934265,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 6090
+    },
+    {
+      "epoch": 0.7076386871914029,
+      "grad_norm": 0.4388028681278229,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 6091
+    },
+    {
+      "epoch": 0.7077548649433634,
+      "grad_norm": 0.41982501745224,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 6092
+    },
+    {
+      "epoch": 0.7078710426953239,
+      "grad_norm": 0.4299807846546173,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 6093
+    },
+    {
+      "epoch": 0.7079872204472843,
+      "grad_norm": 0.42945340275764465,
+      "learning_rate": 0.0001,
+      "loss": 1.5398,
+      "step": 6094
+    },
+    {
+      "epoch": 0.7081033981992448,
+      "grad_norm": 0.4422987401485443,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 6095
+    },
+    {
+      "epoch": 0.7082195759512053,
+      "grad_norm": 0.42367589473724365,
+      "learning_rate": 0.0001,
+      "loss": 1.6466,
+      "step": 6096
+    },
+    {
+      "epoch": 0.7083357537031658,
+      "grad_norm": 0.4437691867351532,
+      "learning_rate": 0.0001,
+      "loss": 1.6916,
+      "step": 6097
+    },
+    {
+      "epoch": 0.7084519314551263,
+      "grad_norm": 0.434927374124527,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 6098
+    },
+    {
+      "epoch": 0.7085681092070868,
+      "grad_norm": 0.5107674598693848,
+      "learning_rate": 0.0001,
+      "loss": 1.6533,
+      "step": 6099
+    },
+    {
+      "epoch": 0.7086842869590474,
+      "grad_norm": 0.41166844964027405,
+      "learning_rate": 0.0001,
+      "loss": 1.5646,
+      "step": 6100
+    },
+    {
+      "epoch": 0.7088004647110079,
+      "grad_norm": 0.42547357082366943,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 6101
+    },
+    {
+      "epoch": 0.7089166424629684,
+      "grad_norm": 0.44167038798332214,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 6102
+    },
+    {
+      "epoch": 0.7090328202149289,
+      "grad_norm": 0.4662346839904785,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 6103
+    },
+    {
+      "epoch": 0.7091489979668893,
+      "grad_norm": 0.43853190541267395,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 6104
+    },
+    {
+      "epoch": 0.7092651757188498,
+      "grad_norm": 0.4340886175632477,
+      "learning_rate": 0.0001,
+      "loss": 1.6179,
+      "step": 6105
+    },
+    {
+      "epoch": 0.7093813534708103,
+      "grad_norm": 0.4345012605190277,
+      "learning_rate": 0.0001,
+      "loss": 1.6163,
+      "step": 6106
+    },
+    {
+      "epoch": 0.7094975312227708,
+      "grad_norm": 0.43658679723739624,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 6107
+    },
+    {
+      "epoch": 0.7096137089747313,
+      "grad_norm": 0.42257410287857056,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 6108
+    },
+    {
+      "epoch": 0.7097298867266918,
+      "grad_norm": 0.4617374539375305,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 6109
+    },
+    {
+      "epoch": 0.7098460644786523,
+      "grad_norm": 0.4545523524284363,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 6110
+    },
+    {
+      "epoch": 0.7099622422306129,
+      "grad_norm": 0.4178352355957031,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 6111
+    },
+    {
+      "epoch": 0.7100784199825734,
+      "grad_norm": 0.4775643050670624,
+      "learning_rate": 0.0001,
+      "loss": 1.6867,
+      "step": 6112
+    },
+    {
+      "epoch": 0.7101945977345339,
+      "grad_norm": 0.4341333508491516,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 6113
+    },
+    {
+      "epoch": 0.7103107754864944,
+      "grad_norm": 0.44754558801651,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 6114
+    },
+    {
+      "epoch": 0.7104269532384548,
+      "grad_norm": 0.40256983041763306,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 6115
+    },
+    {
+      "epoch": 0.7105431309904153,
+      "grad_norm": 0.4549466669559479,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 6116
+    },
+    {
+      "epoch": 0.7106593087423758,
+      "grad_norm": 0.45725107192993164,
+      "learning_rate": 0.0001,
+      "loss": 1.5894,
+      "step": 6117
+    },
+    {
+      "epoch": 0.7107754864943363,
+      "grad_norm": 0.4477519094944,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 6118
+    },
+    {
+      "epoch": 0.7108916642462968,
+      "grad_norm": 0.4426574110984802,
+      "learning_rate": 0.0001,
+      "loss": 1.4692,
+      "step": 6119
+    },
+    {
+      "epoch": 0.7110078419982573,
+      "grad_norm": 0.42777371406555176,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 6120
+    },
+    {
+      "epoch": 0.7111240197502179,
+      "grad_norm": 0.43052244186401367,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 6121
+    },
+    {
+      "epoch": 0.7112401975021784,
+      "grad_norm": 0.42970481514930725,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 6122
+    },
+    {
+      "epoch": 0.7113563752541389,
+      "grad_norm": 0.4782750606536865,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 6123
+    },
+    {
+      "epoch": 0.7114725530060994,
+      "grad_norm": 0.45479482412338257,
+      "learning_rate": 0.0001,
+      "loss": 1.6684,
+      "step": 6124
+    },
+    {
+      "epoch": 0.7115887307580598,
+      "grad_norm": 0.43917152285575867,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 6125
+    },
+    {
+      "epoch": 0.7117049085100203,
+      "grad_norm": 0.4511852562427521,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 6126
+    },
+    {
+      "epoch": 0.7118210862619808,
+      "grad_norm": 0.4351733326911926,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 6127
+    },
+    {
+      "epoch": 0.7119372640139413,
+      "grad_norm": 0.458448588848114,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 6128
+    },
+    {
+      "epoch": 0.7120534417659018,
+      "grad_norm": 0.41360458731651306,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 6129
+    },
+    {
+      "epoch": 0.7121696195178623,
+      "grad_norm": 0.45411717891693115,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 6130
+    },
+    {
+      "epoch": 0.7122857972698228,
+      "grad_norm": 0.4399266242980957,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 6131
+    },
+    {
+      "epoch": 0.7124019750217834,
+      "grad_norm": 0.49353474378585815,
+      "learning_rate": 0.0001,
+      "loss": 1.7608,
+      "step": 6132
+    },
+    {
+      "epoch": 0.7125181527737439,
+      "grad_norm": 0.4594828188419342,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 6133
+    },
+    {
+      "epoch": 0.7126343305257044,
+      "grad_norm": 0.4374282658100128,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 6134
+    },
+    {
+      "epoch": 0.7127505082776648,
+      "grad_norm": 0.45439621806144714,
+      "learning_rate": 0.0001,
+      "loss": 1.5785,
+      "step": 6135
+    },
+    {
+      "epoch": 0.7128666860296253,
+      "grad_norm": 0.4368216097354889,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 6136
+    },
+    {
+      "epoch": 0.7129828637815858,
+      "grad_norm": 0.4321901202201843,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 6137
+    },
+    {
+      "epoch": 0.7130990415335463,
+      "grad_norm": 0.4542466402053833,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 6138
+    },
+    {
+      "epoch": 0.7132152192855068,
+      "grad_norm": 0.43192481994628906,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 6139
+    },
+    {
+      "epoch": 0.7133313970374673,
+      "grad_norm": 0.46249476075172424,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 6140
+    },
+    {
+      "epoch": 0.7134475747894278,
+      "grad_norm": 0.4444749653339386,
+      "learning_rate": 0.0001,
+      "loss": 1.7026,
+      "step": 6141
+    },
+    {
+      "epoch": 0.7135637525413884,
+      "grad_norm": 0.4751194417476654,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 6142
+    },
+    {
+      "epoch": 0.7136799302933489,
+      "grad_norm": 0.4836347699165344,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 6143
+    },
+    {
+      "epoch": 0.7137961080453094,
+      "grad_norm": 0.4470170736312866,
+      "learning_rate": 0.0001,
+      "loss": 1.3949,
+      "step": 6144
+    },
+    {
+      "epoch": 0.7139122857972698,
+      "grad_norm": 0.43671542406082153,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 6145
+    },
+    {
+      "epoch": 0.7140284635492303,
+      "grad_norm": 0.4873625636100769,
+      "learning_rate": 0.0001,
+      "loss": 1.7472,
+      "step": 6146
+    },
+    {
+      "epoch": 0.7141446413011908,
+      "grad_norm": 0.43032151460647583,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 6147
+    },
+    {
+      "epoch": 0.7142608190531513,
+      "grad_norm": 0.445576936006546,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 6148
+    },
+    {
+      "epoch": 0.7143769968051118,
+      "grad_norm": 0.450259804725647,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 6149
+    },
+    {
+      "epoch": 0.7144931745570723,
+      "grad_norm": 0.46049565076828003,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 6150
+    },
+    {
+      "epoch": 0.7146093523090328,
+      "grad_norm": 0.44381004571914673,
+      "learning_rate": 0.0001,
+      "loss": 1.5347,
+      "step": 6151
+    },
+    {
+      "epoch": 0.7147255300609934,
+      "grad_norm": 0.44434526562690735,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 6152
+    },
+    {
+      "epoch": 0.7148417078129539,
+      "grad_norm": 0.4273211658000946,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 6153
+    },
+    {
+      "epoch": 0.7149578855649144,
+      "grad_norm": 0.46949854493141174,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 6154
+    },
+    {
+      "epoch": 0.7150740633168748,
+      "grad_norm": 0.40654122829437256,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 6155
+    },
+    {
+      "epoch": 0.7151902410688353,
+      "grad_norm": 0.4765225946903229,
+      "learning_rate": 0.0001,
+      "loss": 1.7118,
+      "step": 6156
+    },
+    {
+      "epoch": 0.7153064188207958,
+      "grad_norm": 0.46587619185447693,
+      "learning_rate": 0.0001,
+      "loss": 1.6022,
+      "step": 6157
+    },
+    {
+      "epoch": 0.7154225965727563,
+      "grad_norm": 0.4613794684410095,
+      "learning_rate": 0.0001,
+      "loss": 1.7496,
+      "step": 6158
+    },
+    {
+      "epoch": 0.7155387743247168,
+      "grad_norm": 0.4671042263507843,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 6159
+    },
+    {
+      "epoch": 0.7156549520766773,
+      "grad_norm": 0.4612120985984802,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 6160
+    },
+    {
+      "epoch": 0.7157711298286378,
+      "grad_norm": 0.4640682339668274,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 6161
+    },
+    {
+      "epoch": 0.7158873075805983,
+      "grad_norm": 0.43429458141326904,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 6162
+    },
+    {
+      "epoch": 0.7160034853325589,
+      "grad_norm": 0.4491255581378937,
+      "learning_rate": 0.0001,
+      "loss": 1.512,
+      "step": 6163
+    },
+    {
+      "epoch": 0.7161196630845194,
+      "grad_norm": 0.39764824509620667,
+      "learning_rate": 0.0001,
+      "loss": 1.344,
+      "step": 6164
+    },
+    {
+      "epoch": 0.7162358408364798,
+      "grad_norm": 0.5114927887916565,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 6165
+    },
+    {
+      "epoch": 0.7163520185884403,
+      "grad_norm": 0.5151812434196472,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 6166
+    },
+    {
+      "epoch": 0.7164681963404008,
+      "grad_norm": 0.43614163994789124,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 6167
+    },
+    {
+      "epoch": 0.7165843740923613,
+      "grad_norm": 0.46181315183639526,
+      "learning_rate": 0.0001,
+      "loss": 1.6494,
+      "step": 6168
+    },
+    {
+      "epoch": 0.7167005518443218,
+      "grad_norm": 0.43157726526260376,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 6169
+    },
+    {
+      "epoch": 0.7168167295962823,
+      "grad_norm": 0.4304739534854889,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 6170
+    },
+    {
+      "epoch": 0.7169329073482428,
+      "grad_norm": 0.4536237120628357,
+      "learning_rate": 0.0001,
+      "loss": 1.7147,
+      "step": 6171
+    },
+    {
+      "epoch": 0.7170490851002033,
+      "grad_norm": 0.40845513343811035,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 6172
+    },
+    {
+      "epoch": 0.7171652628521639,
+      "grad_norm": 0.45678746700286865,
+      "learning_rate": 0.0001,
+      "loss": 1.6398,
+      "step": 6173
+    },
+    {
+      "epoch": 0.7172814406041244,
+      "grad_norm": 0.42773115634918213,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 6174
+    },
+    {
+      "epoch": 0.7173976183560848,
+      "grad_norm": 0.4427812993526459,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 6175
+    },
+    {
+      "epoch": 0.7175137961080453,
+      "grad_norm": 0.42807790637016296,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 6176
+    },
+    {
+      "epoch": 0.7176299738600058,
+      "grad_norm": 0.4445224404335022,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 6177
+    },
+    {
+      "epoch": 0.7177461516119663,
+      "grad_norm": 0.43124210834503174,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 6178
+    },
+    {
+      "epoch": 0.7178623293639268,
+      "grad_norm": 0.41215458512306213,
+      "learning_rate": 0.0001,
+      "loss": 1.6969,
+      "step": 6179
+    },
+    {
+      "epoch": 0.7179785071158873,
+      "grad_norm": 0.45972126722335815,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 6180
+    },
+    {
+      "epoch": 0.7180946848678478,
+      "grad_norm": 0.44836103916168213,
+      "learning_rate": 0.0001,
+      "loss": 1.7967,
+      "step": 6181
+    },
+    {
+      "epoch": 0.7182108626198083,
+      "grad_norm": 0.4521453082561493,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 6182
+    },
+    {
+      "epoch": 0.7183270403717688,
+      "grad_norm": 0.42196735739707947,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 6183
+    },
+    {
+      "epoch": 0.7184432181237294,
+      "grad_norm": 0.4695570170879364,
+      "learning_rate": 0.0001,
+      "loss": 1.6709,
+      "step": 6184
+    },
+    {
+      "epoch": 0.7185593958756898,
+      "grad_norm": 0.4804345369338989,
+      "learning_rate": 0.0001,
+      "loss": 1.7857,
+      "step": 6185
+    },
+    {
+      "epoch": 0.7186755736276503,
+      "grad_norm": 0.41611528396606445,
+      "learning_rate": 0.0001,
+      "loss": 1.5269,
+      "step": 6186
+    },
+    {
+      "epoch": 0.7187917513796108,
+      "grad_norm": 0.4277372658252716,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 6187
+    },
+    {
+      "epoch": 0.7189079291315713,
+      "grad_norm": 0.45039102435112,
+      "learning_rate": 0.0001,
+      "loss": 1.6271,
+      "step": 6188
+    },
+    {
+      "epoch": 0.7190241068835318,
+      "grad_norm": 0.4118036925792694,
+      "learning_rate": 0.0001,
+      "loss": 1.4019,
+      "step": 6189
+    },
+    {
+      "epoch": 0.7191402846354923,
+      "grad_norm": 0.42078375816345215,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 6190
+    },
+    {
+      "epoch": 0.7192564623874528,
+      "grad_norm": 0.4338337779045105,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 6191
+    },
+    {
+      "epoch": 0.7193726401394133,
+      "grad_norm": 0.440143883228302,
+      "learning_rate": 0.0001,
+      "loss": 1.6647,
+      "step": 6192
+    },
+    {
+      "epoch": 0.7194888178913738,
+      "grad_norm": 0.45626693964004517,
+      "learning_rate": 0.0001,
+      "loss": 1.6489,
+      "step": 6193
+    },
+    {
+      "epoch": 0.7196049956433344,
+      "grad_norm": 0.48190397024154663,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 6194
+    },
+    {
+      "epoch": 0.7197211733952948,
+      "grad_norm": 0.42678698897361755,
+      "learning_rate": 0.0001,
+      "loss": 1.6169,
+      "step": 6195
+    },
+    {
+      "epoch": 0.7198373511472553,
+      "grad_norm": 0.48554766178131104,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 6196
+    },
+    {
+      "epoch": 0.7199535288992158,
+      "grad_norm": 0.4570693075656891,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 6197
+    },
+    {
+      "epoch": 0.7200697066511763,
+      "grad_norm": 0.48876166343688965,
+      "learning_rate": 0.0001,
+      "loss": 1.678,
+      "step": 6198
+    },
+    {
+      "epoch": 0.7201858844031368,
+      "grad_norm": 0.4803100526332855,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 6199
+    },
+    {
+      "epoch": 0.7203020621550973,
+      "grad_norm": 0.44159939885139465,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 6200
+    },
+    {
+      "epoch": 0.7204182399070578,
+      "grad_norm": 0.45352402329444885,
+      "learning_rate": 0.0001,
+      "loss": 1.7286,
+      "step": 6201
+    },
+    {
+      "epoch": 0.7205344176590183,
+      "grad_norm": 0.4505062699317932,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 6202
+    },
+    {
+      "epoch": 0.7206505954109788,
+      "grad_norm": 0.4428583085536957,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 6203
+    },
+    {
+      "epoch": 0.7207667731629392,
+      "grad_norm": 0.43122392892837524,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 6204
+    },
+    {
+      "epoch": 0.7208829509148998,
+      "grad_norm": 0.4514487087726593,
+      "learning_rate": 0.0001,
+      "loss": 1.7102,
+      "step": 6205
+    },
+    {
+      "epoch": 0.7209991286668603,
+      "grad_norm": 0.43971630930900574,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 6206
+    },
+    {
+      "epoch": 0.7211153064188208,
+      "grad_norm": 0.424069344997406,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 6207
+    },
+    {
+      "epoch": 0.7212314841707813,
+      "grad_norm": 0.4398737847805023,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 6208
+    },
+    {
+      "epoch": 0.7213476619227418,
+      "grad_norm": 0.4196438491344452,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 6209
+    },
+    {
+      "epoch": 0.7214638396747023,
+      "grad_norm": 0.45629531145095825,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 6210
+    },
+    {
+      "epoch": 0.7215800174266628,
+      "grad_norm": 0.47182023525238037,
+      "learning_rate": 0.0001,
+      "loss": 1.7316,
+      "step": 6211
+    },
+    {
+      "epoch": 0.7216961951786233,
+      "grad_norm": 0.460814893245697,
+      "learning_rate": 0.0001,
+      "loss": 1.6633,
+      "step": 6212
+    },
+    {
+      "epoch": 0.7218123729305838,
+      "grad_norm": 0.41416749358177185,
+      "learning_rate": 0.0001,
+      "loss": 1.3794,
+      "step": 6213
+    },
+    {
+      "epoch": 0.7219285506825442,
+      "grad_norm": 0.4730556607246399,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 6214
+    },
+    {
+      "epoch": 0.7220447284345048,
+      "grad_norm": 0.4644532799720764,
+      "learning_rate": 0.0001,
+      "loss": 1.5189,
+      "step": 6215
+    },
+    {
+      "epoch": 0.7221609061864653,
+      "grad_norm": 0.44122570753097534,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 6216
+    },
+    {
+      "epoch": 0.7222770839384258,
+      "grad_norm": 0.46849554777145386,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 6217
+    },
+    {
+      "epoch": 0.7223932616903863,
+      "grad_norm": 0.43822774291038513,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 6218
+    },
+    {
+      "epoch": 0.7225094394423468,
+      "grad_norm": 0.42008358240127563,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 6219
+    },
+    {
+      "epoch": 0.7226256171943073,
+      "grad_norm": 0.45318713784217834,
+      "learning_rate": 0.0001,
+      "loss": 1.6414,
+      "step": 6220
+    },
+    {
+      "epoch": 0.7227417949462678,
+      "grad_norm": 0.4390380084514618,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 6221
+    },
+    {
+      "epoch": 0.7228579726982283,
+      "grad_norm": 0.46329495310783386,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 6222
+    },
+    {
+      "epoch": 0.7229741504501888,
+      "grad_norm": 0.47496578097343445,
+      "learning_rate": 0.0001,
+      "loss": 1.6651,
+      "step": 6223
+    },
+    {
+      "epoch": 0.7230903282021492,
+      "grad_norm": 0.46759724617004395,
+      "learning_rate": 0.0001,
+      "loss": 1.7158,
+      "step": 6224
+    },
+    {
+      "epoch": 0.7232065059541097,
+      "grad_norm": 0.46540212631225586,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 6225
+    },
+    {
+      "epoch": 0.7233226837060703,
+      "grad_norm": 0.4282924234867096,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 6226
+    },
+    {
+      "epoch": 0.7234388614580308,
+      "grad_norm": 0.44697490334510803,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 6227
+    },
+    {
+      "epoch": 0.7235550392099913,
+      "grad_norm": 0.4610814154148102,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 6228
+    },
+    {
+      "epoch": 0.7236712169619518,
+      "grad_norm": 0.45639610290527344,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 6229
+    },
+    {
+      "epoch": 0.7237873947139123,
+      "grad_norm": 0.4381502568721771,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 6230
+    },
+    {
+      "epoch": 0.7239035724658728,
+      "grad_norm": 0.440270334482193,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 6231
+    },
+    {
+      "epoch": 0.7240197502178333,
+      "grad_norm": 0.4624209403991699,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 6232
+    },
+    {
+      "epoch": 0.7241359279697938,
+      "grad_norm": 0.4678003489971161,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 6233
+    },
+    {
+      "epoch": 0.7242521057217542,
+      "grad_norm": 0.44897353649139404,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 6234
+    },
+    {
+      "epoch": 0.7243682834737147,
+      "grad_norm": 0.431941419839859,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 6235
+    },
+    {
+      "epoch": 0.7244844612256753,
+      "grad_norm": 0.45239877700805664,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 6236
+    },
+    {
+      "epoch": 0.7246006389776358,
+      "grad_norm": 0.4519427418708801,
+      "learning_rate": 0.0001,
+      "loss": 1.7284,
+      "step": 6237
+    },
+    {
+      "epoch": 0.7247168167295963,
+      "grad_norm": 0.45957690477371216,
+      "learning_rate": 0.0001,
+      "loss": 1.5839,
+      "step": 6238
+    },
+    {
+      "epoch": 0.7248329944815568,
+      "grad_norm": 0.46946290135383606,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 6239
+    },
+    {
+      "epoch": 0.7249491722335173,
+      "grad_norm": 0.4527909457683563,
+      "learning_rate": 0.0001,
+      "loss": 1.6597,
+      "step": 6240
+    },
+    {
+      "epoch": 0.7250653499854778,
+      "grad_norm": 0.4612141251564026,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 6241
+    },
+    {
+      "epoch": 0.7251815277374383,
+      "grad_norm": 0.4190555214881897,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 6242
+    },
+    {
+      "epoch": 0.7252977054893988,
+      "grad_norm": 0.4483005702495575,
+      "learning_rate": 0.0001,
+      "loss": 1.5936,
+      "step": 6243
+    },
+    {
+      "epoch": 0.7254138832413592,
+      "grad_norm": 0.4509059190750122,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 6244
+    },
+    {
+      "epoch": 0.7255300609933197,
+      "grad_norm": 0.47739991545677185,
+      "learning_rate": 0.0001,
+      "loss": 1.759,
+      "step": 6245
+    },
+    {
+      "epoch": 0.7256462387452802,
+      "grad_norm": 0.4522044360637665,
+      "learning_rate": 0.0001,
+      "loss": 1.6458,
+      "step": 6246
+    },
+    {
+      "epoch": 0.7257624164972408,
+      "grad_norm": 0.43035435676574707,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 6247
+    },
+    {
+      "epoch": 0.7258785942492013,
+      "grad_norm": 0.4420571029186249,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 6248
+    },
+    {
+      "epoch": 0.7259947720011618,
+      "grad_norm": 0.47003141045570374,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 6249
+    },
+    {
+      "epoch": 0.7261109497531223,
+      "grad_norm": 0.43513643741607666,
+      "learning_rate": 0.0001,
+      "loss": 1.7224,
+      "step": 6250
+    },
+    {
+      "epoch": 0.7262271275050828,
+      "grad_norm": 0.4268497824668884,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 6251
+    },
+    {
+      "epoch": 0.7263433052570433,
+      "grad_norm": 0.46211907267570496,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 6252
+    },
+    {
+      "epoch": 0.7264594830090038,
+      "grad_norm": 0.4459565281867981,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 6253
+    },
+    {
+      "epoch": 0.7265756607609642,
+      "grad_norm": 0.44835829734802246,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 6254
+    },
+    {
+      "epoch": 0.7266918385129247,
+      "grad_norm": 0.4384959936141968,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 6255
+    },
+    {
+      "epoch": 0.7268080162648852,
+      "grad_norm": 0.4654468894004822,
+      "learning_rate": 0.0001,
+      "loss": 1.6979,
+      "step": 6256
+    },
+    {
+      "epoch": 0.7269241940168458,
+      "grad_norm": 0.4371379613876343,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 6257
+    },
+    {
+      "epoch": 0.7270403717688063,
+      "grad_norm": 0.48108625411987305,
+      "learning_rate": 0.0001,
+      "loss": 1.7885,
+      "step": 6258
+    },
+    {
+      "epoch": 0.7271565495207668,
+      "grad_norm": 0.4787601828575134,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 6259
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.43150654435157776,
+      "learning_rate": 0.0001,
+      "loss": 1.5843,
+      "step": 6260
+    },
+    {
+      "epoch": 0.7273889050246878,
+      "grad_norm": 0.46279093623161316,
+      "learning_rate": 0.0001,
+      "loss": 1.6501,
+      "step": 6261
+    },
+    {
+      "epoch": 0.7275050827766483,
+      "grad_norm": 0.4516332447528839,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 6262
+    },
+    {
+      "epoch": 0.7276212605286088,
+      "grad_norm": 0.43989452719688416,
+      "learning_rate": 0.0001,
+      "loss": 1.6061,
+      "step": 6263
+    },
+    {
+      "epoch": 0.7277374382805692,
+      "grad_norm": 0.44632989168167114,
+      "learning_rate": 0.0001,
+      "loss": 1.5898,
+      "step": 6264
+    },
+    {
+      "epoch": 0.7278536160325297,
+      "grad_norm": 0.44944363832473755,
+      "learning_rate": 0.0001,
+      "loss": 1.6132,
+      "step": 6265
+    },
+    {
+      "epoch": 0.7279697937844902,
+      "grad_norm": 0.4271278977394104,
+      "learning_rate": 0.0001,
+      "loss": 1.48,
+      "step": 6266
+    },
+    {
+      "epoch": 0.7280859715364507,
+      "grad_norm": 0.4227433204650879,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 6267
+    },
+    {
+      "epoch": 0.7282021492884113,
+      "grad_norm": 0.42294958233833313,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 6268
+    },
+    {
+      "epoch": 0.7283183270403718,
+      "grad_norm": 0.43251901865005493,
+      "learning_rate": 0.0001,
+      "loss": 1.576,
+      "step": 6269
+    },
+    {
+      "epoch": 0.7284345047923323,
+      "grad_norm": 0.47049641609191895,
+      "learning_rate": 0.0001,
+      "loss": 1.7649,
+      "step": 6270
+    },
+    {
+      "epoch": 0.7285506825442928,
+      "grad_norm": 0.457476943731308,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 6271
+    },
+    {
+      "epoch": 0.7286668602962533,
+      "grad_norm": 0.4726758599281311,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 6272
+    },
+    {
+      "epoch": 0.7287830380482138,
+      "grad_norm": 0.5140420198440552,
+      "learning_rate": 0.0001,
+      "loss": 1.8907,
+      "step": 6273
+    },
+    {
+      "epoch": 0.7288992158001742,
+      "grad_norm": 0.4506530165672302,
+      "learning_rate": 0.0001,
+      "loss": 1.6638,
+      "step": 6274
+    },
+    {
+      "epoch": 0.7290153935521347,
+      "grad_norm": 0.4550815522670746,
+      "learning_rate": 0.0001,
+      "loss": 1.7825,
+      "step": 6275
+    },
+    {
+      "epoch": 0.7291315713040952,
+      "grad_norm": 0.45727846026420593,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 6276
+    },
+    {
+      "epoch": 0.7292477490560557,
+      "grad_norm": 0.4206952154636383,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 6277
+    },
+    {
+      "epoch": 0.7293639268080163,
+      "grad_norm": 0.4411332905292511,
+      "learning_rate": 0.0001,
+      "loss": 1.7024,
+      "step": 6278
+    },
+    {
+      "epoch": 0.7294801045599768,
+      "grad_norm": 0.4896695613861084,
+      "learning_rate": 0.0001,
+      "loss": 1.7724,
+      "step": 6279
+    },
+    {
+      "epoch": 0.7295962823119373,
+      "grad_norm": 0.463642954826355,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 6280
+    },
+    {
+      "epoch": 0.7297124600638978,
+      "grad_norm": 0.41118043661117554,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 6281
+    },
+    {
+      "epoch": 0.7298286378158583,
+      "grad_norm": 0.4534677565097809,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 6282
+    },
+    {
+      "epoch": 0.7299448155678188,
+      "grad_norm": 0.4662560522556305,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 6283
+    },
+    {
+      "epoch": 0.7300609933197793,
+      "grad_norm": 0.4774499833583832,
+      "learning_rate": 0.0001,
+      "loss": 1.8409,
+      "step": 6284
+    },
+    {
+      "epoch": 0.7301771710717397,
+      "grad_norm": 0.45723921060562134,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 6285
+    },
+    {
+      "epoch": 0.7302933488237002,
+      "grad_norm": 0.44660598039627075,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 6286
+    },
+    {
+      "epoch": 0.7304095265756607,
+      "grad_norm": 0.44800204038619995,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 6287
+    },
+    {
+      "epoch": 0.7305257043276212,
+      "grad_norm": 0.45778778195381165,
+      "learning_rate": 0.0001,
+      "loss": 1.5414,
+      "step": 6288
+    },
+    {
+      "epoch": 0.7306418820795818,
+      "grad_norm": 0.431100457906723,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 6289
+    },
+    {
+      "epoch": 0.7307580598315423,
+      "grad_norm": 0.4813309907913208,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 6290
+    },
+    {
+      "epoch": 0.7308742375835028,
+      "grad_norm": 0.4887335002422333,
+      "learning_rate": 0.0001,
+      "loss": 1.7038,
+      "step": 6291
+    },
+    {
+      "epoch": 0.7309904153354633,
+      "grad_norm": 0.48292210698127747,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 6292
+    },
+    {
+      "epoch": 0.7311065930874238,
+      "grad_norm": 0.42640790343284607,
+      "learning_rate": 0.0001,
+      "loss": 1.5194,
+      "step": 6293
+    },
+    {
+      "epoch": 0.7312227708393843,
+      "grad_norm": 0.41172870993614197,
+      "learning_rate": 0.0001,
+      "loss": 1.4218,
+      "step": 6294
+    },
+    {
+      "epoch": 0.7313389485913447,
+      "grad_norm": 0.48400574922561646,
+      "learning_rate": 0.0001,
+      "loss": 1.6952,
+      "step": 6295
+    },
+    {
+      "epoch": 0.7314551263433052,
+      "grad_norm": 0.47430482506752014,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 6296
+    },
+    {
+      "epoch": 0.7315713040952657,
+      "grad_norm": 0.4179590046405792,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 6297
+    },
+    {
+      "epoch": 0.7316874818472262,
+      "grad_norm": 0.4300008714199066,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 6298
+    },
+    {
+      "epoch": 0.7318036595991868,
+      "grad_norm": 0.4634290933609009,
+      "learning_rate": 0.0001,
+      "loss": 1.7461,
+      "step": 6299
+    },
+    {
+      "epoch": 0.7319198373511473,
+      "grad_norm": 0.4361773133277893,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 6300
+    },
+    {
+      "epoch": 0.7320360151031078,
+      "grad_norm": 0.46928146481513977,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 6301
+    },
+    {
+      "epoch": 0.7321521928550683,
+      "grad_norm": 0.47918713092803955,
+      "learning_rate": 0.0001,
+      "loss": 1.7658,
+      "step": 6302
+    },
+    {
+      "epoch": 0.7322683706070288,
+      "grad_norm": 0.49823591113090515,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 6303
+    },
+    {
+      "epoch": 0.7323845483589893,
+      "grad_norm": 0.5099037289619446,
+      "learning_rate": 0.0001,
+      "loss": 1.8295,
+      "step": 6304
+    },
+    {
+      "epoch": 0.7325007261109497,
+      "grad_norm": 0.49155983328819275,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 6305
+    },
+    {
+      "epoch": 0.7326169038629102,
+      "grad_norm": 0.4120127260684967,
+      "learning_rate": 0.0001,
+      "loss": 1.3311,
+      "step": 6306
+    },
+    {
+      "epoch": 0.7327330816148707,
+      "grad_norm": 0.5385419726371765,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 6307
+    },
+    {
+      "epoch": 0.7328492593668312,
+      "grad_norm": 0.4356537461280823,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 6308
+    },
+    {
+      "epoch": 0.7329654371187917,
+      "grad_norm": 0.46573251485824585,
+      "learning_rate": 0.0001,
+      "loss": 1.6368,
+      "step": 6309
+    },
+    {
+      "epoch": 0.7330816148707523,
+      "grad_norm": 0.42254379391670227,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 6310
+    },
+    {
+      "epoch": 0.7331977926227128,
+      "grad_norm": 0.42466604709625244,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 6311
+    },
+    {
+      "epoch": 0.7333139703746733,
+      "grad_norm": 0.4290546774864197,
+      "learning_rate": 0.0001,
+      "loss": 1.3726,
+      "step": 6312
+    },
+    {
+      "epoch": 0.7334301481266338,
+      "grad_norm": 0.4976603090763092,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 6313
+    },
+    {
+      "epoch": 0.7335463258785943,
+      "grad_norm": 0.457592248916626,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 6314
+    },
+    {
+      "epoch": 0.7336625036305547,
+      "grad_norm": 0.4522661864757538,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 6315
+    },
+    {
+      "epoch": 0.7337786813825152,
+      "grad_norm": 0.43026474118232727,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 6316
+    },
+    {
+      "epoch": 0.7338948591344757,
+      "grad_norm": 0.41863980889320374,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 6317
+    },
+    {
+      "epoch": 0.7340110368864362,
+      "grad_norm": 0.47513189911842346,
+      "learning_rate": 0.0001,
+      "loss": 1.7717,
+      "step": 6318
+    },
+    {
+      "epoch": 0.7341272146383967,
+      "grad_norm": 0.46296703815460205,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 6319
+    },
+    {
+      "epoch": 0.7342433923903573,
+      "grad_norm": 0.40901249647140503,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 6320
+    },
+    {
+      "epoch": 0.7343595701423178,
+      "grad_norm": 0.48210811614990234,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 6321
+    },
+    {
+      "epoch": 0.7344757478942783,
+      "grad_norm": 0.4578268826007843,
+      "learning_rate": 0.0001,
+      "loss": 1.5555,
+      "step": 6322
+    },
+    {
+      "epoch": 0.7345919256462388,
+      "grad_norm": 0.5248122811317444,
+      "learning_rate": 0.0001,
+      "loss": 1.7845,
+      "step": 6323
+    },
+    {
+      "epoch": 0.7347081033981993,
+      "grad_norm": 0.4604836404323578,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 6324
+    },
+    {
+      "epoch": 0.7348242811501597,
+      "grad_norm": 0.4478675425052643,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 6325
+    },
+    {
+      "epoch": 0.7349404589021202,
+      "grad_norm": 0.4665616452693939,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 6326
+    },
+    {
+      "epoch": 0.7350566366540807,
+      "grad_norm": 0.4462968409061432,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 6327
+    },
+    {
+      "epoch": 0.7351728144060412,
+      "grad_norm": 0.4639495313167572,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 6328
+    },
+    {
+      "epoch": 0.7352889921580017,
+      "grad_norm": 0.47614338994026184,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 6329
+    },
+    {
+      "epoch": 0.7354051699099623,
+      "grad_norm": 0.47251033782958984,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 6330
+    },
+    {
+      "epoch": 0.7355213476619228,
+      "grad_norm": 0.4667409062385559,
+      "learning_rate": 0.0001,
+      "loss": 1.6973,
+      "step": 6331
+    },
+    {
+      "epoch": 0.7356375254138833,
+      "grad_norm": 0.43964049220085144,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 6332
+    },
+    {
+      "epoch": 0.7357537031658438,
+      "grad_norm": 0.43899863958358765,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 6333
+    },
+    {
+      "epoch": 0.7358698809178043,
+      "grad_norm": 0.4609370231628418,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 6334
+    },
+    {
+      "epoch": 0.7359860586697647,
+      "grad_norm": 0.4471205472946167,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 6335
+    },
+    {
+      "epoch": 0.7361022364217252,
+      "grad_norm": 0.4316443204879761,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 6336
+    },
+    {
+      "epoch": 0.7362184141736857,
+      "grad_norm": 0.4466089606285095,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 6337
+    },
+    {
+      "epoch": 0.7363345919256462,
+      "grad_norm": 0.4607985317707062,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 6338
+    },
+    {
+      "epoch": 0.7364507696776067,
+      "grad_norm": 0.46263134479522705,
+      "learning_rate": 0.0001,
+      "loss": 1.7685,
+      "step": 6339
+    },
+    {
+      "epoch": 0.7365669474295672,
+      "grad_norm": 0.4616027772426605,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7366831251815278,
+      "grad_norm": 0.4539996385574341,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 6341
+    },
+    {
+      "epoch": 0.7367993029334883,
+      "grad_norm": 0.4599131643772125,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 6342
+    },
+    {
+      "epoch": 0.7369154806854488,
+      "grad_norm": 0.4772374927997589,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 6343
+    },
+    {
+      "epoch": 0.7370316584374093,
+      "grad_norm": 0.4597277343273163,
+      "learning_rate": 0.0001,
+      "loss": 1.6406,
+      "step": 6344
+    },
+    {
+      "epoch": 0.7371478361893697,
+      "grad_norm": 0.4844377934932709,
+      "learning_rate": 0.0001,
+      "loss": 1.6903,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7372640139413302,
+      "grad_norm": 0.4434802532196045,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 6346
+    },
+    {
+      "epoch": 0.7373801916932907,
+      "grad_norm": 0.48957550525665283,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 6347
+    },
+    {
+      "epoch": 0.7374963694452512,
+      "grad_norm": 0.4729391932487488,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 6348
+    },
+    {
+      "epoch": 0.7376125471972117,
+      "grad_norm": 0.4368651807308197,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 6349
+    },
+    {
+      "epoch": 0.7377287249491722,
+      "grad_norm": 0.4742048978805542,
+      "learning_rate": 0.0001,
+      "loss": 1.8506,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7378449027011328,
+      "grad_norm": 0.4797591269016266,
+      "learning_rate": 0.0001,
+      "loss": 1.6867,
+      "step": 6351
+    },
+    {
+      "epoch": 0.7379610804530933,
+      "grad_norm": 0.46037012338638306,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 6352
+    },
+    {
+      "epoch": 0.7380772582050538,
+      "grad_norm": 0.45086851716041565,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 6353
+    },
+    {
+      "epoch": 0.7381934359570143,
+      "grad_norm": 0.4557948410511017,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 6354
+    },
+    {
+      "epoch": 0.7383096137089747,
+      "grad_norm": 0.4523628354072571,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 6355
+    },
+    {
+      "epoch": 0.7384257914609352,
+      "grad_norm": 0.43781834840774536,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 6356
+    },
+    {
+      "epoch": 0.7385419692128957,
+      "grad_norm": 0.5110533833503723,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 6357
+    },
+    {
+      "epoch": 0.7386581469648562,
+      "grad_norm": 0.4702020585536957,
+      "learning_rate": 0.0001,
+      "loss": 1.5689,
+      "step": 6358
+    },
+    {
+      "epoch": 0.7387743247168167,
+      "grad_norm": 0.45719245076179504,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 6359
+    },
+    {
+      "epoch": 0.7388905024687772,
+      "grad_norm": 0.45897620916366577,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7390066802207377,
+      "grad_norm": 0.48604246973991394,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 6361
+    },
+    {
+      "epoch": 0.7391228579726983,
+      "grad_norm": 0.4290804862976074,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 6362
+    },
+    {
+      "epoch": 0.7392390357246588,
+      "grad_norm": 0.47816741466522217,
+      "learning_rate": 0.0001,
+      "loss": 1.5016,
+      "step": 6363
+    },
+    {
+      "epoch": 0.7393552134766193,
+      "grad_norm": 0.5012046098709106,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 6364
+    },
+    {
+      "epoch": 0.7394713912285797,
+      "grad_norm": 0.5031648874282837,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7395875689805402,
+      "grad_norm": 0.46205514669418335,
+      "learning_rate": 0.0001,
+      "loss": 1.7046,
+      "step": 6366
+    },
+    {
+      "epoch": 0.7397037467325007,
+      "grad_norm": 0.4918919801712036,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 6367
+    },
+    {
+      "epoch": 0.7398199244844612,
+      "grad_norm": 0.4417077898979187,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 6368
+    },
+    {
+      "epoch": 0.7399361022364217,
+      "grad_norm": 0.5084665417671204,
+      "learning_rate": 0.0001,
+      "loss": 1.8225,
+      "step": 6369
+    },
+    {
+      "epoch": 0.7400522799883822,
+      "grad_norm": 0.4046757221221924,
+      "learning_rate": 0.0001,
+      "loss": 1.3748,
+      "step": 6370
+    },
+    {
+      "epoch": 0.7401684577403427,
+      "grad_norm": 0.4671691358089447,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 6371
+    },
+    {
+      "epoch": 0.7402846354923033,
+      "grad_norm": 0.41901758313179016,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 6372
+    },
+    {
+      "epoch": 0.7404008132442638,
+      "grad_norm": 0.4258348345756531,
+      "learning_rate": 0.0001,
+      "loss": 1.4846,
+      "step": 6373
+    },
+    {
+      "epoch": 0.7405169909962243,
+      "grad_norm": 0.44732606410980225,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 6374
+    },
+    {
+      "epoch": 0.7406331687481847,
+      "grad_norm": 0.4379178583621979,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 6375
+    },
+    {
+      "epoch": 0.7407493465001452,
+      "grad_norm": 0.44309452176094055,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 6376
+    },
+    {
+      "epoch": 0.7408655242521057,
+      "grad_norm": 0.47775688767433167,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 6377
+    },
+    {
+      "epoch": 0.7409817020040662,
+      "grad_norm": 0.4426514208316803,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 6378
+    },
+    {
+      "epoch": 0.7410978797560267,
+      "grad_norm": 0.45742908120155334,
+      "learning_rate": 0.0001,
+      "loss": 1.5815,
+      "step": 6379
+    },
+    {
+      "epoch": 0.7412140575079872,
+      "grad_norm": 0.46088162064552307,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 6380
+    },
+    {
+      "epoch": 0.7413302352599477,
+      "grad_norm": 0.4453122317790985,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 6381
+    },
+    {
+      "epoch": 0.7414464130119082,
+      "grad_norm": 0.4657224714756012,
+      "learning_rate": 0.0001,
+      "loss": 1.7709,
+      "step": 6382
+    },
+    {
+      "epoch": 0.7415625907638688,
+      "grad_norm": 0.47661951184272766,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 6383
+    },
+    {
+      "epoch": 0.7416787685158293,
+      "grad_norm": 0.4470673203468323,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 6384
+    },
+    {
+      "epoch": 0.7417949462677897,
+      "grad_norm": 0.4273890554904938,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 6385
+    },
+    {
+      "epoch": 0.7419111240197502,
+      "grad_norm": 0.4307868778705597,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 6386
+    },
+    {
+      "epoch": 0.7420273017717107,
+      "grad_norm": 0.5137743353843689,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 6387
+    },
+    {
+      "epoch": 0.7421434795236712,
+      "grad_norm": 0.4485335648059845,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 6388
+    },
+    {
+      "epoch": 0.7422596572756317,
+      "grad_norm": 0.46610766649246216,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 6389
+    },
+    {
+      "epoch": 0.7423758350275922,
+      "grad_norm": 0.4473322629928589,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 6390
+    },
+    {
+      "epoch": 0.7424920127795527,
+      "grad_norm": 0.4895930886268616,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 6391
+    },
+    {
+      "epoch": 0.7426081905315132,
+      "grad_norm": 0.48194587230682373,
+      "learning_rate": 0.0001,
+      "loss": 1.6737,
+      "step": 6392
+    },
+    {
+      "epoch": 0.7427243682834738,
+      "grad_norm": 0.44431936740875244,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 6393
+    },
+    {
+      "epoch": 0.7428405460354343,
+      "grad_norm": 0.4284924864768982,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 6394
+    },
+    {
+      "epoch": 0.7429567237873947,
+      "grad_norm": 0.4462593197822571,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 6395
+    },
+    {
+      "epoch": 0.7430729015393552,
+      "grad_norm": 0.4361656606197357,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 6396
+    },
+    {
+      "epoch": 0.7431890792913157,
+      "grad_norm": 0.4379312992095947,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 6397
+    },
+    {
+      "epoch": 0.7433052570432762,
+      "grad_norm": 0.4425080418586731,
+      "learning_rate": 0.0001,
+      "loss": 1.4492,
+      "step": 6398
+    },
+    {
+      "epoch": 0.7434214347952367,
+      "grad_norm": 0.4721614718437195,
+      "learning_rate": 0.0001,
+      "loss": 1.7602,
+      "step": 6399
+    },
+    {
+      "epoch": 0.7435376125471972,
+      "grad_norm": 0.45762506127357483,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 6400
+    },
+    {
+      "epoch": 0.7436537902991577,
+      "grad_norm": 0.44420626759529114,
+      "learning_rate": 0.0001,
+      "loss": 1.4694,
+      "step": 6401
+    },
+    {
+      "epoch": 0.7437699680511182,
+      "grad_norm": 0.4328571856021881,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 6402
+    },
+    {
+      "epoch": 0.7438861458030787,
+      "grad_norm": 0.4925638735294342,
+      "learning_rate": 0.0001,
+      "loss": 1.731,
+      "step": 6403
+    },
+    {
+      "epoch": 0.7440023235550393,
+      "grad_norm": 0.43524104356765747,
+      "learning_rate": 0.0001,
+      "loss": 1.6355,
+      "step": 6404
+    },
+    {
+      "epoch": 0.7441185013069997,
+      "grad_norm": 0.4498136043548584,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 6405
+    },
+    {
+      "epoch": 0.7442346790589602,
+      "grad_norm": 0.43513327836990356,
+      "learning_rate": 0.0001,
+      "loss": 1.3918,
+      "step": 6406
+    },
+    {
+      "epoch": 0.7443508568109207,
+      "grad_norm": 0.4613160789012909,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 6407
+    },
+    {
+      "epoch": 0.7444670345628812,
+      "grad_norm": 0.4584994614124298,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 6408
+    },
+    {
+      "epoch": 0.7445832123148417,
+      "grad_norm": 0.49195122718811035,
+      "learning_rate": 0.0001,
+      "loss": 1.8131,
+      "step": 6409
+    },
+    {
+      "epoch": 0.7446993900668022,
+      "grad_norm": 0.4556443393230438,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 6410
+    },
+    {
+      "epoch": 0.7448155678187627,
+      "grad_norm": 0.46693944931030273,
+      "learning_rate": 0.0001,
+      "loss": 1.719,
+      "step": 6411
+    },
+    {
+      "epoch": 0.7449317455707232,
+      "grad_norm": 0.46386128664016724,
+      "learning_rate": 0.0001,
+      "loss": 1.6447,
+      "step": 6412
+    },
+    {
+      "epoch": 0.7450479233226837,
+      "grad_norm": 0.43832361698150635,
+      "learning_rate": 0.0001,
+      "loss": 1.5779,
+      "step": 6413
+    },
+    {
+      "epoch": 0.7451641010746443,
+      "grad_norm": 0.4549694359302521,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 6414
+    },
+    {
+      "epoch": 0.7452802788266047,
+      "grad_norm": 0.44481566548347473,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 6415
+    },
+    {
+      "epoch": 0.7453964565785652,
+      "grad_norm": 0.43813270330429077,
+      "learning_rate": 0.0001,
+      "loss": 1.6835,
+      "step": 6416
+    },
+    {
+      "epoch": 0.7455126343305257,
+      "grad_norm": 0.44562163949012756,
+      "learning_rate": 0.0001,
+      "loss": 1.6412,
+      "step": 6417
+    },
+    {
+      "epoch": 0.7456288120824862,
+      "grad_norm": 0.4208749532699585,
+      "learning_rate": 0.0001,
+      "loss": 1.4426,
+      "step": 6418
+    },
+    {
+      "epoch": 0.7457449898344467,
+      "grad_norm": 0.4573444426059723,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 6419
+    },
+    {
+      "epoch": 0.7458611675864072,
+      "grad_norm": 0.4514586329460144,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 6420
+    },
+    {
+      "epoch": 0.7459773453383677,
+      "grad_norm": 0.427177757024765,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 6421
+    },
+    {
+      "epoch": 0.7460935230903282,
+      "grad_norm": 0.43251994252204895,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 6422
+    },
+    {
+      "epoch": 0.7462097008422887,
+      "grad_norm": 0.4547024369239807,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 6423
+    },
+    {
+      "epoch": 0.7463258785942491,
+      "grad_norm": 0.454458087682724,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 6424
+    },
+    {
+      "epoch": 0.7464420563462097,
+      "grad_norm": 0.44266244769096375,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 6425
+    },
+    {
+      "epoch": 0.7465582340981702,
+      "grad_norm": 0.4643990099430084,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 6426
+    },
+    {
+      "epoch": 0.7466744118501307,
+      "grad_norm": 0.4686664938926697,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 6427
+    },
+    {
+      "epoch": 0.7467905896020912,
+      "grad_norm": 0.49029332399368286,
+      "learning_rate": 0.0001,
+      "loss": 1.5791,
+      "step": 6428
+    },
+    {
+      "epoch": 0.7469067673540517,
+      "grad_norm": 0.46670180559158325,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 6429
+    },
+    {
+      "epoch": 0.7470229451060122,
+      "grad_norm": 0.44535404443740845,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 6430
+    },
+    {
+      "epoch": 0.7471391228579727,
+      "grad_norm": 0.4496926963329315,
+      "learning_rate": 0.0001,
+      "loss": 1.4151,
+      "step": 6431
+    },
+    {
+      "epoch": 0.7472553006099332,
+      "grad_norm": 0.4497212767601013,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 6432
+    },
+    {
+      "epoch": 0.7473714783618937,
+      "grad_norm": 0.44064804911613464,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 6433
+    },
+    {
+      "epoch": 0.7474876561138541,
+      "grad_norm": 0.459143728017807,
+      "learning_rate": 0.0001,
+      "loss": 1.6105,
+      "step": 6434
+    },
+    {
+      "epoch": 0.7476038338658147,
+      "grad_norm": 0.4302321672439575,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 6435
+    },
+    {
+      "epoch": 0.7477200116177752,
+      "grad_norm": 0.4803028702735901,
+      "learning_rate": 0.0001,
+      "loss": 1.6813,
+      "step": 6436
+    },
+    {
+      "epoch": 0.7478361893697357,
+      "grad_norm": 0.41635861992836,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 6437
+    },
+    {
+      "epoch": 0.7479523671216962,
+      "grad_norm": 0.44405534863471985,
+      "learning_rate": 0.0001,
+      "loss": 1.6584,
+      "step": 6438
+    },
+    {
+      "epoch": 0.7480685448736567,
+      "grad_norm": 0.4449010491371155,
+      "learning_rate": 0.0001,
+      "loss": 1.7648,
+      "step": 6439
+    },
+    {
+      "epoch": 0.7481847226256172,
+      "grad_norm": 0.45254096388816833,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 6440
+    },
+    {
+      "epoch": 0.7483009003775777,
+      "grad_norm": 0.4546097218990326,
+      "learning_rate": 0.0001,
+      "loss": 1.6735,
+      "step": 6441
+    },
+    {
+      "epoch": 0.7484170781295382,
+      "grad_norm": 0.44353100657463074,
+      "learning_rate": 0.0001,
+      "loss": 1.6668,
+      "step": 6442
+    },
+    {
+      "epoch": 0.7485332558814987,
+      "grad_norm": 0.4421907365322113,
+      "learning_rate": 0.0001,
+      "loss": 1.6233,
+      "step": 6443
+    },
+    {
+      "epoch": 0.7486494336334591,
+      "grad_norm": 0.4550931751728058,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 6444
+    },
+    {
+      "epoch": 0.7487656113854196,
+      "grad_norm": 0.4081590175628662,
+      "learning_rate": 0.0001,
+      "loss": 1.3085,
+      "step": 6445
+    },
+    {
+      "epoch": 0.7488817891373802,
+      "grad_norm": 0.41531556844711304,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 6446
+    },
+    {
+      "epoch": 0.7489979668893407,
+      "grad_norm": 0.46945273876190186,
+      "learning_rate": 0.0001,
+      "loss": 1.6785,
+      "step": 6447
+    },
+    {
+      "epoch": 0.7491141446413012,
+      "grad_norm": 0.4772491455078125,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 6448
+    },
+    {
+      "epoch": 0.7492303223932617,
+      "grad_norm": 0.4390372335910797,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 6449
+    },
+    {
+      "epoch": 0.7493465001452222,
+      "grad_norm": 0.425656795501709,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 6450
+    },
+    {
+      "epoch": 0.7494626778971827,
+      "grad_norm": 0.44285711646080017,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 6451
+    },
+    {
+      "epoch": 0.7495788556491432,
+      "grad_norm": 0.44809186458587646,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 6452
+    },
+    {
+      "epoch": 0.7496950334011037,
+      "grad_norm": 0.44603654742240906,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 6453
+    },
+    {
+      "epoch": 0.7498112111530641,
+      "grad_norm": 0.45368316769599915,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 6454
+    },
+    {
+      "epoch": 0.7499273889050246,
+      "grad_norm": 0.464532732963562,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 6455
+    },
+    {
+      "epoch": 0.7500435666569852,
+      "grad_norm": 0.43661272525787354,
+      "learning_rate": 0.0001,
+      "loss": 1.4989,
+      "step": 6456
+    },
+    {
+      "epoch": 0.7501597444089457,
+      "grad_norm": 0.4373878240585327,
+      "learning_rate": 0.0001,
+      "loss": 1.6482,
+      "step": 6457
+    },
+    {
+      "epoch": 0.7502759221609062,
+      "grad_norm": 0.45285502076148987,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 6458
+    },
+    {
+      "epoch": 0.7503920999128667,
+      "grad_norm": 0.4704199433326721,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 6459
+    },
+    {
+      "epoch": 0.7505082776648272,
+      "grad_norm": 0.45915988087654114,
+      "learning_rate": 0.0001,
+      "loss": 1.7206,
+      "step": 6460
+    },
+    {
+      "epoch": 0.7506244554167877,
+      "grad_norm": 0.46170055866241455,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 6461
+    },
+    {
+      "epoch": 0.7507406331687482,
+      "grad_norm": 0.46552303433418274,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 6462
+    },
+    {
+      "epoch": 0.7508568109207087,
+      "grad_norm": 0.47524699568748474,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 6463
+    },
+    {
+      "epoch": 0.7509729886726692,
+      "grad_norm": 0.4851406514644623,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 6464
+    },
+    {
+      "epoch": 0.7510891664246296,
+      "grad_norm": 0.46212369203567505,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 6465
+    },
+    {
+      "epoch": 0.7512053441765901,
+      "grad_norm": 0.4637957811355591,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 6466
+    },
+    {
+      "epoch": 0.7513215219285507,
+      "grad_norm": 0.46081292629241943,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 6467
+    },
+    {
+      "epoch": 0.7514376996805112,
+      "grad_norm": 0.418517529964447,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 6468
+    },
+    {
+      "epoch": 0.7515538774324717,
+      "grad_norm": 0.42126187682151794,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 6469
+    },
+    {
+      "epoch": 0.7516700551844322,
+      "grad_norm": 0.45902541279792786,
+      "learning_rate": 0.0001,
+      "loss": 1.6279,
+      "step": 6470
+    },
+    {
+      "epoch": 0.7517862329363927,
+      "grad_norm": 0.42754048109054565,
+      "learning_rate": 0.0001,
+      "loss": 1.4632,
+      "step": 6471
+    },
+    {
+      "epoch": 0.7519024106883532,
+      "grad_norm": 0.4438011348247528,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 6472
+    },
+    {
+      "epoch": 0.7520185884403137,
+      "grad_norm": 0.46611130237579346,
+      "learning_rate": 0.0001,
+      "loss": 1.7748,
+      "step": 6473
+    },
+    {
+      "epoch": 0.7521347661922742,
+      "grad_norm": 0.43576914072036743,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 6474
+    },
+    {
+      "epoch": 0.7522509439442346,
+      "grad_norm": 0.4808332622051239,
+      "learning_rate": 0.0001,
+      "loss": 1.6467,
+      "step": 6475
+    },
+    {
+      "epoch": 0.7523671216961951,
+      "grad_norm": 0.4799876809120178,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 6476
+    },
+    {
+      "epoch": 0.7524832994481557,
+      "grad_norm": 0.4838355481624603,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 6477
+    },
+    {
+      "epoch": 0.7525994772001162,
+      "grad_norm": 0.47365084290504456,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 6478
+    },
+    {
+      "epoch": 0.7527156549520767,
+      "grad_norm": 0.4396180510520935,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 6479
+    },
+    {
+      "epoch": 0.7528318327040372,
+      "grad_norm": 0.4406949579715729,
+      "learning_rate": 0.0001,
+      "loss": 1.6357,
+      "step": 6480
+    },
+    {
+      "epoch": 0.7529480104559977,
+      "grad_norm": 0.43334612250328064,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 6481
+    },
+    {
+      "epoch": 0.7530641882079582,
+      "grad_norm": 0.4289649426937103,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 6482
+    },
+    {
+      "epoch": 0.7531803659599187,
+      "grad_norm": 0.4250386953353882,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 6483
+    },
+    {
+      "epoch": 0.7532965437118792,
+      "grad_norm": 0.4700089395046234,
+      "learning_rate": 0.0001,
+      "loss": 1.6436,
+      "step": 6484
+    },
+    {
+      "epoch": 0.7534127214638396,
+      "grad_norm": 0.4487655460834503,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 6485
+    },
+    {
+      "epoch": 0.7535288992158001,
+      "grad_norm": 0.44160258769989014,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 6486
+    },
+    {
+      "epoch": 0.7536450769677606,
+      "grad_norm": 0.43378591537475586,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 6487
+    },
+    {
+      "epoch": 0.7537612547197212,
+      "grad_norm": 0.4056842029094696,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 6488
+    },
+    {
+      "epoch": 0.7538774324716817,
+      "grad_norm": 0.4529426693916321,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 6489
+    },
+    {
+      "epoch": 0.7539936102236422,
+      "grad_norm": 0.5032495856285095,
+      "learning_rate": 0.0001,
+      "loss": 1.7247,
+      "step": 6490
+    },
+    {
+      "epoch": 0.7541097879756027,
+      "grad_norm": 0.4208425283432007,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 6491
+    },
+    {
+      "epoch": 0.7542259657275632,
+      "grad_norm": 0.429240345954895,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 6492
+    },
+    {
+      "epoch": 0.7543421434795237,
+      "grad_norm": 0.45296066999435425,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 6493
+    },
+    {
+      "epoch": 0.7544583212314842,
+      "grad_norm": 0.4730745255947113,
+      "learning_rate": 0.0001,
+      "loss": 1.7686,
+      "step": 6494
+    },
+    {
+      "epoch": 0.7545744989834446,
+      "grad_norm": 0.43891751766204834,
+      "learning_rate": 0.0001,
+      "loss": 1.7096,
+      "step": 6495
+    },
+    {
+      "epoch": 0.7546906767354051,
+      "grad_norm": 0.4880208671092987,
+      "learning_rate": 0.0001,
+      "loss": 1.7739,
+      "step": 6496
+    },
+    {
+      "epoch": 0.7548068544873656,
+      "grad_norm": 0.43888798356056213,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 6497
+    },
+    {
+      "epoch": 0.7549230322393262,
+      "grad_norm": 0.42402908205986023,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 6498
+    },
+    {
+      "epoch": 0.7550392099912867,
+      "grad_norm": 0.43991389870643616,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 6499
+    },
+    {
+      "epoch": 0.7551553877432472,
+      "grad_norm": 0.43535780906677246,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 6500
+    },
+    {
+      "epoch": 0.7552715654952077,
+      "grad_norm": 0.4613569974899292,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 6501
+    },
+    {
+      "epoch": 0.7553877432471682,
+      "grad_norm": 0.4456695020198822,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 6502
+    },
+    {
+      "epoch": 0.7555039209991287,
+      "grad_norm": 0.4481378197669983,
+      "learning_rate": 0.0001,
+      "loss": 1.6161,
+      "step": 6503
+    },
+    {
+      "epoch": 0.7556200987510892,
+      "grad_norm": 0.45384448766708374,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 6504
+    },
+    {
+      "epoch": 0.7557362765030496,
+      "grad_norm": 0.474298894405365,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 6505
+    },
+    {
+      "epoch": 0.7558524542550101,
+      "grad_norm": 0.478977233171463,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 6506
+    },
+    {
+      "epoch": 0.7559686320069706,
+      "grad_norm": 0.45256179571151733,
+      "learning_rate": 0.0001,
+      "loss": 1.745,
+      "step": 6507
+    },
+    {
+      "epoch": 0.7560848097589312,
+      "grad_norm": 0.465461790561676,
+      "learning_rate": 0.0001,
+      "loss": 1.7028,
+      "step": 6508
+    },
+    {
+      "epoch": 0.7562009875108917,
+      "grad_norm": 0.4733600318431854,
+      "learning_rate": 0.0001,
+      "loss": 1.7272,
+      "step": 6509
+    },
+    {
+      "epoch": 0.7563171652628522,
+      "grad_norm": 0.46702197194099426,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 6510
+    },
+    {
+      "epoch": 0.7564333430148127,
+      "grad_norm": 0.4695473909378052,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 6511
+    },
+    {
+      "epoch": 0.7565495207667732,
+      "grad_norm": 0.48179465532302856,
+      "learning_rate": 0.0001,
+      "loss": 1.7386,
+      "step": 6512
+    },
+    {
+      "epoch": 0.7566656985187337,
+      "grad_norm": 0.4340745210647583,
+      "learning_rate": 0.0001,
+      "loss": 1.385,
+      "step": 6513
+    },
+    {
+      "epoch": 0.7567818762706942,
+      "grad_norm": 0.4561876058578491,
+      "learning_rate": 0.0001,
+      "loss": 1.7278,
+      "step": 6514
+    },
+    {
+      "epoch": 0.7568980540226546,
+      "grad_norm": 0.4735701084136963,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 6515
+    },
+    {
+      "epoch": 0.7570142317746151,
+      "grad_norm": 0.5096227526664734,
+      "learning_rate": 0.0001,
+      "loss": 1.6791,
+      "step": 6516
+    },
+    {
+      "epoch": 0.7571304095265756,
+      "grad_norm": 0.5059962868690491,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 6517
+    },
+    {
+      "epoch": 0.7572465872785361,
+      "grad_norm": 0.4230511784553528,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 6518
+    },
+    {
+      "epoch": 0.7573627650304967,
+      "grad_norm": 0.5008574724197388,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 6519
+    },
+    {
+      "epoch": 0.7574789427824572,
+      "grad_norm": 0.45503777265548706,
+      "learning_rate": 0.0001,
+      "loss": 1.582,
+      "step": 6520
+    },
+    {
+      "epoch": 0.7575951205344177,
+      "grad_norm": 0.49144798517227173,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 6521
+    },
+    {
+      "epoch": 0.7577112982863782,
+      "grad_norm": 0.450183629989624,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 6522
+    },
+    {
+      "epoch": 0.7578274760383387,
+      "grad_norm": 0.4511738717556,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 6523
+    },
+    {
+      "epoch": 0.7579436537902992,
+      "grad_norm": 0.45936381816864014,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 6524
+    },
+    {
+      "epoch": 0.7580598315422596,
+      "grad_norm": 0.4909375309944153,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 6525
+    },
+    {
+      "epoch": 0.7581760092942201,
+      "grad_norm": 0.426647812128067,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 6526
+    },
+    {
+      "epoch": 0.7582921870461806,
+      "grad_norm": 0.42700493335723877,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 6527
+    },
+    {
+      "epoch": 0.7584083647981411,
+      "grad_norm": 0.4363797903060913,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 6528
+    },
+    {
+      "epoch": 0.7585245425501017,
+      "grad_norm": 0.47377872467041016,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 6529
+    },
+    {
+      "epoch": 0.7586407203020622,
+      "grad_norm": 0.45347169041633606,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 6530
+    },
+    {
+      "epoch": 0.7587568980540227,
+      "grad_norm": 0.4400898814201355,
+      "learning_rate": 0.0001,
+      "loss": 1.6557,
+      "step": 6531
+    },
+    {
+      "epoch": 0.7588730758059832,
+      "grad_norm": 0.4629836678504944,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 6532
+    },
+    {
+      "epoch": 0.7589892535579437,
+      "grad_norm": 0.43157103657722473,
+      "learning_rate": 0.0001,
+      "loss": 1.3997,
+      "step": 6533
+    },
+    {
+      "epoch": 0.7591054313099042,
+      "grad_norm": 0.44874757528305054,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 6534
+    },
+    {
+      "epoch": 0.7592216090618646,
+      "grad_norm": 0.4572241008281708,
+      "learning_rate": 0.0001,
+      "loss": 1.7732,
+      "step": 6535
+    },
+    {
+      "epoch": 0.7593377868138251,
+      "grad_norm": 0.509334146976471,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 6536
+    },
+    {
+      "epoch": 0.7594539645657856,
+      "grad_norm": 0.45650917291641235,
+      "learning_rate": 0.0001,
+      "loss": 1.7018,
+      "step": 6537
+    },
+    {
+      "epoch": 0.7595701423177461,
+      "grad_norm": 0.4282233715057373,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 6538
+    },
+    {
+      "epoch": 0.7596863200697066,
+      "grad_norm": 0.49342942237854004,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 6539
+    },
+    {
+      "epoch": 0.7598024978216672,
+      "grad_norm": 0.45365840196609497,
+      "learning_rate": 0.0001,
+      "loss": 1.7058,
+      "step": 6540
+    },
+    {
+      "epoch": 0.7599186755736277,
+      "grad_norm": 0.4653255343437195,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 6541
+    },
+    {
+      "epoch": 0.7600348533255882,
+      "grad_norm": 0.4376204013824463,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 6542
+    },
+    {
+      "epoch": 0.7601510310775487,
+      "grad_norm": 0.42907100915908813,
+      "learning_rate": 0.0001,
+      "loss": 1.5606,
+      "step": 6543
+    },
+    {
+      "epoch": 0.7602672088295092,
+      "grad_norm": 0.4851232171058655,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 6544
+    },
+    {
+      "epoch": 0.7603833865814696,
+      "grad_norm": 0.4842982590198517,
+      "learning_rate": 0.0001,
+      "loss": 1.5926,
+      "step": 6545
+    },
+    {
+      "epoch": 0.7604995643334301,
+      "grad_norm": 0.44596225023269653,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 6546
+    },
+    {
+      "epoch": 0.7606157420853906,
+      "grad_norm": 0.42956697940826416,
+      "learning_rate": 0.0001,
+      "loss": 1.5368,
+      "step": 6547
+    },
+    {
+      "epoch": 0.7607319198373511,
+      "grad_norm": 0.4747844934463501,
+      "learning_rate": 0.0001,
+      "loss": 1.6931,
+      "step": 6548
+    },
+    {
+      "epoch": 0.7608480975893116,
+      "grad_norm": 0.4778608977794647,
+      "learning_rate": 0.0001,
+      "loss": 1.7017,
+      "step": 6549
+    },
+    {
+      "epoch": 0.7609642753412722,
+      "grad_norm": 0.4931018352508545,
+      "learning_rate": 0.0001,
+      "loss": 1.7501,
+      "step": 6550
+    },
+    {
+      "epoch": 0.7610804530932327,
+      "grad_norm": 0.4602007567882538,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 6551
+    },
+    {
+      "epoch": 0.7611966308451932,
+      "grad_norm": 0.4534691572189331,
+      "learning_rate": 0.0001,
+      "loss": 1.418,
+      "step": 6552
+    },
+    {
+      "epoch": 0.7613128085971537,
+      "grad_norm": 0.4759400486946106,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 6553
+    },
+    {
+      "epoch": 0.7614289863491142,
+      "grad_norm": 0.44911110401153564,
+      "learning_rate": 0.0001,
+      "loss": 1.7104,
+      "step": 6554
+    },
+    {
+      "epoch": 0.7615451641010746,
+      "grad_norm": 0.4117274582386017,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 6555
+    },
+    {
+      "epoch": 0.7616613418530351,
+      "grad_norm": 0.44577354192733765,
+      "learning_rate": 0.0001,
+      "loss": 1.72,
+      "step": 6556
+    },
+    {
+      "epoch": 0.7617775196049956,
+      "grad_norm": 0.4217338562011719,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 6557
+    },
+    {
+      "epoch": 0.7618936973569561,
+      "grad_norm": 0.4698602855205536,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 6558
+    },
+    {
+      "epoch": 0.7620098751089166,
+      "grad_norm": 0.44901999831199646,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 6559
+    },
+    {
+      "epoch": 0.7621260528608771,
+      "grad_norm": 0.41744062304496765,
+      "learning_rate": 0.0001,
+      "loss": 1.4392,
+      "step": 6560
+    },
+    {
+      "epoch": 0.7622422306128377,
+      "grad_norm": 0.4368325471878052,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 6561
+    },
+    {
+      "epoch": 0.7623584083647982,
+      "grad_norm": 0.43781906366348267,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 6562
+    },
+    {
+      "epoch": 0.7624745861167587,
+      "grad_norm": 0.4672465920448303,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 6563
+    },
+    {
+      "epoch": 0.7625907638687192,
+      "grad_norm": 0.44556140899658203,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 6564
+    },
+    {
+      "epoch": 0.7627069416206796,
+      "grad_norm": 0.47782960534095764,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 6565
+    },
+    {
+      "epoch": 0.7628231193726401,
+      "grad_norm": 0.4672389328479767,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 6566
+    },
+    {
+      "epoch": 0.7629392971246006,
+      "grad_norm": 0.44168704748153687,
+      "learning_rate": 0.0001,
+      "loss": 1.6007,
+      "step": 6567
+    },
+    {
+      "epoch": 0.7630554748765611,
+      "grad_norm": 0.47608694434165955,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 6568
+    },
+    {
+      "epoch": 0.7631716526285216,
+      "grad_norm": 0.4470519423484802,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 6569
+    },
+    {
+      "epoch": 0.7632878303804821,
+      "grad_norm": 0.41022247076034546,
+      "learning_rate": 0.0001,
+      "loss": 1.3155,
+      "step": 6570
+    },
+    {
+      "epoch": 0.7634040081324427,
+      "grad_norm": 0.4504031240940094,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 6571
+    },
+    {
+      "epoch": 0.7635201858844032,
+      "grad_norm": 0.48684489727020264,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 6572
+    },
+    {
+      "epoch": 0.7636363636363637,
+      "grad_norm": 0.43762362003326416,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 6573
+    },
+    {
+      "epoch": 0.7637525413883242,
+      "grad_norm": 0.47748515009880066,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 6574
+    },
+    {
+      "epoch": 0.7638687191402846,
+      "grad_norm": 0.4670480191707611,
+      "learning_rate": 0.0001,
+      "loss": 1.7789,
+      "step": 6575
+    },
+    {
+      "epoch": 0.7639848968922451,
+      "grad_norm": 0.45016875863075256,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 6576
+    },
+    {
+      "epoch": 0.7641010746442056,
+      "grad_norm": 0.4202631115913391,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 6577
+    },
+    {
+      "epoch": 0.7642172523961661,
+      "grad_norm": 0.4666217565536499,
+      "learning_rate": 0.0001,
+      "loss": 1.4498,
+      "step": 6578
+    },
+    {
+      "epoch": 0.7643334301481266,
+      "grad_norm": 0.46688103675842285,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 6579
+    },
+    {
+      "epoch": 0.7644496079000871,
+      "grad_norm": 0.4328000843524933,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 6580
+    },
+    {
+      "epoch": 0.7645657856520476,
+      "grad_norm": 0.4551675617694855,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 6581
+    },
+    {
+      "epoch": 0.7646819634040082,
+      "grad_norm": 0.4549828767776489,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 6582
+    },
+    {
+      "epoch": 0.7647981411559687,
+      "grad_norm": 0.45938196778297424,
+      "learning_rate": 0.0001,
+      "loss": 1.6921,
+      "step": 6583
+    },
+    {
+      "epoch": 0.7649143189079292,
+      "grad_norm": 0.4694269299507141,
+      "learning_rate": 0.0001,
+      "loss": 1.5216,
+      "step": 6584
+    },
+    {
+      "epoch": 0.7650304966598896,
+      "grad_norm": 0.47541555762290955,
+      "learning_rate": 0.0001,
+      "loss": 1.6457,
+      "step": 6585
+    },
+    {
+      "epoch": 0.7651466744118501,
+      "grad_norm": 0.432059645652771,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 6586
+    },
+    {
+      "epoch": 0.7652628521638106,
+      "grad_norm": 0.48397889733314514,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 6587
+    },
+    {
+      "epoch": 0.7653790299157711,
+      "grad_norm": 0.437933087348938,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 6588
+    },
+    {
+      "epoch": 0.7654952076677316,
+      "grad_norm": 0.4440605044364929,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 6589
+    },
+    {
+      "epoch": 0.7656113854196921,
+      "grad_norm": 0.447862833738327,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 6590
+    },
+    {
+      "epoch": 0.7657275631716526,
+      "grad_norm": 0.41971054673194885,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 6591
+    },
+    {
+      "epoch": 0.7658437409236132,
+      "grad_norm": 0.4456073045730591,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 6592
+    },
+    {
+      "epoch": 0.7659599186755737,
+      "grad_norm": 0.43009260296821594,
+      "learning_rate": 0.0001,
+      "loss": 1.5966,
+      "step": 6593
+    },
+    {
+      "epoch": 0.7660760964275342,
+      "grad_norm": 0.482006698846817,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 6594
+    },
+    {
+      "epoch": 0.7661922741794946,
+      "grad_norm": 0.4667458236217499,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 6595
+    },
+    {
+      "epoch": 0.7663084519314551,
+      "grad_norm": 0.4783475995063782,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 6596
+    },
+    {
+      "epoch": 0.7664246296834156,
+      "grad_norm": 0.4709714353084564,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 6597
+    },
+    {
+      "epoch": 0.7665408074353761,
+      "grad_norm": 0.45129141211509705,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 6598
+    },
+    {
+      "epoch": 0.7666569851873366,
+      "grad_norm": 0.44044631719589233,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 6599
+    },
+    {
+      "epoch": 0.7667731629392971,
+      "grad_norm": 0.4458431899547577,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 6600
+    },
+    {
+      "epoch": 0.7668893406912576,
+      "grad_norm": 0.4668479263782501,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 6601
+    },
+    {
+      "epoch": 0.7670055184432181,
+      "grad_norm": 0.4631541073322296,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 6602
+    },
+    {
+      "epoch": 0.7671216961951787,
+      "grad_norm": 0.47970542311668396,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 6603
+    },
+    {
+      "epoch": 0.7672378739471392,
+      "grad_norm": 0.4707159101963043,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 6604
+    },
+    {
+      "epoch": 0.7673540516990996,
+      "grad_norm": 0.48300158977508545,
+      "learning_rate": 0.0001,
+      "loss": 1.5748,
+      "step": 6605
+    },
+    {
+      "epoch": 0.7674702294510601,
+      "grad_norm": 0.47976362705230713,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 6606
+    },
+    {
+      "epoch": 0.7675864072030206,
+      "grad_norm": 0.4959631860256195,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 6607
+    },
+    {
+      "epoch": 0.7677025849549811,
+      "grad_norm": 0.4722089171409607,
+      "learning_rate": 0.0001,
+      "loss": 1.6423,
+      "step": 6608
+    },
+    {
+      "epoch": 0.7678187627069416,
+      "grad_norm": 0.5199443101882935,
+      "learning_rate": 0.0001,
+      "loss": 1.642,
+      "step": 6609
+    },
+    {
+      "epoch": 0.7679349404589021,
+      "grad_norm": 0.4067780375480652,
+      "learning_rate": 0.0001,
+      "loss": 1.3372,
+      "step": 6610
+    },
+    {
+      "epoch": 0.7680511182108626,
+      "grad_norm": 0.43238359689712524,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 6611
+    },
+    {
+      "epoch": 0.7681672959628231,
+      "grad_norm": 0.5341728925704956,
+      "learning_rate": 0.0001,
+      "loss": 1.8739,
+      "step": 6612
+    },
+    {
+      "epoch": 0.7682834737147837,
+      "grad_norm": 0.46490055322647095,
+      "learning_rate": 0.0001,
+      "loss": 1.7022,
+      "step": 6613
+    },
+    {
+      "epoch": 0.7683996514667442,
+      "grad_norm": 0.4309159219264984,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 6614
+    },
+    {
+      "epoch": 0.7685158292187046,
+      "grad_norm": 0.4642525017261505,
+      "learning_rate": 0.0001,
+      "loss": 1.6468,
+      "step": 6615
+    },
+    {
+      "epoch": 0.7686320069706651,
+      "grad_norm": 0.46897363662719727,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 6616
+    },
+    {
+      "epoch": 0.7687481847226256,
+      "grad_norm": 0.43041566014289856,
+      "learning_rate": 0.0001,
+      "loss": 1.5338,
+      "step": 6617
+    },
+    {
+      "epoch": 0.7688643624745861,
+      "grad_norm": 0.42853835225105286,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 6618
+    },
+    {
+      "epoch": 0.7689805402265466,
+      "grad_norm": 0.4620231091976166,
+      "learning_rate": 0.0001,
+      "loss": 1.73,
+      "step": 6619
+    },
+    {
+      "epoch": 0.7690967179785071,
+      "grad_norm": 0.44504088163375854,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 6620
+    },
+    {
+      "epoch": 0.7692128957304676,
+      "grad_norm": 0.48609089851379395,
+      "learning_rate": 0.0001,
+      "loss": 1.5766,
+      "step": 6621
+    },
+    {
+      "epoch": 0.7693290734824281,
+      "grad_norm": 0.483452171087265,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 6622
+    },
+    {
+      "epoch": 0.7694452512343886,
+      "grad_norm": 0.4754545986652374,
+      "learning_rate": 0.0001,
+      "loss": 1.663,
+      "step": 6623
+    },
+    {
+      "epoch": 0.7695614289863492,
+      "grad_norm": 0.4533579349517822,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 6624
+    },
+    {
+      "epoch": 0.7696776067383096,
+      "grad_norm": 0.4664144814014435,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 6625
+    },
+    {
+      "epoch": 0.7697937844902701,
+      "grad_norm": 0.48113691806793213,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 6626
+    },
+    {
+      "epoch": 0.7699099622422306,
+      "grad_norm": 0.4325944185256958,
+      "learning_rate": 0.0001,
+      "loss": 1.5159,
+      "step": 6627
+    },
+    {
+      "epoch": 0.7700261399941911,
+      "grad_norm": 0.4415172338485718,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 6628
+    },
+    {
+      "epoch": 0.7701423177461516,
+      "grad_norm": 0.41996294260025024,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 6629
+    },
+    {
+      "epoch": 0.7702584954981121,
+      "grad_norm": 0.4572359025478363,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 6630
+    },
+    {
+      "epoch": 0.7703746732500726,
+      "grad_norm": 0.4630671441555023,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 6631
+    },
+    {
+      "epoch": 0.7704908510020331,
+      "grad_norm": 0.4352489411830902,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 6632
+    },
+    {
+      "epoch": 0.7706070287539936,
+      "grad_norm": 0.4531479477882385,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 6633
+    },
+    {
+      "epoch": 0.7707232065059542,
+      "grad_norm": 0.5102584362030029,
+      "learning_rate": 0.0001,
+      "loss": 1.5636,
+      "step": 6634
+    },
+    {
+      "epoch": 0.7708393842579147,
+      "grad_norm": 0.45892634987831116,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 6635
+    },
+    {
+      "epoch": 0.7709555620098751,
+      "grad_norm": 0.4718113839626312,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 6636
+    },
+    {
+      "epoch": 0.7710717397618356,
+      "grad_norm": 0.4592994749546051,
+      "learning_rate": 0.0001,
+      "loss": 1.644,
+      "step": 6637
+    },
+    {
+      "epoch": 0.7711879175137961,
+      "grad_norm": 0.4589068293571472,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 6638
+    },
+    {
+      "epoch": 0.7713040952657566,
+      "grad_norm": 0.41312143206596375,
+      "learning_rate": 0.0001,
+      "loss": 1.3861,
+      "step": 6639
+    },
+    {
+      "epoch": 0.7714202730177171,
+      "grad_norm": 0.4337233901023865,
+      "learning_rate": 0.0001,
+      "loss": 1.4143,
+      "step": 6640
+    },
+    {
+      "epoch": 0.7715364507696776,
+      "grad_norm": 0.4643138349056244,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 6641
+    },
+    {
+      "epoch": 0.7716526285216381,
+      "grad_norm": 0.485331654548645,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 6642
+    },
+    {
+      "epoch": 0.7717688062735986,
+      "grad_norm": 0.4634449779987335,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 6643
+    },
+    {
+      "epoch": 0.771884984025559,
+      "grad_norm": 0.45036226511001587,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 6644
+    },
+    {
+      "epoch": 0.7720011617775197,
+      "grad_norm": 0.45549461245536804,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 6645
+    },
+    {
+      "epoch": 0.7721173395294801,
+      "grad_norm": 0.4745498299598694,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 6646
+    },
+    {
+      "epoch": 0.7722335172814406,
+      "grad_norm": 0.5061119198799133,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 6647
+    },
+    {
+      "epoch": 0.7723496950334011,
+      "grad_norm": 0.4423942565917969,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 6648
+    },
+    {
+      "epoch": 0.7724658727853616,
+      "grad_norm": 0.46046459674835205,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 6649
+    },
+    {
+      "epoch": 0.7725820505373221,
+      "grad_norm": 0.4494078755378723,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 6650
+    },
+    {
+      "epoch": 0.7726982282892826,
+      "grad_norm": 0.4689868986606598,
+      "learning_rate": 0.0001,
+      "loss": 1.7184,
+      "step": 6651
+    },
+    {
+      "epoch": 0.7728144060412431,
+      "grad_norm": 0.42947614192962646,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 6652
+    },
+    {
+      "epoch": 0.7729305837932036,
+      "grad_norm": 0.424402117729187,
+      "learning_rate": 0.0001,
+      "loss": 1.3695,
+      "step": 6653
+    },
+    {
+      "epoch": 0.773046761545164,
+      "grad_norm": 0.44652658700942993,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 6654
+    },
+    {
+      "epoch": 0.7731629392971247,
+      "grad_norm": 0.4878056049346924,
+      "learning_rate": 0.0001,
+      "loss": 1.6091,
+      "step": 6655
+    },
+    {
+      "epoch": 0.7732791170490851,
+      "grad_norm": 0.4504278898239136,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 6656
+    },
+    {
+      "epoch": 0.7733952948010456,
+      "grad_norm": 0.43324804306030273,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 6657
+    },
+    {
+      "epoch": 0.7735114725530061,
+      "grad_norm": 0.47235342860221863,
+      "learning_rate": 0.0001,
+      "loss": 1.7402,
+      "step": 6658
+    },
+    {
+      "epoch": 0.7736276503049666,
+      "grad_norm": 0.4258509576320648,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 6659
+    },
+    {
+      "epoch": 0.7737438280569271,
+      "grad_norm": 0.47323077917099,
+      "learning_rate": 0.0001,
+      "loss": 1.5615,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7738600058088876,
+      "grad_norm": 0.44588038325309753,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 6661
+    },
+    {
+      "epoch": 0.7739761835608481,
+      "grad_norm": 0.4483187794685364,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 6662
+    },
+    {
+      "epoch": 0.7740923613128086,
+      "grad_norm": 0.4594082534313202,
+      "learning_rate": 0.0001,
+      "loss": 1.6624,
+      "step": 6663
+    },
+    {
+      "epoch": 0.774208539064769,
+      "grad_norm": 0.4679012894630432,
+      "learning_rate": 0.0001,
+      "loss": 1.7395,
+      "step": 6664
+    },
+    {
+      "epoch": 0.7743247168167295,
+      "grad_norm": 0.5004777312278748,
+      "learning_rate": 0.0001,
+      "loss": 1.7538,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7744408945686901,
+      "grad_norm": 0.4686128795146942,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 6666
+    },
+    {
+      "epoch": 0.7745570723206506,
+      "grad_norm": 0.46381059288978577,
+      "learning_rate": 0.0001,
+      "loss": 1.6585,
+      "step": 6667
+    },
+    {
+      "epoch": 0.7746732500726111,
+      "grad_norm": 0.4782971739768982,
+      "learning_rate": 0.0001,
+      "loss": 1.7547,
+      "step": 6668
+    },
+    {
+      "epoch": 0.7747894278245716,
+      "grad_norm": 0.48163890838623047,
+      "learning_rate": 0.0001,
+      "loss": 1.7483,
+      "step": 6669
+    },
+    {
+      "epoch": 0.7749056055765321,
+      "grad_norm": 0.47830620408058167,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7750217833284926,
+      "grad_norm": 0.4307786524295807,
+      "learning_rate": 0.0001,
+      "loss": 1.7148,
+      "step": 6671
+    },
+    {
+      "epoch": 0.7751379610804531,
+      "grad_norm": 0.4738655090332031,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 6672
+    },
+    {
+      "epoch": 0.7752541388324136,
+      "grad_norm": 0.45517605543136597,
+      "learning_rate": 0.0001,
+      "loss": 1.6803,
+      "step": 6673
+    },
+    {
+      "epoch": 0.775370316584374,
+      "grad_norm": 0.44930917024612427,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 6674
+    },
+    {
+      "epoch": 0.7754864943363345,
+      "grad_norm": 0.4695178270339966,
+      "learning_rate": 0.0001,
+      "loss": 1.6834,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7756026720882951,
+      "grad_norm": 0.4576238691806793,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 6676
+    },
+    {
+      "epoch": 0.7757188498402556,
+      "grad_norm": 0.44012871384620667,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 6677
+    },
+    {
+      "epoch": 0.7758350275922161,
+      "grad_norm": 0.4941727817058563,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 6678
+    },
+    {
+      "epoch": 0.7759512053441766,
+      "grad_norm": 0.4414857029914856,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 6679
+    },
+    {
+      "epoch": 0.7760673830961371,
+      "grad_norm": 0.468009352684021,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7761835608480976,
+      "grad_norm": 0.514825165271759,
+      "learning_rate": 0.0001,
+      "loss": 1.6376,
+      "step": 6681
+    },
+    {
+      "epoch": 0.7762997386000581,
+      "grad_norm": 0.47073277831077576,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 6682
+    },
+    {
+      "epoch": 0.7764159163520186,
+      "grad_norm": 0.46188485622406006,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 6683
+    },
+    {
+      "epoch": 0.776532094103979,
+      "grad_norm": 0.5069568753242493,
+      "learning_rate": 0.0001,
+      "loss": 1.577,
+      "step": 6684
+    },
+    {
+      "epoch": 0.7766482718559395,
+      "grad_norm": 0.4475022256374359,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7767644496079001,
+      "grad_norm": 0.48565706610679626,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 6686
+    },
+    {
+      "epoch": 0.7768806273598606,
+      "grad_norm": 0.4480251967906952,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 6687
+    },
+    {
+      "epoch": 0.7769968051118211,
+      "grad_norm": 0.4224783480167389,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 6688
+    },
+    {
+      "epoch": 0.7771129828637816,
+      "grad_norm": 0.45073410868644714,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7772291606157421,
+      "grad_norm": 0.4535382390022278,
+      "learning_rate": 0.0001,
+      "loss": 1.6862,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7773453383677026,
+      "grad_norm": 0.45997950434684753,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7774615161196631,
+      "grad_norm": 0.4208465814590454,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7775776938716236,
+      "grad_norm": 0.4323180317878723,
+      "learning_rate": 0.0001,
+      "loss": 1.4691,
+      "step": 6693
+    },
+    {
+      "epoch": 0.777693871623584,
+      "grad_norm": 0.47370851039886475,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7778100493755445,
+      "grad_norm": 0.46739041805267334,
+      "learning_rate": 0.0001,
+      "loss": 1.6456,
+      "step": 6695
+    },
+    {
+      "epoch": 0.777926227127505,
+      "grad_norm": 0.44516703486442566,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7780424048794656,
+      "grad_norm": 0.4348163306713104,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 6697
+    },
+    {
+      "epoch": 0.7781585826314261,
+      "grad_norm": 0.5044941902160645,
+      "learning_rate": 0.0001,
+      "loss": 1.7077,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7782747603833866,
+      "grad_norm": 0.48483940958976746,
+      "learning_rate": 0.0001,
+      "loss": 1.7837,
+      "step": 6699
+    },
+    {
+      "epoch": 0.7783909381353471,
+      "grad_norm": 0.46066558361053467,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7785071158873076,
+      "grad_norm": 0.45025432109832764,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 6701
+    },
+    {
+      "epoch": 0.7786232936392681,
+      "grad_norm": 0.4388742744922638,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 6702
+    },
+    {
+      "epoch": 0.7787394713912286,
+      "grad_norm": 0.44050899147987366,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 6703
+    },
+    {
+      "epoch": 0.778855649143189,
+      "grad_norm": 0.5146862864494324,
+      "learning_rate": 0.0001,
+      "loss": 1.8927,
+      "step": 6704
+    },
+    {
+      "epoch": 0.7789718268951495,
+      "grad_norm": 0.43440109491348267,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 6705
+    },
+    {
+      "epoch": 0.77908800464711,
+      "grad_norm": 0.43129873275756836,
+      "learning_rate": 0.0001,
+      "loss": 1.6918,
+      "step": 6706
+    },
+    {
+      "epoch": 0.7792041823990706,
+      "grad_norm": 0.4510217308998108,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 6707
+    },
+    {
+      "epoch": 0.7793203601510311,
+      "grad_norm": 0.5071493983268738,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 6708
+    },
+    {
+      "epoch": 0.7794365379029916,
+      "grad_norm": 0.4631008505821228,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 6709
+    },
+    {
+      "epoch": 0.7795527156549521,
+      "grad_norm": 0.4254882335662842,
+      "learning_rate": 0.0001,
+      "loss": 1.4675,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7796688934069126,
+      "grad_norm": 0.4777772128582001,
+      "learning_rate": 0.0001,
+      "loss": 1.7763,
+      "step": 6711
+    },
+    {
+      "epoch": 0.7797850711588731,
+      "grad_norm": 0.4478752017021179,
+      "learning_rate": 0.0001,
+      "loss": 1.6425,
+      "step": 6712
+    },
+    {
+      "epoch": 0.7799012489108336,
+      "grad_norm": 0.4405520260334015,
+      "learning_rate": 0.0001,
+      "loss": 1.7456,
+      "step": 6713
+    },
+    {
+      "epoch": 0.7800174266627941,
+      "grad_norm": 0.43905889987945557,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 6714
+    },
+    {
+      "epoch": 0.7801336044147545,
+      "grad_norm": 0.46866193413734436,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 6715
+    },
+    {
+      "epoch": 0.780249782166715,
+      "grad_norm": 0.4609702229499817,
+      "learning_rate": 0.0001,
+      "loss": 1.744,
+      "step": 6716
+    },
+    {
+      "epoch": 0.7803659599186755,
+      "grad_norm": 0.4288838803768158,
+      "learning_rate": 0.0001,
+      "loss": 1.5726,
+      "step": 6717
+    },
+    {
+      "epoch": 0.7804821376706361,
+      "grad_norm": 0.4898937940597534,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 6718
+    },
+    {
+      "epoch": 0.7805983154225966,
+      "grad_norm": 0.46226128935813904,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 6719
+    },
+    {
+      "epoch": 0.7807144931745571,
+      "grad_norm": 0.4708699584007263,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 6720
+    },
+    {
+      "epoch": 0.7808306709265176,
+      "grad_norm": 0.4597412943840027,
+      "learning_rate": 0.0001,
+      "loss": 1.7016,
+      "step": 6721
+    },
+    {
+      "epoch": 0.7809468486784781,
+      "grad_norm": 0.44381624460220337,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 6722
+    },
+    {
+      "epoch": 0.7810630264304386,
+      "grad_norm": 0.43926021456718445,
+      "learning_rate": 0.0001,
+      "loss": 1.6971,
+      "step": 6723
+    },
+    {
+      "epoch": 0.7811792041823991,
+      "grad_norm": 0.4407069683074951,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 6724
+    },
+    {
+      "epoch": 0.7812953819343595,
+      "grad_norm": 0.44746869802474976,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 6725
+    },
+    {
+      "epoch": 0.78141155968632,
+      "grad_norm": 0.44408538937568665,
+      "learning_rate": 0.0001,
+      "loss": 1.378,
+      "step": 6726
+    },
+    {
+      "epoch": 0.7815277374382805,
+      "grad_norm": 0.44895273447036743,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 6727
+    },
+    {
+      "epoch": 0.7816439151902411,
+      "grad_norm": 0.4479658901691437,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 6728
+    },
+    {
+      "epoch": 0.7817600929422016,
+      "grad_norm": 0.4522935450077057,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 6729
+    },
+    {
+      "epoch": 0.7818762706941621,
+      "grad_norm": 0.4975508749485016,
+      "learning_rate": 0.0001,
+      "loss": 1.5601,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7819924484461226,
+      "grad_norm": 0.4425521194934845,
+      "learning_rate": 0.0001,
+      "loss": 1.6797,
+      "step": 6731
+    },
+    {
+      "epoch": 0.7821086261980831,
+      "grad_norm": 0.4268142580986023,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 6732
+    },
+    {
+      "epoch": 0.7822248039500436,
+      "grad_norm": 0.45263275504112244,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 6733
+    },
+    {
+      "epoch": 0.7823409817020041,
+      "grad_norm": 0.4428475797176361,
+      "learning_rate": 0.0001,
+      "loss": 1.4382,
+      "step": 6734
+    },
+    {
+      "epoch": 0.7824571594539645,
+      "grad_norm": 0.5304811596870422,
+      "learning_rate": 0.0001,
+      "loss": 1.7288,
+      "step": 6735
+    },
+    {
+      "epoch": 0.782573337205925,
+      "grad_norm": 0.43402206897735596,
+      "learning_rate": 0.0001,
+      "loss": 1.5834,
+      "step": 6736
+    },
+    {
+      "epoch": 0.7826895149578855,
+      "grad_norm": 0.43947911262512207,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 6737
+    },
+    {
+      "epoch": 0.782805692709846,
+      "grad_norm": 0.4783528447151184,
+      "learning_rate": 0.0001,
+      "loss": 1.6796,
+      "step": 6738
+    },
+    {
+      "epoch": 0.7829218704618066,
+      "grad_norm": 0.4822714924812317,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 6739
+    },
+    {
+      "epoch": 0.7830380482137671,
+      "grad_norm": 0.4645395576953888,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7831542259657276,
+      "grad_norm": 0.47700071334838867,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 6741
+    },
+    {
+      "epoch": 0.7832704037176881,
+      "grad_norm": 0.48509228229522705,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 6742
+    },
+    {
+      "epoch": 0.7833865814696486,
+      "grad_norm": 0.43715542554855347,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 6743
+    },
+    {
+      "epoch": 0.7835027592216091,
+      "grad_norm": 0.5839574933052063,
+      "learning_rate": 0.0001,
+      "loss": 1.7659,
+      "step": 6744
+    },
+    {
+      "epoch": 0.7836189369735695,
+      "grad_norm": 0.4410605728626251,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 6745
+    },
+    {
+      "epoch": 0.78373511472553,
+      "grad_norm": 0.44699952006340027,
+      "learning_rate": 0.0001,
+      "loss": 1.4267,
+      "step": 6746
+    },
+    {
+      "epoch": 0.7838512924774905,
+      "grad_norm": 0.43901586532592773,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 6747
+    },
+    {
+      "epoch": 0.783967470229451,
+      "grad_norm": 0.43384185433387756,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 6748
+    },
+    {
+      "epoch": 0.7840836479814116,
+      "grad_norm": 0.41548269987106323,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 6749
+    },
+    {
+      "epoch": 0.7841998257333721,
+      "grad_norm": 0.4889026880264282,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7843160034853326,
+      "grad_norm": 0.46144333481788635,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7844321812372931,
+      "grad_norm": 0.4552895128726959,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7845483589892536,
+      "grad_norm": 0.4651104807853699,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7846645367412141,
+      "grad_norm": 0.4177487790584564,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7847807144931745,
+      "grad_norm": 0.4314608573913574,
+      "learning_rate": 0.0001,
+      "loss": 1.4307,
+      "step": 6755
+    },
+    {
+      "epoch": 0.784896892245135,
+      "grad_norm": 0.4462542235851288,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7850130699970955,
+      "grad_norm": 0.4506712853908539,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 6757
+    },
+    {
+      "epoch": 0.785129247749056,
+      "grad_norm": 0.465562641620636,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 6758
+    },
+    {
+      "epoch": 0.7852454255010165,
+      "grad_norm": 0.5159258842468262,
+      "learning_rate": 0.0001,
+      "loss": 1.7131,
+      "step": 6759
+    },
+    {
+      "epoch": 0.7853616032529771,
+      "grad_norm": 0.4616967439651489,
+      "learning_rate": 0.0001,
+      "loss": 1.6249,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7854777810049376,
+      "grad_norm": 0.4350794553756714,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 6761
+    },
+    {
+      "epoch": 0.7855939587568981,
+      "grad_norm": 0.4288870692253113,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7857101365088586,
+      "grad_norm": 0.4555493891239166,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 6763
+    },
+    {
+      "epoch": 0.7858263142608191,
+      "grad_norm": 0.4390559196472168,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 6764
+    },
+    {
+      "epoch": 0.7859424920127795,
+      "grad_norm": 0.47772979736328125,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 6765
+    },
+    {
+      "epoch": 0.78605866976474,
+      "grad_norm": 0.45631012320518494,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7861748475167005,
+      "grad_norm": 0.4568979740142822,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 6767
+    },
+    {
+      "epoch": 0.786291025268661,
+      "grad_norm": 0.4683852195739746,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 6768
+    },
+    {
+      "epoch": 0.7864072030206215,
+      "grad_norm": 0.45670732855796814,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7865233807725821,
+      "grad_norm": 0.4741782248020172,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7866395585245426,
+      "grad_norm": 0.4552614092826843,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 6771
+    },
+    {
+      "epoch": 0.7867557362765031,
+      "grad_norm": 0.4711034297943115,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 6772
+    },
+    {
+      "epoch": 0.7868719140284636,
+      "grad_norm": 0.4594586491584778,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 6773
+    },
+    {
+      "epoch": 0.7869880917804241,
+      "grad_norm": 0.458446204662323,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 6774
+    },
+    {
+      "epoch": 0.7871042695323845,
+      "grad_norm": 0.48481476306915283,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 6775
+    },
+    {
+      "epoch": 0.787220447284345,
+      "grad_norm": 0.5112566947937012,
+      "learning_rate": 0.0001,
+      "loss": 1.7064,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7873366250363055,
+      "grad_norm": 0.43871763348579407,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 6777
+    },
+    {
+      "epoch": 0.787452802788266,
+      "grad_norm": 0.47852954268455505,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 6778
+    },
+    {
+      "epoch": 0.7875689805402265,
+      "grad_norm": 0.4647180736064911,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 6779
+    },
+    {
+      "epoch": 0.787685158292187,
+      "grad_norm": 0.454166978597641,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 6780
+    },
+    {
+      "epoch": 0.7878013360441476,
+      "grad_norm": 0.447322279214859,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 6781
+    },
+    {
+      "epoch": 0.7879175137961081,
+      "grad_norm": 0.4487408399581909,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 6782
+    },
+    {
+      "epoch": 0.7880336915480686,
+      "grad_norm": 0.46517065167427063,
+      "learning_rate": 0.0001,
+      "loss": 1.725,
+      "step": 6783
+    },
+    {
+      "epoch": 0.7881498693000291,
+      "grad_norm": 0.4516124427318573,
+      "learning_rate": 0.0001,
+      "loss": 1.6411,
+      "step": 6784
+    },
+    {
+      "epoch": 0.7882660470519895,
+      "grad_norm": 0.46726909279823303,
+      "learning_rate": 0.0001,
+      "loss": 1.5802,
+      "step": 6785
+    },
+    {
+      "epoch": 0.78838222480395,
+      "grad_norm": 0.44458258152008057,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 6786
+    },
+    {
+      "epoch": 0.7884984025559105,
+      "grad_norm": 0.44695961475372314,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 6787
+    },
+    {
+      "epoch": 0.788614580307871,
+      "grad_norm": 0.46932363510131836,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 6788
+    },
+    {
+      "epoch": 0.7887307580598315,
+      "grad_norm": 0.4439738392829895,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 6789
+    },
+    {
+      "epoch": 0.788846935811792,
+      "grad_norm": 0.45402002334594727,
+      "learning_rate": 0.0001,
+      "loss": 1.6997,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7889631135637526,
+      "grad_norm": 0.45656514167785645,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 6791
+    },
+    {
+      "epoch": 0.7890792913157131,
+      "grad_norm": 0.4795583486557007,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 6792
+    },
+    {
+      "epoch": 0.7891954690676736,
+      "grad_norm": 0.45822617411613464,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 6793
+    },
+    {
+      "epoch": 0.7893116468196341,
+      "grad_norm": 0.44518178701400757,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 6794
+    },
+    {
+      "epoch": 0.7894278245715945,
+      "grad_norm": 0.4444418251514435,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 6795
+    },
+    {
+      "epoch": 0.789544002323555,
+      "grad_norm": 0.4494422972202301,
+      "learning_rate": 0.0001,
+      "loss": 1.7072,
+      "step": 6796
+    },
+    {
+      "epoch": 0.7896601800755155,
+      "grad_norm": 0.43979451060295105,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 6797
+    },
+    {
+      "epoch": 0.789776357827476,
+      "grad_norm": 0.4433092474937439,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 6798
+    },
+    {
+      "epoch": 0.7898925355794365,
+      "grad_norm": 0.44533100724220276,
+      "learning_rate": 0.0001,
+      "loss": 1.7044,
+      "step": 6799
+    },
+    {
+      "epoch": 0.790008713331397,
+      "grad_norm": 0.4789431393146515,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7901248910833575,
+      "grad_norm": 0.4464295506477356,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 6801
+    },
+    {
+      "epoch": 0.7902410688353181,
+      "grad_norm": 0.48397591710090637,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 6802
+    },
+    {
+      "epoch": 0.7903572465872786,
+      "grad_norm": 0.47774365544319153,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 6803
+    },
+    {
+      "epoch": 0.7904734243392391,
+      "grad_norm": 0.4871893525123596,
+      "learning_rate": 0.0001,
+      "loss": 1.6478,
+      "step": 6804
+    },
+    {
+      "epoch": 0.7905896020911996,
+      "grad_norm": 0.4551542103290558,
+      "learning_rate": 0.0001,
+      "loss": 1.8026,
+      "step": 6805
+    },
+    {
+      "epoch": 0.79070577984316,
+      "grad_norm": 0.4858977496623993,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 6806
+    },
+    {
+      "epoch": 0.7908219575951205,
+      "grad_norm": 0.4872024953365326,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 6807
+    },
+    {
+      "epoch": 0.790938135347081,
+      "grad_norm": 0.45094436407089233,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 6808
+    },
+    {
+      "epoch": 0.7910543130990415,
+      "grad_norm": 0.43320778012275696,
+      "learning_rate": 0.0001,
+      "loss": 1.6605,
+      "step": 6809
+    },
+    {
+      "epoch": 0.791170490851002,
+      "grad_norm": 0.4422098696231842,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7912866686029625,
+      "grad_norm": 0.44262856245040894,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 6811
+    },
+    {
+      "epoch": 0.7914028463549231,
+      "grad_norm": 0.4632859230041504,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 6812
+    },
+    {
+      "epoch": 0.7915190241068836,
+      "grad_norm": 0.4445875883102417,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 6813
+    },
+    {
+      "epoch": 0.7916352018588441,
+      "grad_norm": 0.47772374749183655,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 6814
+    },
+    {
+      "epoch": 0.7917513796108046,
+      "grad_norm": 0.4446466863155365,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 6815
+    },
+    {
+      "epoch": 0.791867557362765,
+      "grad_norm": 0.47620153427124023,
+      "learning_rate": 0.0001,
+      "loss": 1.7399,
+      "step": 6816
+    },
+    {
+      "epoch": 0.7919837351147255,
+      "grad_norm": 0.496855229139328,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 6817
+    },
+    {
+      "epoch": 0.792099912866686,
+      "grad_norm": 0.44331079721450806,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 6818
+    },
+    {
+      "epoch": 0.7922160906186465,
+      "grad_norm": 0.4731941223144531,
+      "learning_rate": 0.0001,
+      "loss": 1.6266,
+      "step": 6819
+    },
+    {
+      "epoch": 0.792332268370607,
+      "grad_norm": 0.507358729839325,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7924484461225675,
+      "grad_norm": 0.46159496903419495,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 6821
+    },
+    {
+      "epoch": 0.792564623874528,
+      "grad_norm": 0.4719959497451782,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 6822
+    },
+    {
+      "epoch": 0.7926808016264886,
+      "grad_norm": 0.4765816330909729,
+      "learning_rate": 0.0001,
+      "loss": 1.6144,
+      "step": 6823
+    },
+    {
+      "epoch": 0.7927969793784491,
+      "grad_norm": 0.43941232562065125,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 6824
+    },
+    {
+      "epoch": 0.7929131571304096,
+      "grad_norm": 0.47497496008872986,
+      "learning_rate": 0.0001,
+      "loss": 1.649,
+      "step": 6825
+    },
+    {
+      "epoch": 0.79302933488237,
+      "grad_norm": 0.43414729833602905,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 6826
+    },
+    {
+      "epoch": 0.7931455126343305,
+      "grad_norm": 0.48383721709251404,
+      "learning_rate": 0.0001,
+      "loss": 1.7554,
+      "step": 6827
+    },
+    {
+      "epoch": 0.793261690386291,
+      "grad_norm": 0.4651719033718109,
+      "learning_rate": 0.0001,
+      "loss": 1.5808,
+      "step": 6828
+    },
+    {
+      "epoch": 0.7933778681382515,
+      "grad_norm": 0.42594683170318604,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 6829
+    },
+    {
+      "epoch": 0.793494045890212,
+      "grad_norm": 0.4392074644565582,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7936102236421725,
+      "grad_norm": 0.43708086013793945,
+      "learning_rate": 0.0001,
+      "loss": 1.3342,
+      "step": 6831
+    },
+    {
+      "epoch": 0.793726401394133,
+      "grad_norm": 0.46385717391967773,
+      "learning_rate": 0.0001,
+      "loss": 1.709,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7938425791460936,
+      "grad_norm": 0.4499506652355194,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 6833
+    },
+    {
+      "epoch": 0.7939587568980541,
+      "grad_norm": 0.46915948390960693,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7940749346500146,
+      "grad_norm": 0.4685263931751251,
+      "learning_rate": 0.0001,
+      "loss": 1.6678,
+      "step": 6835
+    },
+    {
+      "epoch": 0.794191112401975,
+      "grad_norm": 0.45696932077407837,
+      "learning_rate": 0.0001,
+      "loss": 1.6566,
+      "step": 6836
+    },
+    {
+      "epoch": 0.7943072901539355,
+      "grad_norm": 0.47869807481765747,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 6837
+    },
+    {
+      "epoch": 0.794423467905896,
+      "grad_norm": 0.45035579800605774,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7945396456578565,
+      "grad_norm": 0.48194530606269836,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 6839
+    },
+    {
+      "epoch": 0.794655823409817,
+      "grad_norm": 0.47445181012153625,
+      "learning_rate": 0.0001,
+      "loss": 1.7277,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7947720011617775,
+      "grad_norm": 0.4426630139350891,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 6841
+    },
+    {
+      "epoch": 0.794888178913738,
+      "grad_norm": 0.45886486768722534,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 6842
+    },
+    {
+      "epoch": 0.7950043566656985,
+      "grad_norm": 0.4601069986820221,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7951205344176591,
+      "grad_norm": 0.42290568351745605,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7952367121696196,
+      "grad_norm": 0.4962608516216278,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 6845
+    },
+    {
+      "epoch": 0.79535288992158,
+      "grad_norm": 0.47025734186172485,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 6846
+    },
+    {
+      "epoch": 0.7954690676735405,
+      "grad_norm": 0.441499263048172,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 6847
+    },
+    {
+      "epoch": 0.795585245425501,
+      "grad_norm": 0.4636131823062897,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 6848
+    },
+    {
+      "epoch": 0.7957014231774615,
+      "grad_norm": 0.46920451521873474,
+      "learning_rate": 0.0001,
+      "loss": 1.5591,
+      "step": 6849
+    },
+    {
+      "epoch": 0.795817600929422,
+      "grad_norm": 0.49390605092048645,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7959337786813825,
+      "grad_norm": 0.46050357818603516,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 6851
+    },
+    {
+      "epoch": 0.796049956433343,
+      "grad_norm": 0.4840867817401886,
+      "learning_rate": 0.0001,
+      "loss": 1.4195,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7961661341853035,
+      "grad_norm": 0.46949562430381775,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 6853
+    },
+    {
+      "epoch": 0.7962823119372641,
+      "grad_norm": 0.5006006360054016,
+      "learning_rate": 0.0001,
+      "loss": 1.8093,
+      "step": 6854
+    },
+    {
+      "epoch": 0.7963984896892246,
+      "grad_norm": 0.5085091590881348,
+      "learning_rate": 0.0001,
+      "loss": 1.8903,
+      "step": 6855
+    },
+    {
+      "epoch": 0.796514667441185,
+      "grad_norm": 0.4508131742477417,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 6856
+    },
+    {
+      "epoch": 0.7966308451931455,
+      "grad_norm": 0.4634523391723633,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 6857
+    },
+    {
+      "epoch": 0.796747022945106,
+      "grad_norm": 0.4478987753391266,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7968632006970665,
+      "grad_norm": 0.4543631374835968,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 6859
+    },
+    {
+      "epoch": 0.796979378449027,
+      "grad_norm": 0.48462975025177,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7970955562009875,
+      "grad_norm": 0.45995092391967773,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 6861
+    },
+    {
+      "epoch": 0.797211733952948,
+      "grad_norm": 0.48978012800216675,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 6862
+    },
+    {
+      "epoch": 0.7973279117049085,
+      "grad_norm": 0.45018213987350464,
+      "learning_rate": 0.0001,
+      "loss": 1.6156,
+      "step": 6863
+    },
+    {
+      "epoch": 0.797444089456869,
+      "grad_norm": 0.47435247898101807,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7975602672088296,
+      "grad_norm": 0.48207005858421326,
+      "learning_rate": 0.0001,
+      "loss": 1.5935,
+      "step": 6865
+    },
+    {
+      "epoch": 0.79767644496079,
+      "grad_norm": 0.4846184253692627,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 6866
+    },
+    {
+      "epoch": 0.7977926227127505,
+      "grad_norm": 0.42367056012153625,
+      "learning_rate": 0.0001,
+      "loss": 1.3894,
+      "step": 6867
+    },
+    {
+      "epoch": 0.797908800464711,
+      "grad_norm": 0.46958428621292114,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7980249782166715,
+      "grad_norm": 0.4629022777080536,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 6869
+    },
+    {
+      "epoch": 0.798141155968632,
+      "grad_norm": 0.5330944657325745,
+      "learning_rate": 0.0001,
+      "loss": 1.8337,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7982573337205925,
+      "grad_norm": 0.45015132427215576,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 6871
+    },
+    {
+      "epoch": 0.798373511472553,
+      "grad_norm": 0.447221964597702,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7984896892245135,
+      "grad_norm": 0.4698033928871155,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 6873
+    },
+    {
+      "epoch": 0.798605866976474,
+      "grad_norm": 0.4613732099533081,
+      "learning_rate": 0.0001,
+      "loss": 1.7136,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7987220447284346,
+      "grad_norm": 0.45149853825569153,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 6875
+    },
+    {
+      "epoch": 0.798838222480395,
+      "grad_norm": 0.44822239875793457,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7989544002323555,
+      "grad_norm": 0.4859652817249298,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 6877
+    },
+    {
+      "epoch": 0.799070577984316,
+      "grad_norm": 0.4371628165245056,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 6878
+    },
+    {
+      "epoch": 0.7991867557362765,
+      "grad_norm": 0.432081013917923,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 6879
+    },
+    {
+      "epoch": 0.799302933488237,
+      "grad_norm": 0.4589148461818695,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7994191112401975,
+      "grad_norm": 0.43646013736724854,
+      "learning_rate": 0.0001,
+      "loss": 1.5708,
+      "step": 6881
+    },
+    {
+      "epoch": 0.799535288992158,
+      "grad_norm": 0.42627108097076416,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 6882
+    },
+    {
+      "epoch": 0.7996514667441185,
+      "grad_norm": 0.4646453559398651,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 6883
+    },
+    {
+      "epoch": 0.799767644496079,
+      "grad_norm": 0.46586647629737854,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 6884
+    },
+    {
+      "epoch": 0.7998838222480396,
+      "grad_norm": 0.512482762336731,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 6885
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4631596803665161,
+      "learning_rate": 0.0001,
+      "loss": 1.6069,
+      "step": 6886
+    },
+    {
+      "epoch": 0.8001161777519605,
+      "grad_norm": 0.472889244556427,
+      "learning_rate": 0.0001,
+      "loss": 1.7365,
+      "step": 6887
+    },
+    {
+      "epoch": 0.800232355503921,
+      "grad_norm": 0.5098186731338501,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 6888
+    },
+    {
+      "epoch": 0.8003485332558815,
+      "grad_norm": 0.48305651545524597,
+      "learning_rate": 0.0001,
+      "loss": 1.777,
+      "step": 6889
+    },
+    {
+      "epoch": 0.800464711007842,
+      "grad_norm": 0.48298099637031555,
+      "learning_rate": 0.0001,
+      "loss": 1.7043,
+      "step": 6890
+    },
+    {
+      "epoch": 0.8005808887598025,
+      "grad_norm": 0.4996420741081238,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 6891
+    },
+    {
+      "epoch": 0.800697066511763,
+      "grad_norm": 0.44727784395217896,
+      "learning_rate": 0.0001,
+      "loss": 1.6145,
+      "step": 6892
+    },
+    {
+      "epoch": 0.8008132442637235,
+      "grad_norm": 0.4710618555545807,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 6893
+    },
+    {
+      "epoch": 0.800929422015684,
+      "grad_norm": 0.4605047404766083,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 6894
+    },
+    {
+      "epoch": 0.8010455997676444,
+      "grad_norm": 0.4915280342102051,
+      "learning_rate": 0.0001,
+      "loss": 1.6894,
+      "step": 6895
+    },
+    {
+      "epoch": 0.801161777519605,
+      "grad_norm": 0.45687055587768555,
+      "learning_rate": 0.0001,
+      "loss": 1.7078,
+      "step": 6896
+    },
+    {
+      "epoch": 0.8012779552715655,
+      "grad_norm": 0.4345172345638275,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 6897
+    },
+    {
+      "epoch": 0.801394133023526,
+      "grad_norm": 0.4714672863483429,
+      "learning_rate": 0.0001,
+      "loss": 1.4299,
+      "step": 6898
+    },
+    {
+      "epoch": 0.8015103107754865,
+      "grad_norm": 0.46883341670036316,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 6899
+    },
+    {
+      "epoch": 0.801626488527447,
+      "grad_norm": 0.43329375982284546,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 6900
+    },
+    {
+      "epoch": 0.8017426662794075,
+      "grad_norm": 0.5150987505912781,
+      "learning_rate": 0.0001,
+      "loss": 1.8174,
+      "step": 6901
+    },
+    {
+      "epoch": 0.801858844031368,
+      "grad_norm": 0.43868762254714966,
+      "learning_rate": 0.0001,
+      "loss": 1.5489,
+      "step": 6902
+    },
+    {
+      "epoch": 0.8019750217833285,
+      "grad_norm": 0.4465371072292328,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 6903
+    },
+    {
+      "epoch": 0.802091199535289,
+      "grad_norm": 0.4565187096595764,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 6904
+    },
+    {
+      "epoch": 0.8022073772872494,
+      "grad_norm": 0.5145065784454346,
+      "learning_rate": 0.0001,
+      "loss": 1.7764,
+      "step": 6905
+    },
+    {
+      "epoch": 0.80232355503921,
+      "grad_norm": 0.5177064538002014,
+      "learning_rate": 0.0001,
+      "loss": 1.6975,
+      "step": 6906
+    },
+    {
+      "epoch": 0.8024397327911705,
+      "grad_norm": 0.42869505286216736,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 6907
+    },
+    {
+      "epoch": 0.802555910543131,
+      "grad_norm": 0.4332997500896454,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 6908
+    },
+    {
+      "epoch": 0.8026720882950915,
+      "grad_norm": 0.4914308786392212,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 6909
+    },
+    {
+      "epoch": 0.802788266047052,
+      "grad_norm": 0.4771595299243927,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 6910
+    },
+    {
+      "epoch": 0.8029044437990125,
+      "grad_norm": 0.45163875818252563,
+      "learning_rate": 0.0001,
+      "loss": 1.7303,
+      "step": 6911
+    },
+    {
+      "epoch": 0.803020621550973,
+      "grad_norm": 0.4456539452075958,
+      "learning_rate": 0.0001,
+      "loss": 1.6847,
+      "step": 6912
+    },
+    {
+      "epoch": 0.8031367993029335,
+      "grad_norm": 0.42746564745903015,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 6913
+    },
+    {
+      "epoch": 0.803252977054894,
+      "grad_norm": 0.4743218421936035,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 6914
+    },
+    {
+      "epoch": 0.8033691548068544,
+      "grad_norm": 0.5129088759422302,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 6915
+    },
+    {
+      "epoch": 0.8034853325588149,
+      "grad_norm": 0.4551659822463989,
+      "learning_rate": 0.0001,
+      "loss": 1.6385,
+      "step": 6916
+    },
+    {
+      "epoch": 0.8036015103107755,
+      "grad_norm": 0.4753081202507019,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 6917
+    },
+    {
+      "epoch": 0.803717688062736,
+      "grad_norm": 0.4394998252391815,
+      "learning_rate": 0.0001,
+      "loss": 1.575,
+      "step": 6918
+    },
+    {
+      "epoch": 0.8038338658146965,
+      "grad_norm": 0.45924627780914307,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 6919
+    },
+    {
+      "epoch": 0.803950043566657,
+      "grad_norm": 0.44786056876182556,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 6920
+    },
+    {
+      "epoch": 0.8040662213186175,
+      "grad_norm": 0.4409099519252777,
+      "learning_rate": 0.0001,
+      "loss": 1.6561,
+      "step": 6921
+    },
+    {
+      "epoch": 0.804182399070578,
+      "grad_norm": 0.45234692096710205,
+      "learning_rate": 0.0001,
+      "loss": 1.5925,
+      "step": 6922
+    },
+    {
+      "epoch": 0.8042985768225385,
+      "grad_norm": 0.5276682376861572,
+      "learning_rate": 0.0001,
+      "loss": 1.7412,
+      "step": 6923
+    },
+    {
+      "epoch": 0.804414754574499,
+      "grad_norm": 0.4487822353839874,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 6924
+    },
+    {
+      "epoch": 0.8045309323264594,
+      "grad_norm": 0.4596274495124817,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 6925
+    },
+    {
+      "epoch": 0.8046471100784199,
+      "grad_norm": 0.44739964604377747,
+      "learning_rate": 0.0001,
+      "loss": 1.5391,
+      "step": 6926
+    },
+    {
+      "epoch": 0.8047632878303805,
+      "grad_norm": 0.46213769912719727,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 6927
+    },
+    {
+      "epoch": 0.804879465582341,
+      "grad_norm": 0.4504436254501343,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 6928
+    },
+    {
+      "epoch": 0.8049956433343015,
+      "grad_norm": 0.4626392722129822,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 6929
+    },
+    {
+      "epoch": 0.805111821086262,
+      "grad_norm": 0.4738352596759796,
+      "learning_rate": 0.0001,
+      "loss": 1.699,
+      "step": 6930
+    },
+    {
+      "epoch": 0.8052279988382225,
+      "grad_norm": 0.45543593168258667,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 6931
+    },
+    {
+      "epoch": 0.805344176590183,
+      "grad_norm": 0.4740353226661682,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 6932
+    },
+    {
+      "epoch": 0.8054603543421435,
+      "grad_norm": 0.4481174647808075,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 6933
+    },
+    {
+      "epoch": 0.805576532094104,
+      "grad_norm": 0.461093544960022,
+      "learning_rate": 0.0001,
+      "loss": 1.6113,
+      "step": 6934
+    },
+    {
+      "epoch": 0.8056927098460644,
+      "grad_norm": 0.4422491490840912,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 6935
+    },
+    {
+      "epoch": 0.8058088875980249,
+      "grad_norm": 0.4397140443325043,
+      "learning_rate": 0.0001,
+      "loss": 1.5065,
+      "step": 6936
+    },
+    {
+      "epoch": 0.8059250653499854,
+      "grad_norm": 0.47454312443733215,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 6937
+    },
+    {
+      "epoch": 0.806041243101946,
+      "grad_norm": 0.46560293436050415,
+      "learning_rate": 0.0001,
+      "loss": 1.6589,
+      "step": 6938
+    },
+    {
+      "epoch": 0.8061574208539065,
+      "grad_norm": 0.46185746788978577,
+      "learning_rate": 0.0001,
+      "loss": 1.6063,
+      "step": 6939
+    },
+    {
+      "epoch": 0.806273598605867,
+      "grad_norm": 0.4402056634426117,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 6940
+    },
+    {
+      "epoch": 0.8063897763578275,
+      "grad_norm": 0.452150821685791,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 6941
+    },
+    {
+      "epoch": 0.806505954109788,
+      "grad_norm": 0.4527072012424469,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 6942
+    },
+    {
+      "epoch": 0.8066221318617485,
+      "grad_norm": 0.4439299702644348,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 6943
+    },
+    {
+      "epoch": 0.806738309613709,
+      "grad_norm": 0.4649060070514679,
+      "learning_rate": 0.0001,
+      "loss": 1.7965,
+      "step": 6944
+    },
+    {
+      "epoch": 0.8068544873656694,
+      "grad_norm": 0.46530160307884216,
+      "learning_rate": 0.0001,
+      "loss": 1.7508,
+      "step": 6945
+    },
+    {
+      "epoch": 0.8069706651176299,
+      "grad_norm": 0.456967294216156,
+      "learning_rate": 0.0001,
+      "loss": 1.6391,
+      "step": 6946
+    },
+    {
+      "epoch": 0.8070868428695904,
+      "grad_norm": 0.47573670744895935,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 6947
+    },
+    {
+      "epoch": 0.807203020621551,
+      "grad_norm": 0.4661976099014282,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 6948
+    },
+    {
+      "epoch": 0.8073191983735115,
+      "grad_norm": 0.4907025694847107,
+      "learning_rate": 0.0001,
+      "loss": 1.6961,
+      "step": 6949
+    },
+    {
+      "epoch": 0.807435376125472,
+      "grad_norm": 0.4918571412563324,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 6950
+    },
+    {
+      "epoch": 0.8075515538774325,
+      "grad_norm": 0.48200660943984985,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 6951
+    },
+    {
+      "epoch": 0.807667731629393,
+      "grad_norm": 0.42808154225349426,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 6952
+    },
+    {
+      "epoch": 0.8077839093813535,
+      "grad_norm": 0.473086416721344,
+      "learning_rate": 0.0001,
+      "loss": 1.7153,
+      "step": 6953
+    },
+    {
+      "epoch": 0.807900087133314,
+      "grad_norm": 0.44156551361083984,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 6954
+    },
+    {
+      "epoch": 0.8080162648852744,
+      "grad_norm": 0.5178338289260864,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 6955
+    },
+    {
+      "epoch": 0.8081324426372349,
+      "grad_norm": 0.4564967751502991,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 6956
+    },
+    {
+      "epoch": 0.8082486203891954,
+      "grad_norm": 0.44008997082710266,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 6957
+    },
+    {
+      "epoch": 0.8083647981411559,
+      "grad_norm": 0.4413783848285675,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 6958
+    },
+    {
+      "epoch": 0.8084809758931165,
+      "grad_norm": 0.4661567807197571,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 6959
+    },
+    {
+      "epoch": 0.808597153645077,
+      "grad_norm": 0.47281137108802795,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 6960
+    },
+    {
+      "epoch": 0.8087133313970375,
+      "grad_norm": 0.47660407423973083,
+      "learning_rate": 0.0001,
+      "loss": 1.7855,
+      "step": 6961
+    },
+    {
+      "epoch": 0.808829509148998,
+      "grad_norm": 0.4914090633392334,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 6962
+    },
+    {
+      "epoch": 0.8089456869009585,
+      "grad_norm": 0.4621322453022003,
+      "learning_rate": 0.0001,
+      "loss": 1.5594,
+      "step": 6963
+    },
+    {
+      "epoch": 0.809061864652919,
+      "grad_norm": 0.4495106339454651,
+      "learning_rate": 0.0001,
+      "loss": 1.6727,
+      "step": 6964
+    },
+    {
+      "epoch": 0.8091780424048794,
+      "grad_norm": 0.49163323640823364,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 6965
+    },
+    {
+      "epoch": 0.8092942201568399,
+      "grad_norm": 0.47235107421875,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 6966
+    },
+    {
+      "epoch": 0.8094103979088004,
+      "grad_norm": 0.45100662112236023,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 6967
+    },
+    {
+      "epoch": 0.8095265756607609,
+      "grad_norm": 0.46376436948776245,
+      "learning_rate": 0.0001,
+      "loss": 1.7342,
+      "step": 6968
+    },
+    {
+      "epoch": 0.8096427534127215,
+      "grad_norm": 0.490784227848053,
+      "learning_rate": 0.0001,
+      "loss": 1.6433,
+      "step": 6969
+    },
+    {
+      "epoch": 0.809758931164682,
+      "grad_norm": 0.49541133642196655,
+      "learning_rate": 0.0001,
+      "loss": 1.7031,
+      "step": 6970
+    },
+    {
+      "epoch": 0.8098751089166425,
+      "grad_norm": 0.47032544016838074,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 6971
+    },
+    {
+      "epoch": 0.809991286668603,
+      "grad_norm": 0.4594431519508362,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 6972
+    },
+    {
+      "epoch": 0.8101074644205635,
+      "grad_norm": 0.4627406895160675,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 6973
+    },
+    {
+      "epoch": 0.810223642172524,
+      "grad_norm": 0.4937112331390381,
+      "learning_rate": 0.0001,
+      "loss": 1.7939,
+      "step": 6974
+    },
+    {
+      "epoch": 0.8103398199244845,
+      "grad_norm": 0.4783715605735779,
+      "learning_rate": 0.0001,
+      "loss": 1.6928,
+      "step": 6975
+    },
+    {
+      "epoch": 0.8104559976764449,
+      "grad_norm": 0.47671520709991455,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 6976
+    },
+    {
+      "epoch": 0.8105721754284054,
+      "grad_norm": 0.47149521112442017,
+      "learning_rate": 0.0001,
+      "loss": 1.6648,
+      "step": 6977
+    },
+    {
+      "epoch": 0.8106883531803659,
+      "grad_norm": 0.45699021220207214,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 6978
+    },
+    {
+      "epoch": 0.8108045309323264,
+      "grad_norm": 0.4693375527858734,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 6979
+    },
+    {
+      "epoch": 0.810920708684287,
+      "grad_norm": 0.4451500475406647,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 6980
+    },
+    {
+      "epoch": 0.8110368864362475,
+      "grad_norm": 0.45721450448036194,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 6981
+    },
+    {
+      "epoch": 0.811153064188208,
+      "grad_norm": 0.44827860593795776,
+      "learning_rate": 0.0001,
+      "loss": 1.6295,
+      "step": 6982
+    },
+    {
+      "epoch": 0.8112692419401685,
+      "grad_norm": 0.47101083397865295,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 6983
+    },
+    {
+      "epoch": 0.811385419692129,
+      "grad_norm": 0.4695024788379669,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 6984
+    },
+    {
+      "epoch": 0.8115015974440895,
+      "grad_norm": 0.5033894777297974,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 6985
+    },
+    {
+      "epoch": 0.8116177751960499,
+      "grad_norm": 0.49278390407562256,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 6986
+    },
+    {
+      "epoch": 0.8117339529480104,
+      "grad_norm": 0.5109890699386597,
+      "learning_rate": 0.0001,
+      "loss": 1.6748,
+      "step": 6987
+    },
+    {
+      "epoch": 0.8118501306999709,
+      "grad_norm": 0.4835306704044342,
+      "learning_rate": 0.0001,
+      "loss": 1.7316,
+      "step": 6988
+    },
+    {
+      "epoch": 0.8119663084519314,
+      "grad_norm": 0.48664847016334534,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 6989
+    },
+    {
+      "epoch": 0.812082486203892,
+      "grad_norm": 0.4511289596557617,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 6990
+    },
+    {
+      "epoch": 0.8121986639558525,
+      "grad_norm": 0.46078622341156006,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 6991
+    },
+    {
+      "epoch": 0.812314841707813,
+      "grad_norm": 0.5006919503211975,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 6992
+    },
+    {
+      "epoch": 0.8124310194597735,
+      "grad_norm": 0.46968936920166016,
+      "learning_rate": 0.0001,
+      "loss": 1.7114,
+      "step": 6993
+    },
+    {
+      "epoch": 0.812547197211734,
+      "grad_norm": 0.46966198086738586,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 6994
+    },
+    {
+      "epoch": 0.8126633749636945,
+      "grad_norm": 0.5085934400558472,
+      "learning_rate": 0.0001,
+      "loss": 1.7453,
+      "step": 6995
+    },
+    {
+      "epoch": 0.8127795527156549,
+      "grad_norm": 0.42895692586898804,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 6996
+    },
+    {
+      "epoch": 0.8128957304676154,
+      "grad_norm": 0.46299561858177185,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 6997
+    },
+    {
+      "epoch": 0.8130119082195759,
+      "grad_norm": 0.4561334550380707,
+      "learning_rate": 0.0001,
+      "loss": 1.5164,
+      "step": 6998
+    },
+    {
+      "epoch": 0.8131280859715364,
+      "grad_norm": 0.49617594480514526,
+      "learning_rate": 0.0001,
+      "loss": 1.7706,
+      "step": 6999
+    },
+    {
+      "epoch": 0.8132442637234969,
+      "grad_norm": 0.4899764955043793,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 7000
+    },
+    {
+      "epoch": 0.8133604414754575,
+      "grad_norm": 0.4907947778701782,
+      "learning_rate": 0.0001,
+      "loss": 1.7775,
+      "step": 7001
+    },
+    {
+      "epoch": 0.813476619227418,
+      "grad_norm": 0.45711711049079895,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 7002
+    },
+    {
+      "epoch": 0.8135927969793785,
+      "grad_norm": 0.449201375246048,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 7003
+    },
+    {
+      "epoch": 0.813708974731339,
+      "grad_norm": 0.4838063418865204,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 7004
+    },
+    {
+      "epoch": 0.8138251524832995,
+      "grad_norm": 0.46976253390312195,
+      "learning_rate": 0.0001,
+      "loss": 1.6453,
+      "step": 7005
+    },
+    {
+      "epoch": 0.8139413302352599,
+      "grad_norm": 0.4605077803134918,
+      "learning_rate": 0.0001,
+      "loss": 1.6659,
+      "step": 7006
+    },
+    {
+      "epoch": 0.8140575079872204,
+      "grad_norm": 0.4310915470123291,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 7007
+    },
+    {
+      "epoch": 0.8141736857391809,
+      "grad_norm": 0.49787524342536926,
+      "learning_rate": 0.0001,
+      "loss": 1.6964,
+      "step": 7008
+    },
+    {
+      "epoch": 0.8142898634911414,
+      "grad_norm": 0.4633225202560425,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 7009
+    },
+    {
+      "epoch": 0.8144060412431019,
+      "grad_norm": 0.5344363451004028,
+      "learning_rate": 0.0001,
+      "loss": 1.8373,
+      "step": 7010
+    },
+    {
+      "epoch": 0.8145222189950625,
+      "grad_norm": 0.4966965615749359,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 7011
+    },
+    {
+      "epoch": 0.814638396747023,
+      "grad_norm": 0.4843759834766388,
+      "learning_rate": 0.0001,
+      "loss": 1.8051,
+      "step": 7012
+    },
+    {
+      "epoch": 0.8147545744989835,
+      "grad_norm": 0.46458831429481506,
+      "learning_rate": 0.0001,
+      "loss": 1.4644,
+      "step": 7013
+    },
+    {
+      "epoch": 0.814870752250944,
+      "grad_norm": 0.46197426319122314,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 7014
+    },
+    {
+      "epoch": 0.8149869300029045,
+      "grad_norm": 0.4449380934238434,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 7015
+    },
+    {
+      "epoch": 0.8151031077548649,
+      "grad_norm": 0.47970300912857056,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 7016
+    },
+    {
+      "epoch": 0.8152192855068254,
+      "grad_norm": 0.48018476366996765,
+      "learning_rate": 0.0001,
+      "loss": 1.7756,
+      "step": 7017
+    },
+    {
+      "epoch": 0.8153354632587859,
+      "grad_norm": 0.45871859788894653,
+      "learning_rate": 0.0001,
+      "loss": 1.7081,
+      "step": 7018
+    },
+    {
+      "epoch": 0.8154516410107464,
+      "grad_norm": 0.509146511554718,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 7019
+    },
+    {
+      "epoch": 0.8155678187627069,
+      "grad_norm": 0.4545764625072479,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 7020
+    },
+    {
+      "epoch": 0.8156839965146674,
+      "grad_norm": 0.46723830699920654,
+      "learning_rate": 0.0001,
+      "loss": 1.5216,
+      "step": 7021
+    },
+    {
+      "epoch": 0.815800174266628,
+      "grad_norm": 0.4666821360588074,
+      "learning_rate": 0.0001,
+      "loss": 1.6664,
+      "step": 7022
+    },
+    {
+      "epoch": 0.8159163520185885,
+      "grad_norm": 0.45362234115600586,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 7023
+    },
+    {
+      "epoch": 0.816032529770549,
+      "grad_norm": 0.5081374049186707,
+      "learning_rate": 0.0001,
+      "loss": 1.7431,
+      "step": 7024
+    },
+    {
+      "epoch": 0.8161487075225095,
+      "grad_norm": 0.4551977515220642,
+      "learning_rate": 0.0001,
+      "loss": 1.4208,
+      "step": 7025
+    },
+    {
+      "epoch": 0.8162648852744699,
+      "grad_norm": 0.4469984471797943,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 7026
+    },
+    {
+      "epoch": 0.8163810630264304,
+      "grad_norm": 0.4428351819515228,
+      "learning_rate": 0.0001,
+      "loss": 1.5317,
+      "step": 7027
+    },
+    {
+      "epoch": 0.8164972407783909,
+      "grad_norm": 0.46681028604507446,
+      "learning_rate": 0.0001,
+      "loss": 1.7313,
+      "step": 7028
+    },
+    {
+      "epoch": 0.8166134185303514,
+      "grad_norm": 0.47523221373558044,
+      "learning_rate": 0.0001,
+      "loss": 1.5623,
+      "step": 7029
+    },
+    {
+      "epoch": 0.8167295962823119,
+      "grad_norm": 0.42008304595947266,
+      "learning_rate": 0.0001,
+      "loss": 1.2919,
+      "step": 7030
+    },
+    {
+      "epoch": 0.8168457740342724,
+      "grad_norm": 0.4836232364177704,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 7031
+    },
+    {
+      "epoch": 0.816961951786233,
+      "grad_norm": 0.49851134419441223,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 7032
+    },
+    {
+      "epoch": 0.8170781295381935,
+      "grad_norm": 0.4806590974330902,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 7033
+    },
+    {
+      "epoch": 0.817194307290154,
+      "grad_norm": 0.46328291296958923,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 7034
+    },
+    {
+      "epoch": 0.8173104850421145,
+      "grad_norm": 0.44353359937667847,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 7035
+    },
+    {
+      "epoch": 0.8174266627940749,
+      "grad_norm": 0.5083478689193726,
+      "learning_rate": 0.0001,
+      "loss": 1.7822,
+      "step": 7036
+    },
+    {
+      "epoch": 0.8175428405460354,
+      "grad_norm": 0.4776815176010132,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 7037
+    },
+    {
+      "epoch": 0.8176590182979959,
+      "grad_norm": 0.43044784665107727,
+      "learning_rate": 0.0001,
+      "loss": 1.3441,
+      "step": 7038
+    },
+    {
+      "epoch": 0.8177751960499564,
+      "grad_norm": 0.4387872815132141,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 7039
+    },
+    {
+      "epoch": 0.8178913738019169,
+      "grad_norm": 0.46325990557670593,
+      "learning_rate": 0.0001,
+      "loss": 1.7339,
+      "step": 7040
+    },
+    {
+      "epoch": 0.8180075515538774,
+      "grad_norm": 0.4589924216270447,
+      "learning_rate": 0.0001,
+      "loss": 1.6567,
+      "step": 7041
+    },
+    {
+      "epoch": 0.8181237293058379,
+      "grad_norm": 0.4716587960720062,
+      "learning_rate": 0.0001,
+      "loss": 1.6844,
+      "step": 7042
+    },
+    {
+      "epoch": 0.8182399070577985,
+      "grad_norm": 0.4652487635612488,
+      "learning_rate": 0.0001,
+      "loss": 1.7618,
+      "step": 7043
+    },
+    {
+      "epoch": 0.818356084809759,
+      "grad_norm": 0.4787328839302063,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 7044
+    },
+    {
+      "epoch": 0.8184722625617195,
+      "grad_norm": 0.47911664843559265,
+      "learning_rate": 0.0001,
+      "loss": 1.6959,
+      "step": 7045
+    },
+    {
+      "epoch": 0.8185884403136799,
+      "grad_norm": 0.4904523491859436,
+      "learning_rate": 0.0001,
+      "loss": 1.7353,
+      "step": 7046
+    },
+    {
+      "epoch": 0.8187046180656404,
+      "grad_norm": 0.4955161213874817,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 7047
+    },
+    {
+      "epoch": 0.8188207958176009,
+      "grad_norm": 0.44562825560569763,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 7048
+    },
+    {
+      "epoch": 0.8189369735695614,
+      "grad_norm": 0.44809067249298096,
+      "learning_rate": 0.0001,
+      "loss": 1.3182,
+      "step": 7049
+    },
+    {
+      "epoch": 0.8190531513215219,
+      "grad_norm": 0.4440357983112335,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 7050
+    },
+    {
+      "epoch": 0.8191693290734824,
+      "grad_norm": 0.4690278172492981,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 7051
+    },
+    {
+      "epoch": 0.8192855068254429,
+      "grad_norm": 0.4551823139190674,
+      "learning_rate": 0.0001,
+      "loss": 1.6531,
+      "step": 7052
+    },
+    {
+      "epoch": 0.8194016845774035,
+      "grad_norm": 0.48354125022888184,
+      "learning_rate": 0.0001,
+      "loss": 1.7015,
+      "step": 7053
+    },
+    {
+      "epoch": 0.819517862329364,
+      "grad_norm": 0.4836772084236145,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 7054
+    },
+    {
+      "epoch": 0.8196340400813245,
+      "grad_norm": 0.4699805974960327,
+      "learning_rate": 0.0001,
+      "loss": 1.6777,
+      "step": 7055
+    },
+    {
+      "epoch": 0.8197502178332849,
+      "grad_norm": 0.41965410113334656,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 7056
+    },
+    {
+      "epoch": 0.8198663955852454,
+      "grad_norm": 0.5057850480079651,
+      "learning_rate": 0.0001,
+      "loss": 1.7782,
+      "step": 7057
+    },
+    {
+      "epoch": 0.8199825733372059,
+      "grad_norm": 0.4643802046775818,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 7058
+    },
+    {
+      "epoch": 0.8200987510891664,
+      "grad_norm": 0.4550783932209015,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 7059
+    },
+    {
+      "epoch": 0.8202149288411269,
+      "grad_norm": 0.5149709582328796,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 7060
+    },
+    {
+      "epoch": 0.8203311065930874,
+      "grad_norm": 0.46510255336761475,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 7061
+    },
+    {
+      "epoch": 0.8204472843450479,
+      "grad_norm": 0.46841713786125183,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 7062
+    },
+    {
+      "epoch": 0.8205634620970085,
+      "grad_norm": 0.4657427966594696,
+      "learning_rate": 0.0001,
+      "loss": 1.6439,
+      "step": 7063
+    },
+    {
+      "epoch": 0.820679639848969,
+      "grad_norm": 0.43857282400131226,
+      "learning_rate": 0.0001,
+      "loss": 1.3911,
+      "step": 7064
+    },
+    {
+      "epoch": 0.8207958176009295,
+      "grad_norm": 0.457220196723938,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 7065
+    },
+    {
+      "epoch": 0.82091199535289,
+      "grad_norm": 0.462664932012558,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 7066
+    },
+    {
+      "epoch": 0.8210281731048504,
+      "grad_norm": 0.5375829339027405,
+      "learning_rate": 0.0001,
+      "loss": 1.6802,
+      "step": 7067
+    },
+    {
+      "epoch": 0.8211443508568109,
+      "grad_norm": 0.47955790162086487,
+      "learning_rate": 0.0001,
+      "loss": 1.6042,
+      "step": 7068
+    },
+    {
+      "epoch": 0.8212605286087714,
+      "grad_norm": 0.4525638818740845,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 7069
+    },
+    {
+      "epoch": 0.8213767063607319,
+      "grad_norm": 0.6293335556983948,
+      "learning_rate": 0.0001,
+      "loss": 1.7643,
+      "step": 7070
+    },
+    {
+      "epoch": 0.8214928841126924,
+      "grad_norm": 0.46683555841445923,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 7071
+    },
+    {
+      "epoch": 0.8216090618646529,
+      "grad_norm": 0.4560864269733429,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 7072
+    },
+    {
+      "epoch": 0.8217252396166134,
+      "grad_norm": 0.4356245696544647,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 7073
+    },
+    {
+      "epoch": 0.821841417368574,
+      "grad_norm": 0.4432215988636017,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 7074
+    },
+    {
+      "epoch": 0.8219575951205345,
+      "grad_norm": 0.43178898096084595,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 7075
+    },
+    {
+      "epoch": 0.822073772872495,
+      "grad_norm": 0.45065784454345703,
+      "learning_rate": 0.0001,
+      "loss": 1.5992,
+      "step": 7076
+    },
+    {
+      "epoch": 0.8221899506244554,
+      "grad_norm": 0.44402584433555603,
+      "learning_rate": 0.0001,
+      "loss": 1.645,
+      "step": 7077
+    },
+    {
+      "epoch": 0.8223061283764159,
+      "grad_norm": 0.4661881625652313,
+      "learning_rate": 0.0001,
+      "loss": 1.636,
+      "step": 7078
+    },
+    {
+      "epoch": 0.8224223061283764,
+      "grad_norm": 0.46382853388786316,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 7079
+    },
+    {
+      "epoch": 0.8225384838803369,
+      "grad_norm": 0.4864356219768524,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 7080
+    },
+    {
+      "epoch": 0.8226546616322974,
+      "grad_norm": 0.48990336060523987,
+      "learning_rate": 0.0001,
+      "loss": 1.7913,
+      "step": 7081
+    },
+    {
+      "epoch": 0.8227708393842579,
+      "grad_norm": 0.4853106141090393,
+      "learning_rate": 0.0001,
+      "loss": 1.8482,
+      "step": 7082
+    },
+    {
+      "epoch": 0.8228870171362184,
+      "grad_norm": 0.42953288555145264,
+      "learning_rate": 0.0001,
+      "loss": 1.3854,
+      "step": 7083
+    },
+    {
+      "epoch": 0.823003194888179,
+      "grad_norm": 0.4772704541683197,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 7084
+    },
+    {
+      "epoch": 0.8231193726401395,
+      "grad_norm": 0.4406863749027252,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 7085
+    },
+    {
+      "epoch": 0.8232355503921,
+      "grad_norm": 0.47718164324760437,
+      "learning_rate": 0.0001,
+      "loss": 1.6784,
+      "step": 7086
+    },
+    {
+      "epoch": 0.8233517281440604,
+      "grad_norm": 0.4515807628631592,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 7087
+    },
+    {
+      "epoch": 0.8234679058960209,
+      "grad_norm": 0.4711925685405731,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 7088
+    },
+    {
+      "epoch": 0.8235840836479814,
+      "grad_norm": 0.4708985984325409,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 7089
+    },
+    {
+      "epoch": 0.8237002613999419,
+      "grad_norm": 0.48707452416419983,
+      "learning_rate": 0.0001,
+      "loss": 1.5098,
+      "step": 7090
+    },
+    {
+      "epoch": 0.8238164391519024,
+      "grad_norm": 0.5250394940376282,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 7091
+    },
+    {
+      "epoch": 0.8239326169038629,
+      "grad_norm": 0.4977336525917053,
+      "learning_rate": 0.0001,
+      "loss": 1.6806,
+      "step": 7092
+    },
+    {
+      "epoch": 0.8240487946558234,
+      "grad_norm": 0.4466916024684906,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 7093
+    },
+    {
+      "epoch": 0.8241649724077839,
+      "grad_norm": 0.4977428913116455,
+      "learning_rate": 0.0001,
+      "loss": 1.6532,
+      "step": 7094
+    },
+    {
+      "epoch": 0.8242811501597445,
+      "grad_norm": 0.4770013093948364,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 7095
+    },
+    {
+      "epoch": 0.824397327911705,
+      "grad_norm": 0.4783066511154175,
+      "learning_rate": 0.0001,
+      "loss": 1.6102,
+      "step": 7096
+    },
+    {
+      "epoch": 0.8245135056636654,
+      "grad_norm": 0.44045940041542053,
+      "learning_rate": 0.0001,
+      "loss": 1.4502,
+      "step": 7097
+    },
+    {
+      "epoch": 0.8246296834156259,
+      "grad_norm": 0.449008584022522,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 7098
+    },
+    {
+      "epoch": 0.8247458611675864,
+      "grad_norm": 0.45892423391342163,
+      "learning_rate": 0.0001,
+      "loss": 1.6121,
+      "step": 7099
+    },
+    {
+      "epoch": 0.8248620389195469,
+      "grad_norm": 0.4626208245754242,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 7100
+    },
+    {
+      "epoch": 0.8249782166715074,
+      "grad_norm": 0.5038914084434509,
+      "learning_rate": 0.0001,
+      "loss": 1.651,
+      "step": 7101
+    },
+    {
+      "epoch": 0.8250943944234679,
+      "grad_norm": 0.4759518802165985,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 7102
+    },
+    {
+      "epoch": 0.8252105721754284,
+      "grad_norm": 0.48517483472824097,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 7103
+    },
+    {
+      "epoch": 0.8253267499273889,
+      "grad_norm": 0.47098508477211,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 7104
+    },
+    {
+      "epoch": 0.8254429276793495,
+      "grad_norm": 0.47671204805374146,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 7105
+    },
+    {
+      "epoch": 0.82555910543131,
+      "grad_norm": 0.4556223750114441,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 7106
+    },
+    {
+      "epoch": 0.8256752831832704,
+      "grad_norm": 0.47399070858955383,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 7107
+    },
+    {
+      "epoch": 0.8257914609352309,
+      "grad_norm": 0.4748283624649048,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 7108
+    },
+    {
+      "epoch": 0.8259076386871914,
+      "grad_norm": 0.48463886976242065,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 7109
+    },
+    {
+      "epoch": 0.8260238164391519,
+      "grad_norm": 0.46149513125419617,
+      "learning_rate": 0.0001,
+      "loss": 1.4978,
+      "step": 7110
+    },
+    {
+      "epoch": 0.8261399941911124,
+      "grad_norm": 0.47687479853630066,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 7111
+    },
+    {
+      "epoch": 0.8262561719430729,
+      "grad_norm": 0.46911653876304626,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 7112
+    },
+    {
+      "epoch": 0.8263723496950334,
+      "grad_norm": 0.5078893899917603,
+      "learning_rate": 0.0001,
+      "loss": 1.677,
+      "step": 7113
+    },
+    {
+      "epoch": 0.8264885274469939,
+      "grad_norm": 0.455218642950058,
+      "learning_rate": 0.0001,
+      "loss": 1.6192,
+      "step": 7114
+    },
+    {
+      "epoch": 0.8266047051989543,
+      "grad_norm": 0.44931599497795105,
+      "learning_rate": 0.0001,
+      "loss": 1.5572,
+      "step": 7115
+    },
+    {
+      "epoch": 0.826720882950915,
+      "grad_norm": 0.50384521484375,
+      "learning_rate": 0.0001,
+      "loss": 1.7711,
+      "step": 7116
+    },
+    {
+      "epoch": 0.8268370607028754,
+      "grad_norm": 0.48653748631477356,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 7117
+    },
+    {
+      "epoch": 0.8269532384548359,
+      "grad_norm": 0.44801607728004456,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 7118
+    },
+    {
+      "epoch": 0.8270694162067964,
+      "grad_norm": 0.45984333753585815,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 7119
+    },
+    {
+      "epoch": 0.8271855939587569,
+      "grad_norm": 0.49128204584121704,
+      "learning_rate": 0.0001,
+      "loss": 1.6582,
+      "step": 7120
+    },
+    {
+      "epoch": 0.8273017717107174,
+      "grad_norm": 0.47482722997665405,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 7121
+    },
+    {
+      "epoch": 0.8274179494626779,
+      "grad_norm": 0.46835750341415405,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 7122
+    },
+    {
+      "epoch": 0.8275341272146384,
+      "grad_norm": 0.4840836524963379,
+      "learning_rate": 0.0001,
+      "loss": 1.7906,
+      "step": 7123
+    },
+    {
+      "epoch": 0.8276503049665989,
+      "grad_norm": 0.4466482102870941,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 7124
+    },
+    {
+      "epoch": 0.8277664827185593,
+      "grad_norm": 0.47686582803726196,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 7125
+    },
+    {
+      "epoch": 0.82788266047052,
+      "grad_norm": 0.46509337425231934,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 7126
+    },
+    {
+      "epoch": 0.8279988382224804,
+      "grad_norm": 0.44292381405830383,
+      "learning_rate": 0.0001,
+      "loss": 1.6446,
+      "step": 7127
+    },
+    {
+      "epoch": 0.8281150159744409,
+      "grad_norm": 0.42537015676498413,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 7128
+    },
+    {
+      "epoch": 0.8282311937264014,
+      "grad_norm": 0.45168519020080566,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 7129
+    },
+    {
+      "epoch": 0.8283473714783619,
+      "grad_norm": 0.48862767219543457,
+      "learning_rate": 0.0001,
+      "loss": 1.632,
+      "step": 7130
+    },
+    {
+      "epoch": 0.8284635492303224,
+      "grad_norm": 0.4335499405860901,
+      "learning_rate": 0.0001,
+      "loss": 1.3461,
+      "step": 7131
+    },
+    {
+      "epoch": 0.8285797269822829,
+      "grad_norm": 0.4990119934082031,
+      "learning_rate": 0.0001,
+      "loss": 1.7773,
+      "step": 7132
+    },
+    {
+      "epoch": 0.8286959047342434,
+      "grad_norm": 0.4781734347343445,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 7133
+    },
+    {
+      "epoch": 0.8288120824862039,
+      "grad_norm": 0.46874862909317017,
+      "learning_rate": 0.0001,
+      "loss": 1.6284,
+      "step": 7134
+    },
+    {
+      "epoch": 0.8289282602381643,
+      "grad_norm": 0.49614983797073364,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 7135
+    },
+    {
+      "epoch": 0.8290444379901248,
+      "grad_norm": 0.45961683988571167,
+      "learning_rate": 0.0001,
+      "loss": 1.5863,
+      "step": 7136
+    },
+    {
+      "epoch": 0.8291606157420854,
+      "grad_norm": 0.42698389291763306,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 7137
+    },
+    {
+      "epoch": 0.8292767934940459,
+      "grad_norm": 0.5121428370475769,
+      "learning_rate": 0.0001,
+      "loss": 1.6586,
+      "step": 7138
+    },
+    {
+      "epoch": 0.8293929712460064,
+      "grad_norm": 0.45941177010536194,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 7139
+    },
+    {
+      "epoch": 0.8295091489979669,
+      "grad_norm": 0.487868994474411,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8296253267499274,
+      "grad_norm": 0.49762672185897827,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 7141
+    },
+    {
+      "epoch": 0.8297415045018879,
+      "grad_norm": 0.4626729488372803,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 7142
+    },
+    {
+      "epoch": 0.8298576822538484,
+      "grad_norm": 0.4500117301940918,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 7143
+    },
+    {
+      "epoch": 0.8299738600058089,
+      "grad_norm": 0.45620059967041016,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 7144
+    },
+    {
+      "epoch": 0.8300900377577694,
+      "grad_norm": 0.48835253715515137,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 7145
+    },
+    {
+      "epoch": 0.8302062155097298,
+      "grad_norm": 0.468001127243042,
+      "learning_rate": 0.0001,
+      "loss": 1.6553,
+      "step": 7146
+    },
+    {
+      "epoch": 0.8303223932616904,
+      "grad_norm": 0.45825132727622986,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 7147
+    },
+    {
+      "epoch": 0.8304385710136509,
+      "grad_norm": 0.46828025579452515,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 7148
+    },
+    {
+      "epoch": 0.8305547487656114,
+      "grad_norm": 0.49574530124664307,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 7149
+    },
+    {
+      "epoch": 0.8306709265175719,
+      "grad_norm": 0.4577179551124573,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8307871042695324,
+      "grad_norm": 0.47962459921836853,
+      "learning_rate": 0.0001,
+      "loss": 1.622,
+      "step": 7151
+    },
+    {
+      "epoch": 0.8309032820214929,
+      "grad_norm": 0.5886511206626892,
+      "learning_rate": 0.0001,
+      "loss": 1.6682,
+      "step": 7152
+    },
+    {
+      "epoch": 0.8310194597734534,
+      "grad_norm": 0.5050860047340393,
+      "learning_rate": 0.0001,
+      "loss": 1.5805,
+      "step": 7153
+    },
+    {
+      "epoch": 0.8311356375254139,
+      "grad_norm": 0.48599928617477417,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 7154
+    },
+    {
+      "epoch": 0.8312518152773744,
+      "grad_norm": 0.4886664152145386,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 7155
+    },
+    {
+      "epoch": 0.8313679930293348,
+      "grad_norm": 0.502704381942749,
+      "learning_rate": 0.0001,
+      "loss": 1.6681,
+      "step": 7156
+    },
+    {
+      "epoch": 0.8314841707812953,
+      "grad_norm": 0.4662174880504608,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 7157
+    },
+    {
+      "epoch": 0.8316003485332559,
+      "grad_norm": 0.5194843411445618,
+      "learning_rate": 0.0001,
+      "loss": 1.8369,
+      "step": 7158
+    },
+    {
+      "epoch": 0.8317165262852164,
+      "grad_norm": 0.458987832069397,
+      "learning_rate": 0.0001,
+      "loss": 1.738,
+      "step": 7159
+    },
+    {
+      "epoch": 0.8318327040371769,
+      "grad_norm": 0.4516839385032654,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8319488817891374,
+      "grad_norm": 0.47868669033050537,
+      "learning_rate": 0.0001,
+      "loss": 1.4402,
+      "step": 7161
+    },
+    {
+      "epoch": 0.8320650595410979,
+      "grad_norm": 0.4523623585700989,
+      "learning_rate": 0.0001,
+      "loss": 1.7252,
+      "step": 7162
+    },
+    {
+      "epoch": 0.8321812372930584,
+      "grad_norm": 0.4759959876537323,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 7163
+    },
+    {
+      "epoch": 0.8322974150450189,
+      "grad_norm": 0.47161024808883667,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 7164
+    },
+    {
+      "epoch": 0.8324135927969794,
+      "grad_norm": 0.43895629048347473,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 7165
+    },
+    {
+      "epoch": 0.8325297705489398,
+      "grad_norm": 0.47241824865341187,
+      "learning_rate": 0.0001,
+      "loss": 1.4324,
+      "step": 7166
+    },
+    {
+      "epoch": 0.8326459483009003,
+      "grad_norm": 0.43697863817214966,
+      "learning_rate": 0.0001,
+      "loss": 1.4951,
+      "step": 7167
+    },
+    {
+      "epoch": 0.8327621260528609,
+      "grad_norm": 0.5089771747589111,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 7168
+    },
+    {
+      "epoch": 0.8328783038048214,
+      "grad_norm": 0.47586750984191895,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 7169
+    },
+    {
+      "epoch": 0.8329944815567819,
+      "grad_norm": 0.4991025924682617,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8331106593087424,
+      "grad_norm": 0.4552896022796631,
+      "learning_rate": 0.0001,
+      "loss": 1.6522,
+      "step": 7171
+    },
+    {
+      "epoch": 0.8332268370607029,
+      "grad_norm": 0.4508793354034424,
+      "learning_rate": 0.0001,
+      "loss": 1.7194,
+      "step": 7172
+    },
+    {
+      "epoch": 0.8333430148126634,
+      "grad_norm": 0.4660915732383728,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 7173
+    },
+    {
+      "epoch": 0.8334591925646239,
+      "grad_norm": 0.48091113567352295,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 7174
+    },
+    {
+      "epoch": 0.8335753703165844,
+      "grad_norm": 0.4672652781009674,
+      "learning_rate": 0.0001,
+      "loss": 1.6219,
+      "step": 7175
+    },
+    {
+      "epoch": 0.8336915480685448,
+      "grad_norm": 0.46777546405792236,
+      "learning_rate": 0.0001,
+      "loss": 1.6119,
+      "step": 7176
+    },
+    {
+      "epoch": 0.8338077258205053,
+      "grad_norm": 0.422648161649704,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 7177
+    },
+    {
+      "epoch": 0.8339239035724658,
+      "grad_norm": 0.4612855613231659,
+      "learning_rate": 0.0001,
+      "loss": 1.695,
+      "step": 7178
+    },
+    {
+      "epoch": 0.8340400813244264,
+      "grad_norm": 0.4481313228607178,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 7179
+    },
+    {
+      "epoch": 0.8341562590763869,
+      "grad_norm": 0.4890344440937042,
+      "learning_rate": 0.0001,
+      "loss": 1.6548,
+      "step": 7180
+    },
+    {
+      "epoch": 0.8342724368283474,
+      "grad_norm": 0.4647752046585083,
+      "learning_rate": 0.0001,
+      "loss": 1.5956,
+      "step": 7181
+    },
+    {
+      "epoch": 0.8343886145803079,
+      "grad_norm": 0.5206274390220642,
+      "learning_rate": 0.0001,
+      "loss": 1.7667,
+      "step": 7182
+    },
+    {
+      "epoch": 0.8345047923322684,
+      "grad_norm": 0.4441780745983124,
+      "learning_rate": 0.0001,
+      "loss": 1.3936,
+      "step": 7183
+    },
+    {
+      "epoch": 0.8346209700842289,
+      "grad_norm": 0.4659591019153595,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 7184
+    },
+    {
+      "epoch": 0.8347371478361894,
+      "grad_norm": 0.45973634719848633,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 7185
+    },
+    {
+      "epoch": 0.8348533255881498,
+      "grad_norm": 0.43164822459220886,
+      "learning_rate": 0.0001,
+      "loss": 1.4983,
+      "step": 7186
+    },
+    {
+      "epoch": 0.8349695033401103,
+      "grad_norm": 0.4463837742805481,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 7187
+    },
+    {
+      "epoch": 0.8350856810920708,
+      "grad_norm": 0.4984694719314575,
+      "learning_rate": 0.0001,
+      "loss": 1.8769,
+      "step": 7188
+    },
+    {
+      "epoch": 0.8352018588440314,
+      "grad_norm": 0.476712167263031,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 7189
+    },
+    {
+      "epoch": 0.8353180365959919,
+      "grad_norm": 0.4878515899181366,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 7190
+    },
+    {
+      "epoch": 0.8354342143479524,
+      "grad_norm": 0.44668495655059814,
+      "learning_rate": 0.0001,
+      "loss": 1.7159,
+      "step": 7191
+    },
+    {
+      "epoch": 0.8355503920999129,
+      "grad_norm": 0.46840810775756836,
+      "learning_rate": 0.0001,
+      "loss": 1.6875,
+      "step": 7192
+    },
+    {
+      "epoch": 0.8356665698518734,
+      "grad_norm": 0.47092488408088684,
+      "learning_rate": 0.0001,
+      "loss": 1.6715,
+      "step": 7193
+    },
+    {
+      "epoch": 0.8357827476038339,
+      "grad_norm": 0.4265216588973999,
+      "learning_rate": 0.0001,
+      "loss": 1.3333,
+      "step": 7194
+    },
+    {
+      "epoch": 0.8358989253557944,
+      "grad_norm": 0.4651343822479248,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 7195
+    },
+    {
+      "epoch": 0.8360151031077548,
+      "grad_norm": 0.4326002299785614,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 7196
+    },
+    {
+      "epoch": 0.8361312808597153,
+      "grad_norm": 0.46833327412605286,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 7197
+    },
+    {
+      "epoch": 0.8362474586116758,
+      "grad_norm": 0.46911799907684326,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 7198
+    },
+    {
+      "epoch": 0.8363636363636363,
+      "grad_norm": 0.43018439412117004,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 7199
+    },
+    {
+      "epoch": 0.8364798141155969,
+      "grad_norm": 0.5141427516937256,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 7200
+    },
+    {
+      "epoch": 0.8365959918675574,
+      "grad_norm": 0.4784258306026459,
+      "learning_rate": 0.0001,
+      "loss": 1.7193,
+      "step": 7201
+    },
+    {
+      "epoch": 0.8367121696195179,
+      "grad_norm": 0.47277989983558655,
+      "learning_rate": 0.0001,
+      "loss": 1.4349,
+      "step": 7202
+    },
+    {
+      "epoch": 0.8368283473714784,
+      "grad_norm": 0.46376681327819824,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 7203
+    },
+    {
+      "epoch": 0.8369445251234389,
+      "grad_norm": 0.4796522855758667,
+      "learning_rate": 0.0001,
+      "loss": 1.7946,
+      "step": 7204
+    },
+    {
+      "epoch": 0.8370607028753994,
+      "grad_norm": 0.42169952392578125,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 7205
+    },
+    {
+      "epoch": 0.8371768806273598,
+      "grad_norm": 0.4597862660884857,
+      "learning_rate": 0.0001,
+      "loss": 1.6694,
+      "step": 7206
+    },
+    {
+      "epoch": 0.8372930583793203,
+      "grad_norm": 0.4585164189338684,
+      "learning_rate": 0.0001,
+      "loss": 1.5235,
+      "step": 7207
+    },
+    {
+      "epoch": 0.8374092361312808,
+      "grad_norm": 0.4696848392486572,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 7208
+    },
+    {
+      "epoch": 0.8375254138832413,
+      "grad_norm": 0.4311649203300476,
+      "learning_rate": 0.0001,
+      "loss": 1.4602,
+      "step": 7209
+    },
+    {
+      "epoch": 0.8376415916352019,
+      "grad_norm": 0.48763203620910645,
+      "learning_rate": 0.0001,
+      "loss": 1.5663,
+      "step": 7210
+    },
+    {
+      "epoch": 0.8377577693871624,
+      "grad_norm": 0.4315451383590698,
+      "learning_rate": 0.0001,
+      "loss": 1.5024,
+      "step": 7211
+    },
+    {
+      "epoch": 0.8378739471391229,
+      "grad_norm": 0.44684144854545593,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 7212
+    },
+    {
+      "epoch": 0.8379901248910834,
+      "grad_norm": 0.4800936281681061,
+      "learning_rate": 0.0001,
+      "loss": 1.4873,
+      "step": 7213
+    },
+    {
+      "epoch": 0.8381063026430439,
+      "grad_norm": 0.45008590817451477,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 7214
+    },
+    {
+      "epoch": 0.8382224803950044,
+      "grad_norm": 0.5298113822937012,
+      "learning_rate": 0.0001,
+      "loss": 1.6431,
+      "step": 7215
+    },
+    {
+      "epoch": 0.8383386581469648,
+      "grad_norm": 0.4812602400779724,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 7216
+    },
+    {
+      "epoch": 0.8384548358989253,
+      "grad_norm": 0.48588478565216064,
+      "learning_rate": 0.0001,
+      "loss": 1.747,
+      "step": 7217
+    },
+    {
+      "epoch": 0.8385710136508858,
+      "grad_norm": 0.5386226177215576,
+      "learning_rate": 0.0001,
+      "loss": 1.6504,
+      "step": 7218
+    },
+    {
+      "epoch": 0.8386871914028463,
+      "grad_norm": 0.5037118196487427,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 7219
+    },
+    {
+      "epoch": 0.8388033691548068,
+      "grad_norm": 0.48345088958740234,
+      "learning_rate": 0.0001,
+      "loss": 1.7639,
+      "step": 7220
+    },
+    {
+      "epoch": 0.8389195469067674,
+      "grad_norm": 0.4569917321205139,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 7221
+    },
+    {
+      "epoch": 0.8390357246587279,
+      "grad_norm": 0.4810996651649475,
+      "learning_rate": 0.0001,
+      "loss": 1.6776,
+      "step": 7222
+    },
+    {
+      "epoch": 0.8391519024106884,
+      "grad_norm": 0.49533677101135254,
+      "learning_rate": 0.0001,
+      "loss": 1.6395,
+      "step": 7223
+    },
+    {
+      "epoch": 0.8392680801626489,
+      "grad_norm": 0.4612615704536438,
+      "learning_rate": 0.0001,
+      "loss": 1.7384,
+      "step": 7224
+    },
+    {
+      "epoch": 0.8393842579146094,
+      "grad_norm": 0.4266413152217865,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 7225
+    },
+    {
+      "epoch": 0.8395004356665698,
+      "grad_norm": 0.490537166595459,
+      "learning_rate": 0.0001,
+      "loss": 1.8198,
+      "step": 7226
+    },
+    {
+      "epoch": 0.8396166134185303,
+      "grad_norm": 0.45766183733940125,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 7227
+    },
+    {
+      "epoch": 0.8397327911704908,
+      "grad_norm": 0.48349621891975403,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 7228
+    },
+    {
+      "epoch": 0.8398489689224513,
+      "grad_norm": 0.44390740990638733,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 7229
+    },
+    {
+      "epoch": 0.8399651466744118,
+      "grad_norm": 0.4712617099285126,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 7230
+    },
+    {
+      "epoch": 0.8400813244263724,
+      "grad_norm": 0.47112375497817993,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 7231
+    },
+    {
+      "epoch": 0.8401975021783329,
+      "grad_norm": 0.491479754447937,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 7232
+    },
+    {
+      "epoch": 0.8403136799302934,
+      "grad_norm": 0.42350316047668457,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 7233
+    },
+    {
+      "epoch": 0.8404298576822539,
+      "grad_norm": 0.4231340289115906,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 7234
+    },
+    {
+      "epoch": 0.8405460354342144,
+      "grad_norm": 0.47028040885925293,
+      "learning_rate": 0.0001,
+      "loss": 1.7254,
+      "step": 7235
+    },
+    {
+      "epoch": 0.8406622131861748,
+      "grad_norm": 0.4882940351963043,
+      "learning_rate": 0.0001,
+      "loss": 1.7317,
+      "step": 7236
+    },
+    {
+      "epoch": 0.8407783909381353,
+      "grad_norm": 0.4849685728549957,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 7237
+    },
+    {
+      "epoch": 0.8408945686900958,
+      "grad_norm": 0.45504260063171387,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 7238
+    },
+    {
+      "epoch": 0.8410107464420563,
+      "grad_norm": 0.4550893008708954,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 7239
+    },
+    {
+      "epoch": 0.8411269241940168,
+      "grad_norm": 0.5196647047996521,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 7240
+    },
+    {
+      "epoch": 0.8412431019459774,
+      "grad_norm": 0.525698721408844,
+      "learning_rate": 0.0001,
+      "loss": 1.7223,
+      "step": 7241
+    },
+    {
+      "epoch": 0.8413592796979379,
+      "grad_norm": 0.46245935559272766,
+      "learning_rate": 0.0001,
+      "loss": 1.7588,
+      "step": 7242
+    },
+    {
+      "epoch": 0.8414754574498984,
+      "grad_norm": 0.4684200584888458,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 7243
+    },
+    {
+      "epoch": 0.8415916352018589,
+      "grad_norm": 0.5075876116752625,
+      "learning_rate": 0.0001,
+      "loss": 1.5209,
+      "step": 7244
+    },
+    {
+      "epoch": 0.8417078129538194,
+      "grad_norm": 0.4788230061531067,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 7245
+    },
+    {
+      "epoch": 0.8418239907057798,
+      "grad_norm": 0.4579117000102997,
+      "learning_rate": 0.0001,
+      "loss": 1.6237,
+      "step": 7246
+    },
+    {
+      "epoch": 0.8419401684577403,
+      "grad_norm": 0.46064794063568115,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 7247
+    },
+    {
+      "epoch": 0.8420563462097008,
+      "grad_norm": 0.49048542976379395,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 7248
+    },
+    {
+      "epoch": 0.8421725239616613,
+      "grad_norm": 0.46786949038505554,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 7249
+    },
+    {
+      "epoch": 0.8422887017136218,
+      "grad_norm": 0.4660671651363373,
+      "learning_rate": 0.0001,
+      "loss": 1.5582,
+      "step": 7250
+    },
+    {
+      "epoch": 0.8424048794655823,
+      "grad_norm": 0.45976850390434265,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 7251
+    },
+    {
+      "epoch": 0.8425210572175429,
+      "grad_norm": 0.45167800784111023,
+      "learning_rate": 0.0001,
+      "loss": 1.4037,
+      "step": 7252
+    },
+    {
+      "epoch": 0.8426372349695034,
+      "grad_norm": 0.45298877358436584,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 7253
+    },
+    {
+      "epoch": 0.8427534127214639,
+      "grad_norm": 0.4448161721229553,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 7254
+    },
+    {
+      "epoch": 0.8428695904734244,
+      "grad_norm": 0.4482094943523407,
+      "learning_rate": 0.0001,
+      "loss": 1.5979,
+      "step": 7255
+    },
+    {
+      "epoch": 0.8429857682253848,
+      "grad_norm": 0.45849940180778503,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 7256
+    },
+    {
+      "epoch": 0.8431019459773453,
+      "grad_norm": 0.4591526985168457,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 7257
+    },
+    {
+      "epoch": 0.8432181237293058,
+      "grad_norm": 0.4684211015701294,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 7258
+    },
+    {
+      "epoch": 0.8433343014812663,
+      "grad_norm": 0.4497279226779938,
+      "learning_rate": 0.0001,
+      "loss": 1.4289,
+      "step": 7259
+    },
+    {
+      "epoch": 0.8434504792332268,
+      "grad_norm": 0.46544843912124634,
+      "learning_rate": 0.0001,
+      "loss": 1.5504,
+      "step": 7260
+    },
+    {
+      "epoch": 0.8435666569851873,
+      "grad_norm": 0.4916938841342926,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 7261
+    },
+    {
+      "epoch": 0.8436828347371479,
+      "grad_norm": 0.4782385230064392,
+      "learning_rate": 0.0001,
+      "loss": 1.6031,
+      "step": 7262
+    },
+    {
+      "epoch": 0.8437990124891084,
+      "grad_norm": 0.46342313289642334,
+      "learning_rate": 0.0001,
+      "loss": 1.6774,
+      "step": 7263
+    },
+    {
+      "epoch": 0.8439151902410689,
+      "grad_norm": 0.456551194190979,
+      "learning_rate": 0.0001,
+      "loss": 1.7548,
+      "step": 7264
+    },
+    {
+      "epoch": 0.8440313679930294,
+      "grad_norm": 0.4580911695957184,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 7265
+    },
+    {
+      "epoch": 0.8441475457449898,
+      "grad_norm": 0.47985199093818665,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 7266
+    },
+    {
+      "epoch": 0.8442637234969503,
+      "grad_norm": 0.47257447242736816,
+      "learning_rate": 0.0001,
+      "loss": 1.5999,
+      "step": 7267
+    },
+    {
+      "epoch": 0.8443799012489108,
+      "grad_norm": 0.4965929687023163,
+      "learning_rate": 0.0001,
+      "loss": 1.5904,
+      "step": 7268
+    },
+    {
+      "epoch": 0.8444960790008713,
+      "grad_norm": 0.4734271168708801,
+      "learning_rate": 0.0001,
+      "loss": 1.7729,
+      "step": 7269
+    },
+    {
+      "epoch": 0.8446122567528318,
+      "grad_norm": 0.4869614839553833,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 7270
+    },
+    {
+      "epoch": 0.8447284345047923,
+      "grad_norm": 0.4773048162460327,
+      "learning_rate": 0.0001,
+      "loss": 1.5696,
+      "step": 7271
+    },
+    {
+      "epoch": 0.8448446122567528,
+      "grad_norm": 0.47592759132385254,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 7272
+    },
+    {
+      "epoch": 0.8449607900087134,
+      "grad_norm": 0.4415343105792999,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 7273
+    },
+    {
+      "epoch": 0.8450769677606739,
+      "grad_norm": 0.4446001350879669,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 7274
+    },
+    {
+      "epoch": 0.8451931455126344,
+      "grad_norm": 0.4620625674724579,
+      "learning_rate": 0.0001,
+      "loss": 1.5818,
+      "step": 7275
+    },
+    {
+      "epoch": 0.8453093232645948,
+      "grad_norm": 0.4809649884700775,
+      "learning_rate": 0.0001,
+      "loss": 1.5976,
+      "step": 7276
+    },
+    {
+      "epoch": 0.8454255010165553,
+      "grad_norm": 0.45134058594703674,
+      "learning_rate": 0.0001,
+      "loss": 1.6033,
+      "step": 7277
+    },
+    {
+      "epoch": 0.8455416787685158,
+      "grad_norm": 0.44494450092315674,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 7278
+    },
+    {
+      "epoch": 0.8456578565204763,
+      "grad_norm": 0.5055558085441589,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 7279
+    },
+    {
+      "epoch": 0.8457740342724368,
+      "grad_norm": 0.418148398399353,
+      "learning_rate": 0.0001,
+      "loss": 1.4975,
+      "step": 7280
+    },
+    {
+      "epoch": 0.8458902120243973,
+      "grad_norm": 0.4533846378326416,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 7281
+    },
+    {
+      "epoch": 0.8460063897763578,
+      "grad_norm": 0.49870097637176514,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 7282
+    },
+    {
+      "epoch": 0.8461225675283184,
+      "grad_norm": 0.4689449071884155,
+      "learning_rate": 0.0001,
+      "loss": 1.7109,
+      "step": 7283
+    },
+    {
+      "epoch": 0.8462387452802789,
+      "grad_norm": 0.466709166765213,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 7284
+    },
+    {
+      "epoch": 0.8463549230322394,
+      "grad_norm": 0.46668291091918945,
+      "learning_rate": 0.0001,
+      "loss": 1.7183,
+      "step": 7285
+    },
+    {
+      "epoch": 0.8464711007841998,
+      "grad_norm": 0.43597686290740967,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 7286
+    },
+    {
+      "epoch": 0.8465872785361603,
+      "grad_norm": 0.5049334168434143,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 7287
+    },
+    {
+      "epoch": 0.8467034562881208,
+      "grad_norm": 0.46801841259002686,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 7288
+    },
+    {
+      "epoch": 0.8468196340400813,
+      "grad_norm": 0.46203526854515076,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 7289
+    },
+    {
+      "epoch": 0.8469358117920418,
+      "grad_norm": 0.45195382833480835,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 7290
+    },
+    {
+      "epoch": 0.8470519895440023,
+      "grad_norm": 0.4815501868724823,
+      "learning_rate": 0.0001,
+      "loss": 1.5988,
+      "step": 7291
+    },
+    {
+      "epoch": 0.8471681672959628,
+      "grad_norm": 0.4791460931301117,
+      "learning_rate": 0.0001,
+      "loss": 1.6911,
+      "step": 7292
+    },
+    {
+      "epoch": 0.8472843450479233,
+      "grad_norm": 0.4468856155872345,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 7293
+    },
+    {
+      "epoch": 0.8474005227998839,
+      "grad_norm": 0.45526090264320374,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 7294
+    },
+    {
+      "epoch": 0.8475167005518444,
+      "grad_norm": 0.5207736492156982,
+      "learning_rate": 0.0001,
+      "loss": 1.7663,
+      "step": 7295
+    },
+    {
+      "epoch": 0.8476328783038048,
+      "grad_norm": 0.47375261783599854,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 7296
+    },
+    {
+      "epoch": 0.8477490560557653,
+      "grad_norm": 0.4951147139072418,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 7297
+    },
+    {
+      "epoch": 0.8478652338077258,
+      "grad_norm": 0.4843972325325012,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 7298
+    },
+    {
+      "epoch": 0.8479814115596863,
+      "grad_norm": 0.44949018955230713,
+      "learning_rate": 0.0001,
+      "loss": 1.5033,
+      "step": 7299
+    },
+    {
+      "epoch": 0.8480975893116468,
+      "grad_norm": 0.467597633600235,
+      "learning_rate": 0.0001,
+      "loss": 1.6343,
+      "step": 7300
+    },
+    {
+      "epoch": 0.8482137670636073,
+      "grad_norm": 0.44474077224731445,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 7301
+    },
+    {
+      "epoch": 0.8483299448155678,
+      "grad_norm": 0.455203115940094,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 7302
+    },
+    {
+      "epoch": 0.8484461225675283,
+      "grad_norm": 0.44709402322769165,
+      "learning_rate": 0.0001,
+      "loss": 1.3972,
+      "step": 7303
+    },
+    {
+      "epoch": 0.8485623003194889,
+      "grad_norm": 0.4874759316444397,
+      "learning_rate": 0.0001,
+      "loss": 1.7365,
+      "step": 7304
+    },
+    {
+      "epoch": 0.8486784780714494,
+      "grad_norm": 0.4757806956768036,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 7305
+    },
+    {
+      "epoch": 0.8487946558234098,
+      "grad_norm": 0.4443889260292053,
+      "learning_rate": 0.0001,
+      "loss": 1.4886,
+      "step": 7306
+    },
+    {
+      "epoch": 0.8489108335753703,
+      "grad_norm": 0.4977072477340698,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 7307
+    },
+    {
+      "epoch": 0.8490270113273308,
+      "grad_norm": 0.5108775496482849,
+      "learning_rate": 0.0001,
+      "loss": 1.7174,
+      "step": 7308
+    },
+    {
+      "epoch": 0.8491431890792913,
+      "grad_norm": 0.5324129462242126,
+      "learning_rate": 0.0001,
+      "loss": 1.7534,
+      "step": 7309
+    },
+    {
+      "epoch": 0.8492593668312518,
+      "grad_norm": 0.46875059604644775,
+      "learning_rate": 0.0001,
+      "loss": 1.7199,
+      "step": 7310
+    },
+    {
+      "epoch": 0.8493755445832123,
+      "grad_norm": 0.4918031096458435,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 7311
+    },
+    {
+      "epoch": 0.8494917223351728,
+      "grad_norm": 0.4679587185382843,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 7312
+    },
+    {
+      "epoch": 0.8496079000871333,
+      "grad_norm": 0.4425588846206665,
+      "learning_rate": 0.0001,
+      "loss": 1.4041,
+      "step": 7313
+    },
+    {
+      "epoch": 0.8497240778390938,
+      "grad_norm": 0.48238202929496765,
+      "learning_rate": 0.0001,
+      "loss": 1.8244,
+      "step": 7314
+    },
+    {
+      "epoch": 0.8498402555910544,
+      "grad_norm": 0.42519932985305786,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 7315
+    },
+    {
+      "epoch": 0.8499564333430148,
+      "grad_norm": 0.4286966919898987,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 7316
+    },
+    {
+      "epoch": 0.8500726110949753,
+      "grad_norm": 0.4630890190601349,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 7317
+    },
+    {
+      "epoch": 0.8501887888469358,
+      "grad_norm": 0.4862578809261322,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 7318
+    },
+    {
+      "epoch": 0.8503049665988963,
+      "grad_norm": 0.45061859488487244,
+      "learning_rate": 0.0001,
+      "loss": 1.6935,
+      "step": 7319
+    },
+    {
+      "epoch": 0.8504211443508568,
+      "grad_norm": 0.47194933891296387,
+      "learning_rate": 0.0001,
+      "loss": 1.639,
+      "step": 7320
+    },
+    {
+      "epoch": 0.8505373221028173,
+      "grad_norm": 0.457969605922699,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 7321
+    },
+    {
+      "epoch": 0.8506534998547778,
+      "grad_norm": 0.4440132677555084,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 7322
+    },
+    {
+      "epoch": 0.8507696776067383,
+      "grad_norm": 0.5148141384124756,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 7323
+    },
+    {
+      "epoch": 0.8508858553586988,
+      "grad_norm": 0.481653094291687,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 7324
+    },
+    {
+      "epoch": 0.8510020331106594,
+      "grad_norm": 0.4533710777759552,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 7325
+    },
+    {
+      "epoch": 0.8511182108626199,
+      "grad_norm": 0.4878830909729004,
+      "learning_rate": 0.0001,
+      "loss": 1.6365,
+      "step": 7326
+    },
+    {
+      "epoch": 0.8512343886145803,
+      "grad_norm": 0.5046993494033813,
+      "learning_rate": 0.0001,
+      "loss": 1.8043,
+      "step": 7327
+    },
+    {
+      "epoch": 0.8513505663665408,
+      "grad_norm": 0.48228132724761963,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 7328
+    },
+    {
+      "epoch": 0.8514667441185013,
+      "grad_norm": 0.47209233045578003,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 7329
+    },
+    {
+      "epoch": 0.8515829218704618,
+      "grad_norm": 0.4597385823726654,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 7330
+    },
+    {
+      "epoch": 0.8516990996224223,
+      "grad_norm": 0.478766530752182,
+      "learning_rate": 0.0001,
+      "loss": 1.6755,
+      "step": 7331
+    },
+    {
+      "epoch": 0.8518152773743828,
+      "grad_norm": 0.4749451279640198,
+      "learning_rate": 0.0001,
+      "loss": 1.5885,
+      "step": 7332
+    },
+    {
+      "epoch": 0.8519314551263433,
+      "grad_norm": 0.4681585431098938,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 7333
+    },
+    {
+      "epoch": 0.8520476328783038,
+      "grad_norm": 0.5120987892150879,
+      "learning_rate": 0.0001,
+      "loss": 1.651,
+      "step": 7334
+    },
+    {
+      "epoch": 0.8521638106302643,
+      "grad_norm": 0.4379187822341919,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 7335
+    },
+    {
+      "epoch": 0.8522799883822249,
+      "grad_norm": 0.5244159698486328,
+      "learning_rate": 0.0001,
+      "loss": 1.7338,
+      "step": 7336
+    },
+    {
+      "epoch": 0.8523961661341853,
+      "grad_norm": 0.4366852343082428,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 7337
+    },
+    {
+      "epoch": 0.8525123438861458,
+      "grad_norm": 0.468220591545105,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 7338
+    },
+    {
+      "epoch": 0.8526285216381063,
+      "grad_norm": 0.43382033705711365,
+      "learning_rate": 0.0001,
+      "loss": 1.4661,
+      "step": 7339
+    },
+    {
+      "epoch": 0.8527446993900668,
+      "grad_norm": 0.48206859827041626,
+      "learning_rate": 0.0001,
+      "loss": 1.7871,
+      "step": 7340
+    },
+    {
+      "epoch": 0.8528608771420273,
+      "grad_norm": 0.448639839887619,
+      "learning_rate": 0.0001,
+      "loss": 1.3691,
+      "step": 7341
+    },
+    {
+      "epoch": 0.8529770548939878,
+      "grad_norm": 0.48015591502189636,
+      "learning_rate": 0.0001,
+      "loss": 1.7805,
+      "step": 7342
+    },
+    {
+      "epoch": 0.8530932326459483,
+      "grad_norm": 0.44810763001441956,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 7343
+    },
+    {
+      "epoch": 0.8532094103979088,
+      "grad_norm": 0.5089064240455627,
+      "learning_rate": 0.0001,
+      "loss": 1.7226,
+      "step": 7344
+    },
+    {
+      "epoch": 0.8533255881498693,
+      "grad_norm": 0.4648088216781616,
+      "learning_rate": 0.0001,
+      "loss": 1.3528,
+      "step": 7345
+    },
+    {
+      "epoch": 0.8534417659018299,
+      "grad_norm": 0.46756070852279663,
+      "learning_rate": 0.0001,
+      "loss": 1.6056,
+      "step": 7346
+    },
+    {
+      "epoch": 0.8535579436537903,
+      "grad_norm": 0.4743233621120453,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 7347
+    },
+    {
+      "epoch": 0.8536741214057508,
+      "grad_norm": 0.4871358573436737,
+      "learning_rate": 0.0001,
+      "loss": 1.7681,
+      "step": 7348
+    },
+    {
+      "epoch": 0.8537902991577113,
+      "grad_norm": 0.4481533467769623,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 7349
+    },
+    {
+      "epoch": 0.8539064769096718,
+      "grad_norm": 0.49137020111083984,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 7350
+    },
+    {
+      "epoch": 0.8540226546616323,
+      "grad_norm": 0.45045387744903564,
+      "learning_rate": 0.0001,
+      "loss": 1.6635,
+      "step": 7351
+    },
+    {
+      "epoch": 0.8541388324135928,
+      "grad_norm": 0.45191293954849243,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 7352
+    },
+    {
+      "epoch": 0.8542550101655533,
+      "grad_norm": 0.46247225999832153,
+      "learning_rate": 0.0001,
+      "loss": 1.7451,
+      "step": 7353
+    },
+    {
+      "epoch": 0.8543711879175138,
+      "grad_norm": 0.4777053892612457,
+      "learning_rate": 0.0001,
+      "loss": 1.6937,
+      "step": 7354
+    },
+    {
+      "epoch": 0.8544873656694743,
+      "grad_norm": 0.44908279180526733,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 7355
+    },
+    {
+      "epoch": 0.8546035434214347,
+      "grad_norm": 0.4714983403682709,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 7356
+    },
+    {
+      "epoch": 0.8547197211733953,
+      "grad_norm": 0.4688902497291565,
+      "learning_rate": 0.0001,
+      "loss": 1.6913,
+      "step": 7357
+    },
+    {
+      "epoch": 0.8548358989253558,
+      "grad_norm": 0.46939048171043396,
+      "learning_rate": 0.0001,
+      "loss": 1.67,
+      "step": 7358
+    },
+    {
+      "epoch": 0.8549520766773163,
+      "grad_norm": 0.47825509309768677,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 7359
+    },
+    {
+      "epoch": 0.8550682544292768,
+      "grad_norm": 0.49258682131767273,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 7360
+    },
+    {
+      "epoch": 0.8551844321812373,
+      "grad_norm": 0.5022354125976562,
+      "learning_rate": 0.0001,
+      "loss": 1.7366,
+      "step": 7361
+    },
+    {
+      "epoch": 0.8553006099331978,
+      "grad_norm": 0.4815436899662018,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 7362
+    },
+    {
+      "epoch": 0.8554167876851583,
+      "grad_norm": 0.4701344668865204,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 7363
+    },
+    {
+      "epoch": 0.8555329654371188,
+      "grad_norm": 0.4595559537410736,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 7364
+    },
+    {
+      "epoch": 0.8556491431890793,
+      "grad_norm": 0.45530787110328674,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 7365
+    },
+    {
+      "epoch": 0.8557653209410397,
+      "grad_norm": 0.45153534412384033,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 7366
+    },
+    {
+      "epoch": 0.8558814986930003,
+      "grad_norm": 0.4568248391151428,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 7367
+    },
+    {
+      "epoch": 0.8559976764449608,
+      "grad_norm": 0.5031229853630066,
+      "learning_rate": 0.0001,
+      "loss": 1.6887,
+      "step": 7368
+    },
+    {
+      "epoch": 0.8561138541969213,
+      "grad_norm": 0.5217198729515076,
+      "learning_rate": 0.0001,
+      "loss": 1.7427,
+      "step": 7369
+    },
+    {
+      "epoch": 0.8562300319488818,
+      "grad_norm": 0.46275079250335693,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 7370
+    },
+    {
+      "epoch": 0.8563462097008423,
+      "grad_norm": 0.47801512479782104,
+      "learning_rate": 0.0001,
+      "loss": 1.6526,
+      "step": 7371
+    },
+    {
+      "epoch": 0.8564623874528028,
+      "grad_norm": 0.48373323678970337,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 7372
+    },
+    {
+      "epoch": 0.8565785652047633,
+      "grad_norm": 0.4380189776420593,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 7373
+    },
+    {
+      "epoch": 0.8566947429567238,
+      "grad_norm": 0.48953357338905334,
+      "learning_rate": 0.0001,
+      "loss": 1.7664,
+      "step": 7374
+    },
+    {
+      "epoch": 0.8568109207086843,
+      "grad_norm": 0.4443288743495941,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 7375
+    },
+    {
+      "epoch": 0.8569270984606447,
+      "grad_norm": 0.4462949335575104,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 7376
+    },
+    {
+      "epoch": 0.8570432762126052,
+      "grad_norm": 0.5047136545181274,
+      "learning_rate": 0.0001,
+      "loss": 1.6205,
+      "step": 7377
+    },
+    {
+      "epoch": 0.8571594539645658,
+      "grad_norm": 0.440891832113266,
+      "learning_rate": 0.0001,
+      "loss": 1.6558,
+      "step": 7378
+    },
+    {
+      "epoch": 0.8572756317165263,
+      "grad_norm": 0.49732422828674316,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 7379
+    },
+    {
+      "epoch": 0.8573918094684868,
+      "grad_norm": 0.47000056505203247,
+      "learning_rate": 0.0001,
+      "loss": 1.6352,
+      "step": 7380
+    },
+    {
+      "epoch": 0.8575079872204473,
+      "grad_norm": 0.4483802020549774,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 7381
+    },
+    {
+      "epoch": 0.8576241649724078,
+      "grad_norm": 0.4606070816516876,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 7382
+    },
+    {
+      "epoch": 0.8577403427243683,
+      "grad_norm": 0.4387550950050354,
+      "learning_rate": 0.0001,
+      "loss": 1.5059,
+      "step": 7383
+    },
+    {
+      "epoch": 0.8578565204763288,
+      "grad_norm": 0.4888121783733368,
+      "learning_rate": 0.0001,
+      "loss": 1.6821,
+      "step": 7384
+    },
+    {
+      "epoch": 0.8579726982282893,
+      "grad_norm": 0.4582982659339905,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 7385
+    },
+    {
+      "epoch": 0.8580888759802497,
+      "grad_norm": 0.46754390001296997,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 7386
+    },
+    {
+      "epoch": 0.8582050537322102,
+      "grad_norm": 0.44355496764183044,
+      "learning_rate": 0.0001,
+      "loss": 1.6246,
+      "step": 7387
+    },
+    {
+      "epoch": 0.8583212314841708,
+      "grad_norm": 0.628402829170227,
+      "learning_rate": 0.0001,
+      "loss": 1.4312,
+      "step": 7388
+    },
+    {
+      "epoch": 0.8584374092361313,
+      "grad_norm": 0.5051394701004028,
+      "learning_rate": 0.0001,
+      "loss": 1.6685,
+      "step": 7389
+    },
+    {
+      "epoch": 0.8585535869880918,
+      "grad_norm": 0.4363487660884857,
+      "learning_rate": 0.0001,
+      "loss": 1.3303,
+      "step": 7390
+    },
+    {
+      "epoch": 0.8586697647400523,
+      "grad_norm": 0.4447265863418579,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 7391
+    },
+    {
+      "epoch": 0.8587859424920128,
+      "grad_norm": 0.4490847885608673,
+      "learning_rate": 0.0001,
+      "loss": 1.3502,
+      "step": 7392
+    },
+    {
+      "epoch": 0.8589021202439733,
+      "grad_norm": 0.6224667429924011,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 7393
+    },
+    {
+      "epoch": 0.8590182979959338,
+      "grad_norm": 0.4750422537326813,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 7394
+    },
+    {
+      "epoch": 0.8591344757478943,
+      "grad_norm": 0.47161757946014404,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 7395
+    },
+    {
+      "epoch": 0.8592506534998547,
+      "grad_norm": 0.4862884283065796,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 7396
+    },
+    {
+      "epoch": 0.8593668312518152,
+      "grad_norm": 0.6144393682479858,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 7397
+    },
+    {
+      "epoch": 0.8594830090037757,
+      "grad_norm": 0.4535844326019287,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 7398
+    },
+    {
+      "epoch": 0.8595991867557363,
+      "grad_norm": 0.4707335829734802,
+      "learning_rate": 0.0001,
+      "loss": 1.6917,
+      "step": 7399
+    },
+    {
+      "epoch": 0.8597153645076968,
+      "grad_norm": 0.45508313179016113,
+      "learning_rate": 0.0001,
+      "loss": 1.5563,
+      "step": 7400
+    },
+    {
+      "epoch": 0.8598315422596573,
+      "grad_norm": 0.45613574981689453,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 7401
+    },
+    {
+      "epoch": 0.8599477200116178,
+      "grad_norm": 0.46452754735946655,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 7402
+    },
+    {
+      "epoch": 0.8600638977635783,
+      "grad_norm": 0.5027807354927063,
+      "learning_rate": 0.0001,
+      "loss": 1.7318,
+      "step": 7403
+    },
+    {
+      "epoch": 0.8601800755155388,
+      "grad_norm": 0.4823223352432251,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 7404
+    },
+    {
+      "epoch": 0.8602962532674993,
+      "grad_norm": 0.4800436794757843,
+      "learning_rate": 0.0001,
+      "loss": 1.6692,
+      "step": 7405
+    },
+    {
+      "epoch": 0.8604124310194597,
+      "grad_norm": 0.48085838556289673,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 7406
+    },
+    {
+      "epoch": 0.8605286087714202,
+      "grad_norm": 0.45660898089408875,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 7407
+    },
+    {
+      "epoch": 0.8606447865233807,
+      "grad_norm": 0.4891624450683594,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 7408
+    },
+    {
+      "epoch": 0.8607609642753413,
+      "grad_norm": 0.4807065427303314,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 7409
+    },
+    {
+      "epoch": 0.8608771420273018,
+      "grad_norm": 0.45160871744155884,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 7410
+    },
+    {
+      "epoch": 0.8609933197792623,
+      "grad_norm": 0.4672984182834625,
+      "learning_rate": 0.0001,
+      "loss": 1.5387,
+      "step": 7411
+    },
+    {
+      "epoch": 0.8611094975312228,
+      "grad_norm": 0.4706302881240845,
+      "learning_rate": 0.0001,
+      "loss": 1.7187,
+      "step": 7412
+    },
+    {
+      "epoch": 0.8612256752831833,
+      "grad_norm": 0.4920727014541626,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 7413
+    },
+    {
+      "epoch": 0.8613418530351438,
+      "grad_norm": 0.46764302253723145,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 7414
+    },
+    {
+      "epoch": 0.8614580307871043,
+      "grad_norm": 0.4732871651649475,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 7415
+    },
+    {
+      "epoch": 0.8615742085390647,
+      "grad_norm": 0.4615955054759979,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 7416
+    },
+    {
+      "epoch": 0.8616903862910252,
+      "grad_norm": 1.543933629989624,
+      "learning_rate": 0.0001,
+      "loss": 1.7691,
+      "step": 7417
+    },
+    {
+      "epoch": 0.8618065640429857,
+      "grad_norm": 0.4970760941505432,
+      "learning_rate": 0.0001,
+      "loss": 1.5598,
+      "step": 7418
+    },
+    {
+      "epoch": 0.8619227417949463,
+      "grad_norm": 0.45834749937057495,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 7419
+    },
+    {
+      "epoch": 0.8620389195469068,
+      "grad_norm": 0.49793627858161926,
+      "learning_rate": 0.0001,
+      "loss": 1.7069,
+      "step": 7420
+    },
+    {
+      "epoch": 0.8621550972988673,
+      "grad_norm": 0.4896315336227417,
+      "learning_rate": 0.0001,
+      "loss": 1.5341,
+      "step": 7421
+    },
+    {
+      "epoch": 0.8622712750508278,
+      "grad_norm": 0.46836981177330017,
+      "learning_rate": 0.0001,
+      "loss": 1.5196,
+      "step": 7422
+    },
+    {
+      "epoch": 0.8623874528027883,
+      "grad_norm": 0.47271421551704407,
+      "learning_rate": 0.0001,
+      "loss": 1.6342,
+      "step": 7423
+    },
+    {
+      "epoch": 0.8625036305547488,
+      "grad_norm": 0.5026324987411499,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 7424
+    },
+    {
+      "epoch": 0.8626198083067093,
+      "grad_norm": 0.515398383140564,
+      "learning_rate": 0.0001,
+      "loss": 1.813,
+      "step": 7425
+    },
+    {
+      "epoch": 0.8627359860586697,
+      "grad_norm": 0.49251848459243774,
+      "learning_rate": 0.0001,
+      "loss": 1.6419,
+      "step": 7426
+    },
+    {
+      "epoch": 0.8628521638106302,
+      "grad_norm": 0.44753625988960266,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 7427
+    },
+    {
+      "epoch": 0.8629683415625907,
+      "grad_norm": 0.4866905212402344,
+      "learning_rate": 0.0001,
+      "loss": 1.7084,
+      "step": 7428
+    },
+    {
+      "epoch": 0.8630845193145512,
+      "grad_norm": 0.4585065245628357,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 7429
+    },
+    {
+      "epoch": 0.8632006970665118,
+      "grad_norm": 0.47094056010246277,
+      "learning_rate": 0.0001,
+      "loss": 1.7554,
+      "step": 7430
+    },
+    {
+      "epoch": 0.8633168748184723,
+      "grad_norm": 0.45176732540130615,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 7431
+    },
+    {
+      "epoch": 0.8634330525704328,
+      "grad_norm": 0.48504602909088135,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 7432
+    },
+    {
+      "epoch": 0.8635492303223933,
+      "grad_norm": 0.49550312757492065,
+      "learning_rate": 0.0001,
+      "loss": 1.6723,
+      "step": 7433
+    },
+    {
+      "epoch": 0.8636654080743538,
+      "grad_norm": 0.4960317313671112,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 7434
+    },
+    {
+      "epoch": 0.8637815858263143,
+      "grad_norm": 0.4894813597202301,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 7435
+    },
+    {
+      "epoch": 0.8638977635782747,
+      "grad_norm": 0.47758471965789795,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 7436
+    },
+    {
+      "epoch": 0.8640139413302352,
+      "grad_norm": 0.5021971464157104,
+      "learning_rate": 0.0001,
+      "loss": 1.7103,
+      "step": 7437
+    },
+    {
+      "epoch": 0.8641301190821957,
+      "grad_norm": 0.5115821957588196,
+      "learning_rate": 0.0001,
+      "loss": 1.8215,
+      "step": 7438
+    },
+    {
+      "epoch": 0.8642462968341562,
+      "grad_norm": 0.45494914054870605,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 7439
+    },
+    {
+      "epoch": 0.8643624745861168,
+      "grad_norm": 0.4635239839553833,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 7440
+    },
+    {
+      "epoch": 0.8644786523380773,
+      "grad_norm": 0.4838850200176239,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 7441
+    },
+    {
+      "epoch": 0.8645948300900378,
+      "grad_norm": 0.48782113194465637,
+      "learning_rate": 0.0001,
+      "loss": 1.8607,
+      "step": 7442
+    },
+    {
+      "epoch": 0.8647110078419983,
+      "grad_norm": 0.4964597225189209,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 7443
+    },
+    {
+      "epoch": 0.8648271855939588,
+      "grad_norm": 0.4533204734325409,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 7444
+    },
+    {
+      "epoch": 0.8649433633459193,
+      "grad_norm": 0.4995560348033905,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 7445
+    },
+    {
+      "epoch": 0.8650595410978797,
+      "grad_norm": 0.48218008875846863,
+      "learning_rate": 0.0001,
+      "loss": 1.5235,
+      "step": 7446
+    },
+    {
+      "epoch": 0.8651757188498402,
+      "grad_norm": 0.4977954626083374,
+      "learning_rate": 0.0001,
+      "loss": 1.6671,
+      "step": 7447
+    },
+    {
+      "epoch": 0.8652918966018007,
+      "grad_norm": 0.4879086911678314,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 7448
+    },
+    {
+      "epoch": 0.8654080743537612,
+      "grad_norm": 0.4788166880607605,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 7449
+    },
+    {
+      "epoch": 0.8655242521057217,
+      "grad_norm": 0.5052685737609863,
+      "learning_rate": 0.0001,
+      "loss": 1.6799,
+      "step": 7450
+    },
+    {
+      "epoch": 0.8656404298576823,
+      "grad_norm": 0.48076626658439636,
+      "learning_rate": 0.0001,
+      "loss": 1.5764,
+      "step": 7451
+    },
+    {
+      "epoch": 0.8657566076096428,
+      "grad_norm": 0.481800377368927,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 7452
+    },
+    {
+      "epoch": 0.8658727853616033,
+      "grad_norm": 0.4728825092315674,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 7453
+    },
+    {
+      "epoch": 0.8659889631135638,
+      "grad_norm": 0.4901738464832306,
+      "learning_rate": 0.0001,
+      "loss": 1.6488,
+      "step": 7454
+    },
+    {
+      "epoch": 0.8661051408655243,
+      "grad_norm": 0.47152256965637207,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 7455
+    },
+    {
+      "epoch": 0.8662213186174847,
+      "grad_norm": 0.4627699851989746,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 7456
+    },
+    {
+      "epoch": 0.8663374963694452,
+      "grad_norm": 0.47442302107810974,
+      "learning_rate": 0.0001,
+      "loss": 1.6379,
+      "step": 7457
+    },
+    {
+      "epoch": 0.8664536741214057,
+      "grad_norm": 0.45434120297431946,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 7458
+    },
+    {
+      "epoch": 0.8665698518733662,
+      "grad_norm": 0.5147332549095154,
+      "learning_rate": 0.0001,
+      "loss": 1.6957,
+      "step": 7459
+    },
+    {
+      "epoch": 0.8666860296253267,
+      "grad_norm": 0.4700101315975189,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 7460
+    },
+    {
+      "epoch": 0.8668022073772873,
+      "grad_norm": 0.5192029476165771,
+      "learning_rate": 0.0001,
+      "loss": 1.562,
+      "step": 7461
+    },
+    {
+      "epoch": 0.8669183851292478,
+      "grad_norm": 0.44885125756263733,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 7462
+    },
+    {
+      "epoch": 0.8670345628812083,
+      "grad_norm": 0.45993509888648987,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 7463
+    },
+    {
+      "epoch": 0.8671507406331688,
+      "grad_norm": 0.46040210127830505,
+      "learning_rate": 0.0001,
+      "loss": 1.3868,
+      "step": 7464
+    },
+    {
+      "epoch": 0.8672669183851293,
+      "grad_norm": 0.47678741812705994,
+      "learning_rate": 0.0001,
+      "loss": 1.6829,
+      "step": 7465
+    },
+    {
+      "epoch": 0.8673830961370897,
+      "grad_norm": 0.4918549358844757,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 7466
+    },
+    {
+      "epoch": 0.8674992738890502,
+      "grad_norm": 0.45147234201431274,
+      "learning_rate": 0.0001,
+      "loss": 1.3795,
+      "step": 7467
+    },
+    {
+      "epoch": 0.8676154516410107,
+      "grad_norm": 0.4544316530227661,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 7468
+    },
+    {
+      "epoch": 0.8677316293929712,
+      "grad_norm": 0.4829142987728119,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 7469
+    },
+    {
+      "epoch": 0.8678478071449317,
+      "grad_norm": 0.5195518136024475,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 7470
+    },
+    {
+      "epoch": 0.8679639848968922,
+      "grad_norm": 0.48266440629959106,
+      "learning_rate": 0.0001,
+      "loss": 1.6076,
+      "step": 7471
+    },
+    {
+      "epoch": 0.8680801626488528,
+      "grad_norm": 0.5128846764564514,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 7472
+    },
+    {
+      "epoch": 0.8681963404008133,
+      "grad_norm": 0.48383504152297974,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 7473
+    },
+    {
+      "epoch": 0.8683125181527738,
+      "grad_norm": 0.49934422969818115,
+      "learning_rate": 0.0001,
+      "loss": 1.7374,
+      "step": 7474
+    },
+    {
+      "epoch": 0.8684286959047343,
+      "grad_norm": 0.4719254672527313,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 7475
+    },
+    {
+      "epoch": 0.8685448736566947,
+      "grad_norm": 0.45778128504753113,
+      "learning_rate": 0.0001,
+      "loss": 1.4926,
+      "step": 7476
+    },
+    {
+      "epoch": 0.8686610514086552,
+      "grad_norm": 0.4514407813549042,
+      "learning_rate": 0.0001,
+      "loss": 1.6567,
+      "step": 7477
+    },
+    {
+      "epoch": 0.8687772291606157,
+      "grad_norm": 0.4867977201938629,
+      "learning_rate": 0.0001,
+      "loss": 1.6724,
+      "step": 7478
+    },
+    {
+      "epoch": 0.8688934069125762,
+      "grad_norm": 0.4617425799369812,
+      "learning_rate": 0.0001,
+      "loss": 1.684,
+      "step": 7479
+    },
+    {
+      "epoch": 0.8690095846645367,
+      "grad_norm": 0.44265344738960266,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 7480
+    },
+    {
+      "epoch": 0.8691257624164972,
+      "grad_norm": 0.4775390625,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 7481
+    },
+    {
+      "epoch": 0.8692419401684578,
+      "grad_norm": 0.5168774127960205,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 7482
+    },
+    {
+      "epoch": 0.8693581179204183,
+      "grad_norm": 0.48702213168144226,
+      "learning_rate": 0.0001,
+      "loss": 1.7377,
+      "step": 7483
+    },
+    {
+      "epoch": 0.8694742956723788,
+      "grad_norm": 0.47748100757598877,
+      "learning_rate": 0.0001,
+      "loss": 1.7014,
+      "step": 7484
+    },
+    {
+      "epoch": 0.8695904734243393,
+      "grad_norm": 0.45373064279556274,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 7485
+    },
+    {
+      "epoch": 0.8697066511762997,
+      "grad_norm": 0.4954831898212433,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 7486
+    },
+    {
+      "epoch": 0.8698228289282602,
+      "grad_norm": 0.4588451683521271,
+      "learning_rate": 0.0001,
+      "loss": 1.3947,
+      "step": 7487
+    },
+    {
+      "epoch": 0.8699390066802207,
+      "grad_norm": 0.4915754199028015,
+      "learning_rate": 0.0001,
+      "loss": 1.6084,
+      "step": 7488
+    },
+    {
+      "epoch": 0.8700551844321812,
+      "grad_norm": 0.4902786612510681,
+      "learning_rate": 0.0001,
+      "loss": 1.4462,
+      "step": 7489
+    },
+    {
+      "epoch": 0.8701713621841417,
+      "grad_norm": 0.4758920967578888,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 7490
+    },
+    {
+      "epoch": 0.8702875399361022,
+      "grad_norm": 0.4647005796432495,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 7491
+    },
+    {
+      "epoch": 0.8704037176880627,
+      "grad_norm": 0.4310697615146637,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 7492
+    },
+    {
+      "epoch": 0.8705198954400233,
+      "grad_norm": 0.4658608138561249,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 7493
+    },
+    {
+      "epoch": 0.8706360731919838,
+      "grad_norm": 0.5647340416908264,
+      "learning_rate": 0.0001,
+      "loss": 1.7218,
+      "step": 7494
+    },
+    {
+      "epoch": 0.8707522509439443,
+      "grad_norm": 0.4370400607585907,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 7495
+    },
+    {
+      "epoch": 0.8708684286959048,
+      "grad_norm": 0.5290013551712036,
+      "learning_rate": 0.0001,
+      "loss": 1.7937,
+      "step": 7496
+    },
+    {
+      "epoch": 0.8709846064478652,
+      "grad_norm": 0.49788692593574524,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 7497
+    },
+    {
+      "epoch": 0.8711007841998257,
+      "grad_norm": 0.5271779894828796,
+      "learning_rate": 0.0001,
+      "loss": 1.4642,
+      "step": 7498
+    },
+    {
+      "epoch": 0.8712169619517862,
+      "grad_norm": 0.5094662308692932,
+      "learning_rate": 0.0001,
+      "loss": 1.766,
+      "step": 7499
+    },
+    {
+      "epoch": 0.8713331397037467,
+      "grad_norm": 0.45193493366241455,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 7500
+    },
+    {
+      "epoch": 0.8714493174557072,
+      "grad_norm": 0.4743613600730896,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 7501
+    },
+    {
+      "epoch": 0.8715654952076677,
+      "grad_norm": 0.5035197734832764,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 7502
+    },
+    {
+      "epoch": 0.8716816729596283,
+      "grad_norm": 0.4376752972602844,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 7503
+    },
+    {
+      "epoch": 0.8717978507115888,
+      "grad_norm": 0.5224100947380066,
+      "learning_rate": 0.0001,
+      "loss": 1.7625,
+      "step": 7504
+    },
+    {
+      "epoch": 0.8719140284635493,
+      "grad_norm": 0.44658803939819336,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 7505
+    },
+    {
+      "epoch": 0.8720302062155098,
+      "grad_norm": 0.44167301058769226,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 7506
+    },
+    {
+      "epoch": 0.8721463839674702,
+      "grad_norm": 0.45513877272605896,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 7507
+    },
+    {
+      "epoch": 0.8722625617194307,
+      "grad_norm": 0.5717625617980957,
+      "learning_rate": 0.0001,
+      "loss": 1.7618,
+      "step": 7508
+    },
+    {
+      "epoch": 0.8723787394713912,
+      "grad_norm": 0.4953731298446655,
+      "learning_rate": 0.0001,
+      "loss": 1.7237,
+      "step": 7509
+    },
+    {
+      "epoch": 0.8724949172233517,
+      "grad_norm": 0.49298036098480225,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 7510
+    },
+    {
+      "epoch": 0.8726110949753122,
+      "grad_norm": 0.4598557651042938,
+      "learning_rate": 0.0001,
+      "loss": 1.5216,
+      "step": 7511
+    },
+    {
+      "epoch": 0.8727272727272727,
+      "grad_norm": 0.46120190620422363,
+      "learning_rate": 0.0001,
+      "loss": 1.587,
+      "step": 7512
+    },
+    {
+      "epoch": 0.8728434504792332,
+      "grad_norm": 0.48204609751701355,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 7513
+    },
+    {
+      "epoch": 0.8729596282311938,
+      "grad_norm": 0.5170610547065735,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 7514
+    },
+    {
+      "epoch": 0.8730758059831543,
+      "grad_norm": 0.47179120779037476,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 7515
+    },
+    {
+      "epoch": 0.8731919837351148,
+      "grad_norm": 0.5068784952163696,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 7516
+    },
+    {
+      "epoch": 0.8733081614870752,
+      "grad_norm": 0.46290868520736694,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 7517
+    },
+    {
+      "epoch": 0.8734243392390357,
+      "grad_norm": 0.45205244421958923,
+      "learning_rate": 0.0001,
+      "loss": 1.3574,
+      "step": 7518
+    },
+    {
+      "epoch": 0.8735405169909962,
+      "grad_norm": 0.934231162071228,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 7519
+    },
+    {
+      "epoch": 0.8736566947429567,
+      "grad_norm": 0.45129621028900146,
+      "learning_rate": 0.0001,
+      "loss": 1.4467,
+      "step": 7520
+    },
+    {
+      "epoch": 0.8737728724949172,
+      "grad_norm": 0.456668496131897,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 7521
+    },
+    {
+      "epoch": 0.8738890502468777,
+      "grad_norm": 0.5230050086975098,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 7522
+    },
+    {
+      "epoch": 0.8740052279988382,
+      "grad_norm": 0.47268834710121155,
+      "learning_rate": 0.0001,
+      "loss": 1.6878,
+      "step": 7523
+    },
+    {
+      "epoch": 0.8741214057507988,
+      "grad_norm": 0.5117980241775513,
+      "learning_rate": 0.0001,
+      "loss": 1.5417,
+      "step": 7524
+    },
+    {
+      "epoch": 0.8742375835027593,
+      "grad_norm": 0.44424542784690857,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 7525
+    },
+    {
+      "epoch": 0.8743537612547198,
+      "grad_norm": 0.479316771030426,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 7526
+    },
+    {
+      "epoch": 0.8744699390066802,
+      "grad_norm": 0.442326158285141,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 7527
+    },
+    {
+      "epoch": 0.8745861167586407,
+      "grad_norm": 0.4649803936481476,
+      "learning_rate": 0.0001,
+      "loss": 1.5859,
+      "step": 7528
+    },
+    {
+      "epoch": 0.8747022945106012,
+      "grad_norm": 0.41753268241882324,
+      "learning_rate": 0.0001,
+      "loss": 1.4568,
+      "step": 7529
+    },
+    {
+      "epoch": 0.8748184722625617,
+      "grad_norm": 0.5000750422477722,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 7530
+    },
+    {
+      "epoch": 0.8749346500145222,
+      "grad_norm": 0.4585053026676178,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 7531
+    },
+    {
+      "epoch": 0.8750508277664827,
+      "grad_norm": 0.48473891615867615,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 7532
+    },
+    {
+      "epoch": 0.8751670055184432,
+      "grad_norm": 0.4806990325450897,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 7533
+    },
+    {
+      "epoch": 0.8752831832704037,
+      "grad_norm": 0.4711168110370636,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 7534
+    },
+    {
+      "epoch": 0.8753993610223643,
+      "grad_norm": 0.4625207483768463,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 7535
+    },
+    {
+      "epoch": 0.8755155387743248,
+      "grad_norm": 0.47735491394996643,
+      "learning_rate": 0.0001,
+      "loss": 1.5856,
+      "step": 7536
+    },
+    {
+      "epoch": 0.8756317165262852,
+      "grad_norm": 0.47820740938186646,
+      "learning_rate": 0.0001,
+      "loss": 1.5657,
+      "step": 7537
+    },
+    {
+      "epoch": 0.8757478942782457,
+      "grad_norm": 0.43879976868629456,
+      "learning_rate": 0.0001,
+      "loss": 1.4218,
+      "step": 7538
+    },
+    {
+      "epoch": 0.8758640720302062,
+      "grad_norm": 0.5144027471542358,
+      "learning_rate": 0.0001,
+      "loss": 1.6866,
+      "step": 7539
+    },
+    {
+      "epoch": 0.8759802497821667,
+      "grad_norm": 0.5422682166099548,
+      "learning_rate": 0.0001,
+      "loss": 1.765,
+      "step": 7540
+    },
+    {
+      "epoch": 0.8760964275341272,
+      "grad_norm": 0.49460479617118835,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 7541
+    },
+    {
+      "epoch": 0.8762126052860877,
+      "grad_norm": 0.4671952426433563,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 7542
+    },
+    {
+      "epoch": 0.8763287830380482,
+      "grad_norm": 0.4591098725795746,
+      "learning_rate": 0.0001,
+      "loss": 1.5281,
+      "step": 7543
+    },
+    {
+      "epoch": 0.8764449607900087,
+      "grad_norm": 0.49451717734336853,
+      "learning_rate": 0.0001,
+      "loss": 1.6752,
+      "step": 7544
+    },
+    {
+      "epoch": 0.8765611385419693,
+      "grad_norm": 0.48347243666648865,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 7545
+    },
+    {
+      "epoch": 0.8766773162939298,
+      "grad_norm": 0.4897051453590393,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 7546
+    },
+    {
+      "epoch": 0.8767934940458902,
+      "grad_norm": 0.45810526609420776,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 7547
+    },
+    {
+      "epoch": 0.8769096717978507,
+      "grad_norm": 0.4647473394870758,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 7548
+    },
+    {
+      "epoch": 0.8770258495498112,
+      "grad_norm": 0.4482647776603699,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 7549
+    },
+    {
+      "epoch": 0.8771420273017717,
+      "grad_norm": 0.44632047414779663,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 7550
+    },
+    {
+      "epoch": 0.8772582050537322,
+      "grad_norm": 0.5201831459999084,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 7551
+    },
+    {
+      "epoch": 0.8773743828056927,
+      "grad_norm": 0.46722298860549927,
+      "learning_rate": 0.0001,
+      "loss": 1.7115,
+      "step": 7552
+    },
+    {
+      "epoch": 0.8774905605576532,
+      "grad_norm": 0.5357077121734619,
+      "learning_rate": 0.0001,
+      "loss": 1.7269,
+      "step": 7553
+    },
+    {
+      "epoch": 0.8776067383096137,
+      "grad_norm": 0.481431782245636,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 7554
+    },
+    {
+      "epoch": 0.8777229160615742,
+      "grad_norm": 0.4767739474773407,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 7555
+    },
+    {
+      "epoch": 0.8778390938135348,
+      "grad_norm": 0.4820916950702667,
+      "learning_rate": 0.0001,
+      "loss": 1.3223,
+      "step": 7556
+    },
+    {
+      "epoch": 0.8779552715654952,
+      "grad_norm": 0.49660539627075195,
+      "learning_rate": 0.0001,
+      "loss": 1.5837,
+      "step": 7557
+    },
+    {
+      "epoch": 0.8780714493174557,
+      "grad_norm": 0.4909195303916931,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 7558
+    },
+    {
+      "epoch": 0.8781876270694162,
+      "grad_norm": 0.47243842482566833,
+      "learning_rate": 0.0001,
+      "loss": 1.5658,
+      "step": 7559
+    },
+    {
+      "epoch": 0.8783038048213767,
+      "grad_norm": 0.501683235168457,
+      "learning_rate": 0.0001,
+      "loss": 1.7332,
+      "step": 7560
+    },
+    {
+      "epoch": 0.8784199825733372,
+      "grad_norm": 0.4446217119693756,
+      "learning_rate": 0.0001,
+      "loss": 1.4357,
+      "step": 7561
+    },
+    {
+      "epoch": 0.8785361603252977,
+      "grad_norm": 0.5229566693305969,
+      "learning_rate": 0.0001,
+      "loss": 1.7222,
+      "step": 7562
+    },
+    {
+      "epoch": 0.8786523380772582,
+      "grad_norm": 0.479064404964447,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 7563
+    },
+    {
+      "epoch": 0.8787685158292187,
+      "grad_norm": 0.4563969075679779,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 7564
+    },
+    {
+      "epoch": 0.8788846935811792,
+      "grad_norm": 0.47300323843955994,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 7565
+    },
+    {
+      "epoch": 0.8790008713331398,
+      "grad_norm": 0.48597297072410583,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 7566
+    },
+    {
+      "epoch": 0.8791170490851002,
+      "grad_norm": 0.4700767695903778,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 7567
+    },
+    {
+      "epoch": 0.8792332268370607,
+      "grad_norm": 0.50390625,
+      "learning_rate": 0.0001,
+      "loss": 1.7064,
+      "step": 7568
+    },
+    {
+      "epoch": 0.8793494045890212,
+      "grad_norm": 0.4735111892223358,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 7569
+    },
+    {
+      "epoch": 0.8794655823409817,
+      "grad_norm": 0.43248268961906433,
+      "learning_rate": 0.0001,
+      "loss": 1.4049,
+      "step": 7570
+    },
+    {
+      "epoch": 0.8795817600929422,
+      "grad_norm": 0.4565393328666687,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 7571
+    },
+    {
+      "epoch": 0.8796979378449027,
+      "grad_norm": 0.46394291520118713,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 7572
+    },
+    {
+      "epoch": 0.8798141155968632,
+      "grad_norm": 0.47517409920692444,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 7573
+    },
+    {
+      "epoch": 0.8799302933488237,
+      "grad_norm": 0.515457272529602,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 7574
+    },
+    {
+      "epoch": 0.8800464711007842,
+      "grad_norm": 0.4533648192882538,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 7575
+    },
+    {
+      "epoch": 0.8801626488527446,
+      "grad_norm": 0.4539215862751007,
+      "learning_rate": 0.0001,
+      "loss": 1.504,
+      "step": 7576
+    },
+    {
+      "epoch": 0.8802788266047052,
+      "grad_norm": 0.5214062333106995,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 7577
+    },
+    {
+      "epoch": 0.8803950043566657,
+      "grad_norm": 0.4697798192501068,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 7578
+    },
+    {
+      "epoch": 0.8805111821086262,
+      "grad_norm": 0.4855616092681885,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 7579
+    },
+    {
+      "epoch": 0.8806273598605867,
+      "grad_norm": 0.49214091897010803,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 7580
+    },
+    {
+      "epoch": 0.8807435376125472,
+      "grad_norm": 0.5328443646430969,
+      "learning_rate": 0.0001,
+      "loss": 1.7461,
+      "step": 7581
+    },
+    {
+      "epoch": 0.8808597153645077,
+      "grad_norm": 0.5216013789176941,
+      "learning_rate": 0.0001,
+      "loss": 1.8298,
+      "step": 7582
+    },
+    {
+      "epoch": 0.8809758931164682,
+      "grad_norm": 0.46522417664527893,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 7583
+    },
+    {
+      "epoch": 0.8810920708684287,
+      "grad_norm": 0.4373832046985626,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 7584
+    },
+    {
+      "epoch": 0.8812082486203892,
+      "grad_norm": 0.49534451961517334,
+      "learning_rate": 0.0001,
+      "loss": 1.7228,
+      "step": 7585
+    },
+    {
+      "epoch": 0.8813244263723496,
+      "grad_norm": 0.42766979336738586,
+      "learning_rate": 0.0001,
+      "loss": 1.3932,
+      "step": 7586
+    },
+    {
+      "epoch": 0.8814406041243102,
+      "grad_norm": 0.47507309913635254,
+      "learning_rate": 0.0001,
+      "loss": 1.6604,
+      "step": 7587
+    },
+    {
+      "epoch": 0.8815567818762707,
+      "grad_norm": 0.4551963210105896,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 7588
+    },
+    {
+      "epoch": 0.8816729596282312,
+      "grad_norm": 0.46101507544517517,
+      "learning_rate": 0.0001,
+      "loss": 1.7364,
+      "step": 7589
+    },
+    {
+      "epoch": 0.8817891373801917,
+      "grad_norm": 0.45468607544898987,
+      "learning_rate": 0.0001,
+      "loss": 1.3932,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8819053151321522,
+      "grad_norm": 0.4612409174442291,
+      "learning_rate": 0.0001,
+      "loss": 1.592,
+      "step": 7591
+    },
+    {
+      "epoch": 0.8820214928841127,
+      "grad_norm": 0.4820188581943512,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 7592
+    },
+    {
+      "epoch": 0.8821376706360732,
+      "grad_norm": 0.49493300914764404,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 7593
+    },
+    {
+      "epoch": 0.8822538483880337,
+      "grad_norm": 0.45542287826538086,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 7594
+    },
+    {
+      "epoch": 0.8823700261399942,
+      "grad_norm": 0.47837620973587036,
+      "learning_rate": 0.0001,
+      "loss": 1.6641,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8824862038919546,
+      "grad_norm": 0.45609262585639954,
+      "learning_rate": 0.0001,
+      "loss": 1.4127,
+      "step": 7596
+    },
+    {
+      "epoch": 0.8826023816439151,
+      "grad_norm": 0.46420490741729736,
+      "learning_rate": 0.0001,
+      "loss": 1.56,
+      "step": 7597
+    },
+    {
+      "epoch": 0.8827185593958757,
+      "grad_norm": 0.48322582244873047,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 7598
+    },
+    {
+      "epoch": 0.8828347371478362,
+      "grad_norm": 0.4640806317329407,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 7599
+    },
+    {
+      "epoch": 0.8829509148997967,
+      "grad_norm": 0.47058844566345215,
+      "learning_rate": 0.0001,
+      "loss": 1.6128,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8830670926517572,
+      "grad_norm": 0.5269159078598022,
+      "learning_rate": 0.0001,
+      "loss": 1.7702,
+      "step": 7601
+    },
+    {
+      "epoch": 0.8831832704037177,
+      "grad_norm": 0.4986213445663452,
+      "learning_rate": 0.0001,
+      "loss": 1.6251,
+      "step": 7602
+    },
+    {
+      "epoch": 0.8832994481556782,
+      "grad_norm": 0.4375273585319519,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 7603
+    },
+    {
+      "epoch": 0.8834156259076387,
+      "grad_norm": 0.45944470167160034,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 7604
+    },
+    {
+      "epoch": 0.8835318036595992,
+      "grad_norm": 0.4776669442653656,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8836479814115596,
+      "grad_norm": 0.47542616724967957,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 7606
+    },
+    {
+      "epoch": 0.8837641591635201,
+      "grad_norm": 0.4733422100543976,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 7607
+    },
+    {
+      "epoch": 0.8838803369154807,
+      "grad_norm": 0.47752857208251953,
+      "learning_rate": 0.0001,
+      "loss": 1.4314,
+      "step": 7608
+    },
+    {
+      "epoch": 0.8839965146674412,
+      "grad_norm": 0.4597974121570587,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 7609
+    },
+    {
+      "epoch": 0.8841126924194017,
+      "grad_norm": 0.46217969059944153,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8842288701713622,
+      "grad_norm": 0.5063923001289368,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 7611
+    },
+    {
+      "epoch": 0.8843450479233227,
+      "grad_norm": 0.4687163233757019,
+      "learning_rate": 0.0001,
+      "loss": 1.6046,
+      "step": 7612
+    },
+    {
+      "epoch": 0.8844612256752832,
+      "grad_norm": 0.45635896921157837,
+      "learning_rate": 0.0001,
+      "loss": 1.4355,
+      "step": 7613
+    },
+    {
+      "epoch": 0.8845774034272437,
+      "grad_norm": 0.46440935134887695,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 7614
+    },
+    {
+      "epoch": 0.8846935811792042,
+      "grad_norm": 0.48333290219306946,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8848097589311646,
+      "grad_norm": 0.4907698929309845,
+      "learning_rate": 0.0001,
+      "loss": 1.6455,
+      "step": 7616
+    },
+    {
+      "epoch": 0.8849259366831251,
+      "grad_norm": 0.48579859733581543,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 7617
+    },
+    {
+      "epoch": 0.8850421144350857,
+      "grad_norm": 0.46631836891174316,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 7618
+    },
+    {
+      "epoch": 0.8851582921870462,
+      "grad_norm": 0.47939544916152954,
+      "learning_rate": 0.0001,
+      "loss": 1.6185,
+      "step": 7619
+    },
+    {
+      "epoch": 0.8852744699390067,
+      "grad_norm": 0.4607776701450348,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8853906476909672,
+      "grad_norm": 0.47920143604278564,
+      "learning_rate": 0.0001,
+      "loss": 1.7053,
+      "step": 7621
+    },
+    {
+      "epoch": 0.8855068254429277,
+      "grad_norm": 0.48284971714019775,
+      "learning_rate": 0.0001,
+      "loss": 1.6131,
+      "step": 7622
+    },
+    {
+      "epoch": 0.8856230031948882,
+      "grad_norm": 0.4558715224266052,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 7623
+    },
+    {
+      "epoch": 0.8857391809468487,
+      "grad_norm": 0.47470736503601074,
+      "learning_rate": 0.0001,
+      "loss": 1.669,
+      "step": 7624
+    },
+    {
+      "epoch": 0.8858553586988092,
+      "grad_norm": 0.4660652279853821,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8859715364507696,
+      "grad_norm": 0.5015971064567566,
+      "learning_rate": 0.0001,
+      "loss": 1.8973,
+      "step": 7626
+    },
+    {
+      "epoch": 0.8860877142027301,
+      "grad_norm": 0.48925477266311646,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 7627
+    },
+    {
+      "epoch": 0.8862038919546906,
+      "grad_norm": 0.4653438925743103,
+      "learning_rate": 0.0001,
+      "loss": 1.5789,
+      "step": 7628
+    },
+    {
+      "epoch": 0.8863200697066512,
+      "grad_norm": 0.4613575041294098,
+      "learning_rate": 0.0001,
+      "loss": 1.5857,
+      "step": 7629
+    },
+    {
+      "epoch": 0.8864362474586117,
+      "grad_norm": 0.4613388776779175,
+      "learning_rate": 0.0001,
+      "loss": 1.7134,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8865524252105722,
+      "grad_norm": 0.4807787835597992,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 7631
+    },
+    {
+      "epoch": 0.8866686029625327,
+      "grad_norm": 0.4956919252872467,
+      "learning_rate": 0.0001,
+      "loss": 1.5629,
+      "step": 7632
+    },
+    {
+      "epoch": 0.8867847807144932,
+      "grad_norm": 0.4471133053302765,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 7633
+    },
+    {
+      "epoch": 0.8869009584664537,
+      "grad_norm": 0.42097005248069763,
+      "learning_rate": 0.0001,
+      "loss": 1.4112,
+      "step": 7634
+    },
+    {
+      "epoch": 0.8870171362184142,
+      "grad_norm": 0.45755913853645325,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8871333139703746,
+      "grad_norm": 0.46528616547584534,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 7636
+    },
+    {
+      "epoch": 0.8872494917223351,
+      "grad_norm": 0.44611406326293945,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 7637
+    },
+    {
+      "epoch": 0.8873656694742956,
+      "grad_norm": 0.4587550759315491,
+      "learning_rate": 0.0001,
+      "loss": 1.6289,
+      "step": 7638
+    },
+    {
+      "epoch": 0.8874818472262562,
+      "grad_norm": 0.48842063546180725,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 7639
+    },
+    {
+      "epoch": 0.8875980249782167,
+      "grad_norm": 0.4626847505569458,
+      "learning_rate": 0.0001,
+      "loss": 1.5549,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8877142027301772,
+      "grad_norm": 0.45964398980140686,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 7641
+    },
+    {
+      "epoch": 0.8878303804821377,
+      "grad_norm": 0.5278240442276001,
+      "learning_rate": 0.0001,
+      "loss": 1.6286,
+      "step": 7642
+    },
+    {
+      "epoch": 0.8879465582340982,
+      "grad_norm": 0.4622969329357147,
+      "learning_rate": 0.0001,
+      "loss": 1.716,
+      "step": 7643
+    },
+    {
+      "epoch": 0.8880627359860587,
+      "grad_norm": 0.47453299164772034,
+      "learning_rate": 0.0001,
+      "loss": 1.6328,
+      "step": 7644
+    },
+    {
+      "epoch": 0.8881789137380192,
+      "grad_norm": 0.4532757103443146,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 7645
+    },
+    {
+      "epoch": 0.8882950914899796,
+      "grad_norm": 0.47530779242515564,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 7646
+    },
+    {
+      "epoch": 0.8884112692419401,
+      "grad_norm": 0.4514038562774658,
+      "learning_rate": 0.0001,
+      "loss": 1.462,
+      "step": 7647
+    },
+    {
+      "epoch": 0.8885274469939006,
+      "grad_norm": 0.48598358035087585,
+      "learning_rate": 0.0001,
+      "loss": 1.6245,
+      "step": 7648
+    },
+    {
+      "epoch": 0.8886436247458611,
+      "grad_norm": 0.4912118911743164,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 7649
+    },
+    {
+      "epoch": 0.8887598024978217,
+      "grad_norm": 0.4736858010292053,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8888759802497822,
+      "grad_norm": 0.4599950611591339,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 7651
+    },
+    {
+      "epoch": 0.8889921580017427,
+      "grad_norm": 0.45858752727508545,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 7652
+    },
+    {
+      "epoch": 0.8891083357537032,
+      "grad_norm": 0.4588874280452728,
+      "learning_rate": 0.0001,
+      "loss": 1.3829,
+      "step": 7653
+    },
+    {
+      "epoch": 0.8892245135056637,
+      "grad_norm": 0.4724344313144684,
+      "learning_rate": 0.0001,
+      "loss": 1.6388,
+      "step": 7654
+    },
+    {
+      "epoch": 0.8893406912576242,
+      "grad_norm": 0.46815189719200134,
+      "learning_rate": 0.0001,
+      "loss": 1.6232,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8894568690095846,
+      "grad_norm": 0.47574499249458313,
+      "learning_rate": 0.0001,
+      "loss": 1.6022,
+      "step": 7656
+    },
+    {
+      "epoch": 0.8895730467615451,
+      "grad_norm": 0.5131924152374268,
+      "learning_rate": 0.0001,
+      "loss": 1.7868,
+      "step": 7657
+    },
+    {
+      "epoch": 0.8896892245135056,
+      "grad_norm": 0.47800272703170776,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 7658
+    },
+    {
+      "epoch": 0.8898054022654661,
+      "grad_norm": 0.4923827052116394,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 7659
+    },
+    {
+      "epoch": 0.8899215800174267,
+      "grad_norm": 0.4540409743785858,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8900377577693872,
+      "grad_norm": 0.4510956108570099,
+      "learning_rate": 0.0001,
+      "loss": 1.428,
+      "step": 7661
+    },
+    {
+      "epoch": 0.8901539355213477,
+      "grad_norm": 0.47217369079589844,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 7662
+    },
+    {
+      "epoch": 0.8902701132733082,
+      "grad_norm": 0.5065828561782837,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 7663
+    },
+    {
+      "epoch": 0.8903862910252687,
+      "grad_norm": 0.47260862588882446,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 7664
+    },
+    {
+      "epoch": 0.8905024687772292,
+      "grad_norm": 0.4974091947078705,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 7665
+    },
+    {
+      "epoch": 0.8906186465291897,
+      "grad_norm": 0.5020241737365723,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 7666
+    },
+    {
+      "epoch": 0.8907348242811501,
+      "grad_norm": 0.47812047600746155,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 7667
+    },
+    {
+      "epoch": 0.8908510020331106,
+      "grad_norm": 0.5125158429145813,
+      "learning_rate": 0.0001,
+      "loss": 1.8338,
+      "step": 7668
+    },
+    {
+      "epoch": 0.8909671797850711,
+      "grad_norm": 0.4838445484638214,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 7669
+    },
+    {
+      "epoch": 0.8910833575370316,
+      "grad_norm": 0.4530005156993866,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8911995352889922,
+      "grad_norm": 0.4553331732749939,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 7671
+    },
+    {
+      "epoch": 0.8913157130409527,
+      "grad_norm": 0.46588602662086487,
+      "learning_rate": 0.0001,
+      "loss": 1.3147,
+      "step": 7672
+    },
+    {
+      "epoch": 0.8914318907929132,
+      "grad_norm": 0.46226534247398376,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 7673
+    },
+    {
+      "epoch": 0.8915480685448737,
+      "grad_norm": 0.506314754486084,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 7674
+    },
+    {
+      "epoch": 0.8916642462968342,
+      "grad_norm": 0.49548977613449097,
+      "learning_rate": 0.0001,
+      "loss": 1.7037,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8917804240487947,
+      "grad_norm": 0.4775904715061188,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 7676
+    },
+    {
+      "epoch": 0.8918966018007551,
+      "grad_norm": 0.4993993639945984,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 7677
+    },
+    {
+      "epoch": 0.8920127795527156,
+      "grad_norm": 0.49439752101898193,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 7678
+    },
+    {
+      "epoch": 0.8921289573046761,
+      "grad_norm": 0.4990589916706085,
+      "learning_rate": 0.0001,
+      "loss": 1.6649,
+      "step": 7679
+    },
+    {
+      "epoch": 0.8922451350566366,
+      "grad_norm": 0.4424578547477722,
+      "learning_rate": 0.0001,
+      "loss": 1.2997,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8923613128085972,
+      "grad_norm": 0.4604630768299103,
+      "learning_rate": 0.0001,
+      "loss": 1.6151,
+      "step": 7681
+    },
+    {
+      "epoch": 0.8924774905605577,
+      "grad_norm": 0.4625977575778961,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 7682
+    },
+    {
+      "epoch": 0.8925936683125182,
+      "grad_norm": 0.4803260266780853,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 7683
+    },
+    {
+      "epoch": 0.8927098460644787,
+      "grad_norm": 0.464473694562912,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 7684
+    },
+    {
+      "epoch": 0.8928260238164392,
+      "grad_norm": 0.44447073340415955,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8929422015683997,
+      "grad_norm": 0.44388943910598755,
+      "learning_rate": 0.0001,
+      "loss": 1.6461,
+      "step": 7686
+    },
+    {
+      "epoch": 0.8930583793203601,
+      "grad_norm": 0.5310772657394409,
+      "learning_rate": 0.0001,
+      "loss": 1.6304,
+      "step": 7687
+    },
+    {
+      "epoch": 0.8931745570723206,
+      "grad_norm": 0.4848250150680542,
+      "learning_rate": 0.0001,
+      "loss": 1.4879,
+      "step": 7688
+    },
+    {
+      "epoch": 0.8932907348242811,
+      "grad_norm": 0.4669688940048218,
+      "learning_rate": 0.0001,
+      "loss": 1.5178,
+      "step": 7689
+    },
+    {
+      "epoch": 0.8934069125762416,
+      "grad_norm": 0.47490620613098145,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8935230903282021,
+      "grad_norm": 0.49228766560554504,
+      "learning_rate": 0.0001,
+      "loss": 1.6218,
+      "step": 7691
+    },
+    {
+      "epoch": 0.8936392680801627,
+      "grad_norm": 0.5638328790664673,
+      "learning_rate": 0.0001,
+      "loss": 1.9294,
+      "step": 7692
+    },
+    {
+      "epoch": 0.8937554458321232,
+      "grad_norm": 0.48162275552749634,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 7693
+    },
+    {
+      "epoch": 0.8938716235840837,
+      "grad_norm": 0.4998082220554352,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 7694
+    },
+    {
+      "epoch": 0.8939878013360442,
+      "grad_norm": 0.4761600196361542,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8941039790880047,
+      "grad_norm": 0.4868890345096588,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 7696
+    },
+    {
+      "epoch": 0.8942201568399651,
+      "grad_norm": 0.4814538061618805,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 7697
+    },
+    {
+      "epoch": 0.8943363345919256,
+      "grad_norm": 0.46864473819732666,
+      "learning_rate": 0.0001,
+      "loss": 1.7473,
+      "step": 7698
+    },
+    {
+      "epoch": 0.8944525123438861,
+      "grad_norm": 0.514991044998169,
+      "learning_rate": 0.0001,
+      "loss": 1.6549,
+      "step": 7699
+    },
+    {
+      "epoch": 0.8945686900958466,
+      "grad_norm": 0.4469473361968994,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8946848678478071,
+      "grad_norm": 0.42899852991104126,
+      "learning_rate": 0.0001,
+      "loss": 1.4535,
+      "step": 7701
+    },
+    {
+      "epoch": 0.8948010455997677,
+      "grad_norm": 0.44255563616752625,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 7702
+    },
+    {
+      "epoch": 0.8949172233517282,
+      "grad_norm": 0.4918566942214966,
+      "learning_rate": 0.0001,
+      "loss": 1.578,
+      "step": 7703
+    },
+    {
+      "epoch": 0.8950334011036887,
+      "grad_norm": 0.506798505783081,
+      "learning_rate": 0.0001,
+      "loss": 1.6809,
+      "step": 7704
+    },
+    {
+      "epoch": 0.8951495788556492,
+      "grad_norm": 0.4588955044746399,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 7705
+    },
+    {
+      "epoch": 0.8952657566076097,
+      "grad_norm": 0.4813918173313141,
+      "learning_rate": 0.0001,
+      "loss": 1.5812,
+      "step": 7706
+    },
+    {
+      "epoch": 0.8953819343595701,
+      "grad_norm": 0.4796220064163208,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 7707
+    },
+    {
+      "epoch": 0.8954981121115306,
+      "grad_norm": 0.5280551910400391,
+      "learning_rate": 0.0001,
+      "loss": 1.7209,
+      "step": 7708
+    },
+    {
+      "epoch": 0.8956142898634911,
+      "grad_norm": 0.5038082599639893,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 7709
+    },
+    {
+      "epoch": 0.8957304676154516,
+      "grad_norm": 0.45545947551727295,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8958466453674121,
+      "grad_norm": 0.5030218958854675,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 7711
+    },
+    {
+      "epoch": 0.8959628231193726,
+      "grad_norm": 0.4473370909690857,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 7712
+    },
+    {
+      "epoch": 0.8960790008713332,
+      "grad_norm": 0.7040799856185913,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 7713
+    },
+    {
+      "epoch": 0.8961951786232937,
+      "grad_norm": 0.46077761054039,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 7714
+    },
+    {
+      "epoch": 0.8963113563752542,
+      "grad_norm": 0.44979485869407654,
+      "learning_rate": 0.0001,
+      "loss": 1.5214,
+      "step": 7715
+    },
+    {
+      "epoch": 0.8964275341272147,
+      "grad_norm": 0.46903368830680847,
+      "learning_rate": 0.0001,
+      "loss": 1.5326,
+      "step": 7716
+    },
+    {
+      "epoch": 0.8965437118791751,
+      "grad_norm": 0.4809918701648712,
+      "learning_rate": 0.0001,
+      "loss": 1.6342,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8966598896311356,
+      "grad_norm": 0.4725591838359833,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 7718
+    },
+    {
+      "epoch": 0.8967760673830961,
+      "grad_norm": 0.49973586201667786,
+      "learning_rate": 0.0001,
+      "loss": 1.783,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8968922451350566,
+      "grad_norm": 0.4993938207626343,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8970084228870171,
+      "grad_norm": 0.5148173570632935,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 7721
+    },
+    {
+      "epoch": 0.8971246006389776,
+      "grad_norm": 0.49831125140190125,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 7722
+    },
+    {
+      "epoch": 0.8972407783909382,
+      "grad_norm": 0.46448323130607605,
+      "learning_rate": 0.0001,
+      "loss": 1.6107,
+      "step": 7723
+    },
+    {
+      "epoch": 0.8973569561428987,
+      "grad_norm": 0.47163113951683044,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 7724
+    },
+    {
+      "epoch": 0.8974731338948592,
+      "grad_norm": 0.45095062255859375,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8975893116468197,
+      "grad_norm": 0.4865681231021881,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8977054893987801,
+      "grad_norm": 0.4516942501068115,
+      "learning_rate": 0.0001,
+      "loss": 1.6462,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8978216671507406,
+      "grad_norm": 0.45870745182037354,
+      "learning_rate": 0.0001,
+      "loss": 1.5657,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8979378449027011,
+      "grad_norm": 0.5079413056373596,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 7729
+    },
+    {
+      "epoch": 0.8980540226546616,
+      "grad_norm": 0.46579909324645996,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 7730
+    },
+    {
+      "epoch": 0.8981702004066221,
+      "grad_norm": 0.46913209557533264,
+      "learning_rate": 0.0001,
+      "loss": 1.6732,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8982863781585826,
+      "grad_norm": 0.46891793608665466,
+      "learning_rate": 0.0001,
+      "loss": 1.5816,
+      "step": 7732
+    },
+    {
+      "epoch": 0.8984025559105431,
+      "grad_norm": 0.5008355379104614,
+      "learning_rate": 0.0001,
+      "loss": 1.654,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8985187336625037,
+      "grad_norm": 0.48343855142593384,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 7734
+    },
+    {
+      "epoch": 0.8986349114144642,
+      "grad_norm": 0.47268950939178467,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8987510891664247,
+      "grad_norm": 0.5122753381729126,
+      "learning_rate": 0.0001,
+      "loss": 1.7408,
+      "step": 7736
+    },
+    {
+      "epoch": 0.8988672669183851,
+      "grad_norm": 0.45394930243492126,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 7737
+    },
+    {
+      "epoch": 0.8989834446703456,
+      "grad_norm": 0.4667399823665619,
+      "learning_rate": 0.0001,
+      "loss": 1.6136,
+      "step": 7738
+    },
+    {
+      "epoch": 0.8990996224223061,
+      "grad_norm": 0.5103424191474915,
+      "learning_rate": 0.0001,
+      "loss": 1.7418,
+      "step": 7739
+    },
+    {
+      "epoch": 0.8992158001742666,
+      "grad_norm": 0.5049645304679871,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8993319779262271,
+      "grad_norm": 0.45100435614585876,
+      "learning_rate": 0.0001,
+      "loss": 1.3736,
+      "step": 7741
+    },
+    {
+      "epoch": 0.8994481556781876,
+      "grad_norm": 0.4995354115962982,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 7742
+    },
+    {
+      "epoch": 0.8995643334301481,
+      "grad_norm": 0.5154250264167786,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8996805111821087,
+      "grad_norm": 0.4704994261264801,
+      "learning_rate": 0.0001,
+      "loss": 1.6447,
+      "step": 7744
+    },
+    {
+      "epoch": 0.8997966889340692,
+      "grad_norm": 0.4912912845611572,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8999128666860297,
+      "grad_norm": 0.4594641327857971,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 7746
+    },
+    {
+      "epoch": 0.9000290444379901,
+      "grad_norm": 0.48515209555625916,
+      "learning_rate": 0.0001,
+      "loss": 1.4787,
+      "step": 7747
+    },
+    {
+      "epoch": 0.9001452221899506,
+      "grad_norm": 0.4500701129436493,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 7748
+    },
+    {
+      "epoch": 0.9002613999419111,
+      "grad_norm": 0.5031145811080933,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 7749
+    },
+    {
+      "epoch": 0.9003775776938716,
+      "grad_norm": 0.46528854966163635,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 7750
+    },
+    {
+      "epoch": 0.9004937554458321,
+      "grad_norm": 0.4735311269760132,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 7751
+    },
+    {
+      "epoch": 0.9006099331977926,
+      "grad_norm": 0.4692305028438568,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 7752
+    },
+    {
+      "epoch": 0.9007261109497531,
+      "grad_norm": 0.48819053173065186,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 7753
+    },
+    {
+      "epoch": 0.9008422887017136,
+      "grad_norm": 0.494018018245697,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 7754
+    },
+    {
+      "epoch": 0.9009584664536742,
+      "grad_norm": 0.48794278502464294,
+      "learning_rate": 0.0001,
+      "loss": 1.6929,
+      "step": 7755
+    },
+    {
+      "epoch": 0.9010746442056347,
+      "grad_norm": 0.48159360885620117,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 7756
+    },
+    {
+      "epoch": 0.9011908219575951,
+      "grad_norm": 0.5053829550743103,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 7757
+    },
+    {
+      "epoch": 0.9013069997095556,
+      "grad_norm": 0.47055330872535706,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 7758
+    },
+    {
+      "epoch": 0.9014231774615161,
+      "grad_norm": 0.46060261130332947,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 7759
+    },
+    {
+      "epoch": 0.9015393552134766,
+      "grad_norm": 0.46970701217651367,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 7760
+    },
+    {
+      "epoch": 0.9016555329654371,
+      "grad_norm": 0.45572981238365173,
+      "learning_rate": 0.0001,
+      "loss": 1.4527,
+      "step": 7761
+    },
+    {
+      "epoch": 0.9017717107173976,
+      "grad_norm": 0.44444963335990906,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 7762
+    },
+    {
+      "epoch": 0.9018878884693581,
+      "grad_norm": 0.43386489152908325,
+      "learning_rate": 0.0001,
+      "loss": 1.3637,
+      "step": 7763
+    },
+    {
+      "epoch": 0.9020040662213186,
+      "grad_norm": 0.4735313057899475,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 7764
+    },
+    {
+      "epoch": 0.9021202439732792,
+      "grad_norm": 0.5182105898857117,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 7765
+    },
+    {
+      "epoch": 0.9022364217252397,
+      "grad_norm": 0.5161397457122803,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 7766
+    },
+    {
+      "epoch": 0.9023525994772001,
+      "grad_norm": 0.45563217997550964,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 7767
+    },
+    {
+      "epoch": 0.9024687772291606,
+      "grad_norm": 0.5249543786048889,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 7768
+    },
+    {
+      "epoch": 0.9025849549811211,
+      "grad_norm": 0.49920180439949036,
+      "learning_rate": 0.0001,
+      "loss": 1.7509,
+      "step": 7769
+    },
+    {
+      "epoch": 0.9027011327330816,
+      "grad_norm": 0.43166783452033997,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 7770
+    },
+    {
+      "epoch": 0.9028173104850421,
+      "grad_norm": 0.4483865201473236,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 7771
+    },
+    {
+      "epoch": 0.9029334882370026,
+      "grad_norm": 0.43736976385116577,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 7772
+    },
+    {
+      "epoch": 0.9030496659889631,
+      "grad_norm": 0.48580455780029297,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 7773
+    },
+    {
+      "epoch": 0.9031658437409236,
+      "grad_norm": 0.47388365864753723,
+      "learning_rate": 0.0001,
+      "loss": 1.5931,
+      "step": 7774
+    },
+    {
+      "epoch": 0.9032820214928841,
+      "grad_norm": 0.49823474884033203,
+      "learning_rate": 0.0001,
+      "loss": 1.6047,
+      "step": 7775
+    },
+    {
+      "epoch": 0.9033981992448447,
+      "grad_norm": 0.451608270406723,
+      "learning_rate": 0.0001,
+      "loss": 1.5266,
+      "step": 7776
+    },
+    {
+      "epoch": 0.9035143769968051,
+      "grad_norm": 0.4777389168739319,
+      "learning_rate": 0.0001,
+      "loss": 1.7107,
+      "step": 7777
+    },
+    {
+      "epoch": 0.9036305547487656,
+      "grad_norm": 0.4792795777320862,
+      "learning_rate": 0.0001,
+      "loss": 1.655,
+      "step": 7778
+    },
+    {
+      "epoch": 0.9037467325007261,
+      "grad_norm": 0.4765215814113617,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 7779
+    },
+    {
+      "epoch": 0.9038629102526866,
+      "grad_norm": 0.46686357259750366,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 7780
+    },
+    {
+      "epoch": 0.9039790880046471,
+      "grad_norm": 0.5087867379188538,
+      "learning_rate": 0.0001,
+      "loss": 1.9357,
+      "step": 7781
+    },
+    {
+      "epoch": 0.9040952657566076,
+      "grad_norm": 0.46744218468666077,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 7782
+    },
+    {
+      "epoch": 0.9042114435085681,
+      "grad_norm": 0.5227089524269104,
+      "learning_rate": 0.0001,
+      "loss": 1.5938,
+      "step": 7783
+    },
+    {
+      "epoch": 0.9043276212605286,
+      "grad_norm": 0.4557849168777466,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 7784
+    },
+    {
+      "epoch": 0.9044437990124891,
+      "grad_norm": 0.4599103033542633,
+      "learning_rate": 0.0001,
+      "loss": 1.6792,
+      "step": 7785
+    },
+    {
+      "epoch": 0.9045599767644497,
+      "grad_norm": 0.5397565364837646,
+      "learning_rate": 0.0001,
+      "loss": 1.8295,
+      "step": 7786
+    },
+    {
+      "epoch": 0.9046761545164101,
+      "grad_norm": 0.5359669923782349,
+      "learning_rate": 0.0001,
+      "loss": 1.7149,
+      "step": 7787
+    },
+    {
+      "epoch": 0.9047923322683706,
+      "grad_norm": 0.49455010890960693,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 7788
+    },
+    {
+      "epoch": 0.9049085100203311,
+      "grad_norm": 0.4927608072757721,
+      "learning_rate": 0.0001,
+      "loss": 1.668,
+      "step": 7789
+    },
+    {
+      "epoch": 0.9050246877722916,
+      "grad_norm": 0.45446282625198364,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 7790
+    },
+    {
+      "epoch": 0.9051408655242521,
+      "grad_norm": 0.4715528190135956,
+      "learning_rate": 0.0001,
+      "loss": 1.5217,
+      "step": 7791
+    },
+    {
+      "epoch": 0.9052570432762126,
+      "grad_norm": 0.47455599904060364,
+      "learning_rate": 0.0001,
+      "loss": 1.6861,
+      "step": 7792
+    },
+    {
+      "epoch": 0.9053732210281731,
+      "grad_norm": 0.4646841585636139,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 7793
+    },
+    {
+      "epoch": 0.9054893987801336,
+      "grad_norm": 0.4721209406852722,
+      "learning_rate": 0.0001,
+      "loss": 1.6885,
+      "step": 7794
+    },
+    {
+      "epoch": 0.9056055765320941,
+      "grad_norm": 0.4585552215576172,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 7795
+    },
+    {
+      "epoch": 0.9057217542840547,
+      "grad_norm": 0.49558770656585693,
+      "learning_rate": 0.0001,
+      "loss": 1.6104,
+      "step": 7796
+    },
+    {
+      "epoch": 0.9058379320360151,
+      "grad_norm": 0.45324012637138367,
+      "learning_rate": 0.0001,
+      "loss": 1.6386,
+      "step": 7797
+    },
+    {
+      "epoch": 0.9059541097879756,
+      "grad_norm": 0.44756069779396057,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 7798
+    },
+    {
+      "epoch": 0.9060702875399361,
+      "grad_norm": 0.48996299505233765,
+      "learning_rate": 0.0001,
+      "loss": 1.8288,
+      "step": 7799
+    },
+    {
+      "epoch": 0.9061864652918966,
+      "grad_norm": 0.47252756357192993,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 7800
+    },
+    {
+      "epoch": 0.9063026430438571,
+      "grad_norm": 0.4882994294166565,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 7801
+    },
+    {
+      "epoch": 0.9064188207958176,
+      "grad_norm": 0.49475234746932983,
+      "learning_rate": 0.0001,
+      "loss": 1.7413,
+      "step": 7802
+    },
+    {
+      "epoch": 0.9065349985477781,
+      "grad_norm": 0.467546671628952,
+      "learning_rate": 0.0001,
+      "loss": 1.3964,
+      "step": 7803
+    },
+    {
+      "epoch": 0.9066511762997386,
+      "grad_norm": 0.47339338064193726,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 7804
+    },
+    {
+      "epoch": 0.9067673540516991,
+      "grad_norm": 0.4243144094944,
+      "learning_rate": 0.0001,
+      "loss": 1.2828,
+      "step": 7805
+    },
+    {
+      "epoch": 0.9068835318036595,
+      "grad_norm": 0.4705776572227478,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 7806
+    },
+    {
+      "epoch": 0.9069997095556201,
+      "grad_norm": 0.48262354731559753,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 7807
+    },
+    {
+      "epoch": 0.9071158873075806,
+      "grad_norm": 0.46224817633628845,
+      "learning_rate": 0.0001,
+      "loss": 1.4478,
+      "step": 7808
+    },
+    {
+      "epoch": 0.9072320650595411,
+      "grad_norm": 0.48688849806785583,
+      "learning_rate": 0.0001,
+      "loss": 1.6116,
+      "step": 7809
+    },
+    {
+      "epoch": 0.9073482428115016,
+      "grad_norm": 0.47182363271713257,
+      "learning_rate": 0.0001,
+      "loss": 1.6092,
+      "step": 7810
+    },
+    {
+      "epoch": 0.9074644205634621,
+      "grad_norm": 0.5122577548027039,
+      "learning_rate": 0.0001,
+      "loss": 1.5642,
+      "step": 7811
+    },
+    {
+      "epoch": 0.9075805983154226,
+      "grad_norm": 0.4881529211997986,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 7812
+    },
+    {
+      "epoch": 0.9076967760673831,
+      "grad_norm": 0.5338044166564941,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 7813
+    },
+    {
+      "epoch": 0.9078129538193436,
+      "grad_norm": 0.4970211684703827,
+      "learning_rate": 0.0001,
+      "loss": 1.7056,
+      "step": 7814
+    },
+    {
+      "epoch": 0.9079291315713041,
+      "grad_norm": 0.574657142162323,
+      "learning_rate": 0.0001,
+      "loss": 1.3376,
+      "step": 7815
+    },
+    {
+      "epoch": 0.9080453093232645,
+      "grad_norm": 0.44657719135284424,
+      "learning_rate": 0.0001,
+      "loss": 1.4165,
+      "step": 7816
+    },
+    {
+      "epoch": 0.9081614870752251,
+      "grad_norm": 0.4609968066215515,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 7817
+    },
+    {
+      "epoch": 0.9082776648271856,
+      "grad_norm": 0.4657510221004486,
+      "learning_rate": 0.0001,
+      "loss": 1.6129,
+      "step": 7818
+    },
+    {
+      "epoch": 0.9083938425791461,
+      "grad_norm": 0.4992734491825104,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 7819
+    },
+    {
+      "epoch": 0.9085100203311066,
+      "grad_norm": 0.5076075196266174,
+      "learning_rate": 0.0001,
+      "loss": 1.7326,
+      "step": 7820
+    },
+    {
+      "epoch": 0.9086261980830671,
+      "grad_norm": 0.4936392605304718,
+      "learning_rate": 0.0001,
+      "loss": 1.7559,
+      "step": 7821
+    },
+    {
+      "epoch": 0.9087423758350276,
+      "grad_norm": 0.47953271865844727,
+      "learning_rate": 0.0001,
+      "loss": 1.5399,
+      "step": 7822
+    },
+    {
+      "epoch": 0.9088585535869881,
+      "grad_norm": 0.4652204215526581,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 7823
+    },
+    {
+      "epoch": 0.9089747313389486,
+      "grad_norm": 0.4872530698776245,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 7824
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 0.4713476300239563,
+      "learning_rate": 0.0001,
+      "loss": 1.5699,
+      "step": 7825
+    },
+    {
+      "epoch": 0.9092070868428695,
+      "grad_norm": 0.4624953269958496,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 7826
+    },
+    {
+      "epoch": 0.90932326459483,
+      "grad_norm": 0.49820998311042786,
+      "learning_rate": 0.0001,
+      "loss": 1.6088,
+      "step": 7827
+    },
+    {
+      "epoch": 0.9094394423467906,
+      "grad_norm": 0.4489380419254303,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 7828
+    },
+    {
+      "epoch": 0.9095556200987511,
+      "grad_norm": 0.48509761691093445,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 7829
+    },
+    {
+      "epoch": 0.9096717978507116,
+      "grad_norm": 0.5044791102409363,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 7830
+    },
+    {
+      "epoch": 0.9097879756026721,
+      "grad_norm": 0.47124698758125305,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 7831
+    },
+    {
+      "epoch": 0.9099041533546326,
+      "grad_norm": 0.4781220555305481,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 7832
+    },
+    {
+      "epoch": 0.9100203311065931,
+      "grad_norm": 0.47968727350234985,
+      "learning_rate": 0.0001,
+      "loss": 1.7076,
+      "step": 7833
+    },
+    {
+      "epoch": 0.9101365088585536,
+      "grad_norm": 0.5137500166893005,
+      "learning_rate": 0.0001,
+      "loss": 1.7598,
+      "step": 7834
+    },
+    {
+      "epoch": 0.9102526866105141,
+      "grad_norm": 0.4631037414073944,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 7835
+    },
+    {
+      "epoch": 0.9103688643624746,
+      "grad_norm": 0.5219733715057373,
+      "learning_rate": 0.0001,
+      "loss": 1.7391,
+      "step": 7836
+    },
+    {
+      "epoch": 0.910485042114435,
+      "grad_norm": 0.4807119071483612,
+      "learning_rate": 0.0001,
+      "loss": 1.6519,
+      "step": 7837
+    },
+    {
+      "epoch": 0.9106012198663956,
+      "grad_norm": 0.5032156705856323,
+      "learning_rate": 0.0001,
+      "loss": 1.7154,
+      "step": 7838
+    },
+    {
+      "epoch": 0.9107173976183561,
+      "grad_norm": 0.7016561031341553,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 7839
+    },
+    {
+      "epoch": 0.9108335753703166,
+      "grad_norm": 0.4516977071762085,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 7840
+    },
+    {
+      "epoch": 0.9109497531222771,
+      "grad_norm": 0.4732464849948883,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 7841
+    },
+    {
+      "epoch": 0.9110659308742376,
+      "grad_norm": 0.4913138151168823,
+      "learning_rate": 0.0001,
+      "loss": 1.6617,
+      "step": 7842
+    },
+    {
+      "epoch": 0.9111821086261981,
+      "grad_norm": 0.5138623118400574,
+      "learning_rate": 0.0001,
+      "loss": 1.7446,
+      "step": 7843
+    },
+    {
+      "epoch": 0.9112982863781586,
+      "grad_norm": 0.46417149901390076,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 7844
+    },
+    {
+      "epoch": 0.9114144641301191,
+      "grad_norm": 0.47088223695755005,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 7845
+    },
+    {
+      "epoch": 0.9115306418820796,
+      "grad_norm": 0.45091691613197327,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 7846
+    },
+    {
+      "epoch": 0.91164681963404,
+      "grad_norm": 0.45545732975006104,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 7847
+    },
+    {
+      "epoch": 0.9117629973860005,
+      "grad_norm": 0.45981520414352417,
+      "learning_rate": 0.0001,
+      "loss": 1.4383,
+      "step": 7848
+    },
+    {
+      "epoch": 0.9118791751379611,
+      "grad_norm": 0.472409725189209,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 7849
+    },
+    {
+      "epoch": 0.9119953528899216,
+      "grad_norm": 0.5488510131835938,
+      "learning_rate": 0.0001,
+      "loss": 1.7462,
+      "step": 7850
+    },
+    {
+      "epoch": 0.9121115306418821,
+      "grad_norm": 0.48794418573379517,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 7851
+    },
+    {
+      "epoch": 0.9122277083938426,
+      "grad_norm": 0.4115673899650574,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 7852
+    },
+    {
+      "epoch": 0.9123438861458031,
+      "grad_norm": 0.46749910712242126,
+      "learning_rate": 0.0001,
+      "loss": 1.6613,
+      "step": 7853
+    },
+    {
+      "epoch": 0.9124600638977636,
+      "grad_norm": 0.47917798161506653,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 7854
+    },
+    {
+      "epoch": 0.9125762416497241,
+      "grad_norm": 0.46126171946525574,
+      "learning_rate": 0.0001,
+      "loss": 1.6636,
+      "step": 7855
+    },
+    {
+      "epoch": 0.9126924194016846,
+      "grad_norm": 0.4459565281867981,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 7856
+    },
+    {
+      "epoch": 0.912808597153645,
+      "grad_norm": 0.4595984220504761,
+      "learning_rate": 0.0001,
+      "loss": 1.624,
+      "step": 7857
+    },
+    {
+      "epoch": 0.9129247749056055,
+      "grad_norm": 0.45594367384910583,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 7858
+    },
+    {
+      "epoch": 0.9130409526575661,
+      "grad_norm": 0.4766393303871155,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 7859
+    },
+    {
+      "epoch": 0.9131571304095266,
+      "grad_norm": 0.5518864393234253,
+      "learning_rate": 0.0001,
+      "loss": 1.7039,
+      "step": 7860
+    },
+    {
+      "epoch": 0.9132733081614871,
+      "grad_norm": 0.48562371730804443,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 7861
+    },
+    {
+      "epoch": 0.9133894859134476,
+      "grad_norm": 0.5005999207496643,
+      "learning_rate": 0.0001,
+      "loss": 1.7962,
+      "step": 7862
+    },
+    {
+      "epoch": 0.9135056636654081,
+      "grad_norm": 0.4571722149848938,
+      "learning_rate": 0.0001,
+      "loss": 1.5681,
+      "step": 7863
+    },
+    {
+      "epoch": 0.9136218414173686,
+      "grad_norm": 0.4847254753112793,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 7864
+    },
+    {
+      "epoch": 0.9137380191693291,
+      "grad_norm": 0.5055410265922546,
+      "learning_rate": 0.0001,
+      "loss": 1.7978,
+      "step": 7865
+    },
+    {
+      "epoch": 0.9138541969212896,
+      "grad_norm": 0.5655259490013123,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 7866
+    },
+    {
+      "epoch": 0.91397037467325,
+      "grad_norm": 0.46798625588417053,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 7867
+    },
+    {
+      "epoch": 0.9140865524252105,
+      "grad_norm": 0.48041674494743347,
+      "learning_rate": 0.0001,
+      "loss": 1.6535,
+      "step": 7868
+    },
+    {
+      "epoch": 0.914202730177171,
+      "grad_norm": 0.4875914752483368,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 7869
+    },
+    {
+      "epoch": 0.9143189079291316,
+      "grad_norm": 0.49609968066215515,
+      "learning_rate": 0.0001,
+      "loss": 1.7816,
+      "step": 7870
+    },
+    {
+      "epoch": 0.9144350856810921,
+      "grad_norm": 0.43055441975593567,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 7871
+    },
+    {
+      "epoch": 0.9145512634330526,
+      "grad_norm": 0.4403667747974396,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 7872
+    },
+    {
+      "epoch": 0.9146674411850131,
+      "grad_norm": 0.4711524546146393,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 7873
+    },
+    {
+      "epoch": 0.9147836189369736,
+      "grad_norm": 0.49612268805503845,
+      "learning_rate": 0.0001,
+      "loss": 1.7506,
+      "step": 7874
+    },
+    {
+      "epoch": 0.9148997966889341,
+      "grad_norm": 0.4640529751777649,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 7875
+    },
+    {
+      "epoch": 0.9150159744408946,
+      "grad_norm": 0.45342904329299927,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 7876
+    },
+    {
+      "epoch": 0.915132152192855,
+      "grad_norm": 0.4637288749217987,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 7877
+    },
+    {
+      "epoch": 0.9152483299448155,
+      "grad_norm": 0.5066947937011719,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 7878
+    },
+    {
+      "epoch": 0.915364507696776,
+      "grad_norm": 0.47863486409187317,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 7879
+    },
+    {
+      "epoch": 0.9154806854487366,
+      "grad_norm": 0.4799175262451172,
+      "learning_rate": 0.0001,
+      "loss": 1.6418,
+      "step": 7880
+    },
+    {
+      "epoch": 0.9155968632006971,
+      "grad_norm": 0.49706557393074036,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 7881
+    },
+    {
+      "epoch": 0.9157130409526576,
+      "grad_norm": 0.49601349234580994,
+      "learning_rate": 0.0001,
+      "loss": 1.6441,
+      "step": 7882
+    },
+    {
+      "epoch": 0.9158292187046181,
+      "grad_norm": 0.5072823762893677,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 7883
+    },
+    {
+      "epoch": 0.9159453964565786,
+      "grad_norm": 0.5159144401550293,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 7884
+    },
+    {
+      "epoch": 0.9160615742085391,
+      "grad_norm": 0.46649453043937683,
+      "learning_rate": 0.0001,
+      "loss": 1.489,
+      "step": 7885
+    },
+    {
+      "epoch": 0.9161777519604996,
+      "grad_norm": 0.48782074451446533,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 7886
+    },
+    {
+      "epoch": 0.91629392971246,
+      "grad_norm": 0.5250930190086365,
+      "learning_rate": 0.0001,
+      "loss": 1.6332,
+      "step": 7887
+    },
+    {
+      "epoch": 0.9164101074644205,
+      "grad_norm": 0.4878339469432831,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 7888
+    },
+    {
+      "epoch": 0.916526285216381,
+      "grad_norm": 0.45729678869247437,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 7889
+    },
+    {
+      "epoch": 0.9166424629683415,
+      "grad_norm": 0.47600996494293213,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 7890
+    },
+    {
+      "epoch": 0.9167586407203021,
+      "grad_norm": 0.5157418847084045,
+      "learning_rate": 0.0001,
+      "loss": 1.6465,
+      "step": 7891
+    },
+    {
+      "epoch": 0.9168748184722626,
+      "grad_norm": 0.5000462532043457,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 7892
+    },
+    {
+      "epoch": 0.9169909962242231,
+      "grad_norm": 0.48779532313346863,
+      "learning_rate": 0.0001,
+      "loss": 1.3709,
+      "step": 7893
+    },
+    {
+      "epoch": 0.9171071739761836,
+      "grad_norm": 0.49037328362464905,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 7894
+    },
+    {
+      "epoch": 0.9172233517281441,
+      "grad_norm": 0.47151613235473633,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 7895
+    },
+    {
+      "epoch": 0.9173395294801046,
+      "grad_norm": 0.5137687921524048,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 7896
+    },
+    {
+      "epoch": 0.917455707232065,
+      "grad_norm": 0.4644807279109955,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 7897
+    },
+    {
+      "epoch": 0.9175718849840255,
+      "grad_norm": 0.45312634110450745,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 7898
+    },
+    {
+      "epoch": 0.917688062735986,
+      "grad_norm": 0.5129163861274719,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 7899
+    },
+    {
+      "epoch": 0.9178042404879465,
+      "grad_norm": 0.474354088306427,
+      "learning_rate": 0.0001,
+      "loss": 1.5625,
+      "step": 7900
+    },
+    {
+      "epoch": 0.9179204182399071,
+      "grad_norm": 0.47130098938941956,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 7901
+    },
+    {
+      "epoch": 0.9180365959918676,
+      "grad_norm": 0.4573417901992798,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 7902
+    },
+    {
+      "epoch": 0.9181527737438281,
+      "grad_norm": 0.46235814690589905,
+      "learning_rate": 0.0001,
+      "loss": 1.758,
+      "step": 7903
+    },
+    {
+      "epoch": 0.9182689514957886,
+      "grad_norm": 0.5168223977088928,
+      "learning_rate": 0.0001,
+      "loss": 1.2772,
+      "step": 7904
+    },
+    {
+      "epoch": 0.9183851292477491,
+      "grad_norm": 0.4721146523952484,
+      "learning_rate": 0.0001,
+      "loss": 1.6981,
+      "step": 7905
+    },
+    {
+      "epoch": 0.9185013069997096,
+      "grad_norm": 0.49531880021095276,
+      "learning_rate": 0.0001,
+      "loss": 1.5861,
+      "step": 7906
+    },
+    {
+      "epoch": 0.91861748475167,
+      "grad_norm": 0.49574729800224304,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 7907
+    },
+    {
+      "epoch": 0.9187336625036305,
+      "grad_norm": 0.4491695761680603,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 7908
+    },
+    {
+      "epoch": 0.918849840255591,
+      "grad_norm": 0.45107486844062805,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 7909
+    },
+    {
+      "epoch": 0.9189660180075515,
+      "grad_norm": 0.4571503698825836,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 7910
+    },
+    {
+      "epoch": 0.919082195759512,
+      "grad_norm": 0.49854135513305664,
+      "learning_rate": 0.0001,
+      "loss": 1.7238,
+      "step": 7911
+    },
+    {
+      "epoch": 0.9191983735114726,
+      "grad_norm": 0.4973702132701874,
+      "learning_rate": 0.0001,
+      "loss": 1.5851,
+      "step": 7912
+    },
+    {
+      "epoch": 0.9193145512634331,
+      "grad_norm": 0.47900450229644775,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 7913
+    },
+    {
+      "epoch": 0.9194307290153936,
+      "grad_norm": 0.42754286527633667,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 7914
+    },
+    {
+      "epoch": 0.9195469067673541,
+      "grad_norm": 0.5205299854278564,
+      "learning_rate": 0.0001,
+      "loss": 1.6868,
+      "step": 7915
+    },
+    {
+      "epoch": 0.9196630845193146,
+      "grad_norm": 0.49486666917800903,
+      "learning_rate": 0.0001,
+      "loss": 1.685,
+      "step": 7916
+    },
+    {
+      "epoch": 0.919779262271275,
+      "grad_norm": 0.47060513496398926,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 7917
+    },
+    {
+      "epoch": 0.9198954400232355,
+      "grad_norm": 0.4912223517894745,
+      "learning_rate": 0.0001,
+      "loss": 1.7563,
+      "step": 7918
+    },
+    {
+      "epoch": 0.920011617775196,
+      "grad_norm": 0.46059104800224304,
+      "learning_rate": 0.0001,
+      "loss": 1.3697,
+      "step": 7919
+    },
+    {
+      "epoch": 0.9201277955271565,
+      "grad_norm": 0.4507901072502136,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 7920
+    },
+    {
+      "epoch": 0.920243973279117,
+      "grad_norm": 0.49140775203704834,
+      "learning_rate": 0.0001,
+      "loss": 1.4478,
+      "step": 7921
+    },
+    {
+      "epoch": 0.9203601510310776,
+      "grad_norm": 0.4491073489189148,
+      "learning_rate": 0.0001,
+      "loss": 1.3941,
+      "step": 7922
+    },
+    {
+      "epoch": 0.9204763287830381,
+      "grad_norm": 0.49387529492378235,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 7923
+    },
+    {
+      "epoch": 0.9205925065349986,
+      "grad_norm": 0.4687308669090271,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 7924
+    },
+    {
+      "epoch": 0.9207086842869591,
+      "grad_norm": 0.5228381752967834,
+      "learning_rate": 0.0001,
+      "loss": 1.8851,
+      "step": 7925
+    },
+    {
+      "epoch": 0.9208248620389196,
+      "grad_norm": 0.49759143590927124,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 7926
+    },
+    {
+      "epoch": 0.92094103979088,
+      "grad_norm": 0.47675225138664246,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 7927
+    },
+    {
+      "epoch": 0.9210572175428405,
+      "grad_norm": 0.45978063344955444,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 7928
+    },
+    {
+      "epoch": 0.921173395294801,
+      "grad_norm": 0.5017638802528381,
+      "learning_rate": 0.0001,
+      "loss": 1.7266,
+      "step": 7929
+    },
+    {
+      "epoch": 0.9212895730467615,
+      "grad_norm": 0.467979371547699,
+      "learning_rate": 0.0001,
+      "loss": 1.6362,
+      "step": 7930
+    },
+    {
+      "epoch": 0.921405750798722,
+      "grad_norm": 0.4964231848716736,
+      "learning_rate": 0.0001,
+      "loss": 1.4573,
+      "step": 7931
+    },
+    {
+      "epoch": 0.9215219285506825,
+      "grad_norm": 0.46611857414245605,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 7932
+    },
+    {
+      "epoch": 0.9216381063026431,
+      "grad_norm": 0.47594428062438965,
+      "learning_rate": 0.0001,
+      "loss": 1.6423,
+      "step": 7933
+    },
+    {
+      "epoch": 0.9217542840546036,
+      "grad_norm": 0.4936331510543823,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 7934
+    },
+    {
+      "epoch": 0.9218704618065641,
+      "grad_norm": 0.4947721064090729,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 7935
+    },
+    {
+      "epoch": 0.9219866395585246,
+      "grad_norm": 0.4754369854927063,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 7936
+    },
+    {
+      "epoch": 0.922102817310485,
+      "grad_norm": 0.46831005811691284,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 7937
+    },
+    {
+      "epoch": 0.9222189950624455,
+      "grad_norm": 0.498220294713974,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 7938
+    },
+    {
+      "epoch": 0.922335172814406,
+      "grad_norm": 0.4809699058532715,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 7939
+    },
+    {
+      "epoch": 0.9224513505663665,
+      "grad_norm": 0.4623294174671173,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 7940
+    },
+    {
+      "epoch": 0.922567528318327,
+      "grad_norm": 0.49049609899520874,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 7941
+    },
+    {
+      "epoch": 0.9226837060702875,
+      "grad_norm": 0.47794732451438904,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 7942
+    },
+    {
+      "epoch": 0.9227998838222481,
+      "grad_norm": 0.49489712715148926,
+      "learning_rate": 0.0001,
+      "loss": 1.7071,
+      "step": 7943
+    },
+    {
+      "epoch": 0.9229160615742086,
+      "grad_norm": 0.4838644564151764,
+      "learning_rate": 0.0001,
+      "loss": 1.6534,
+      "step": 7944
+    },
+    {
+      "epoch": 0.9230322393261691,
+      "grad_norm": 0.542927622795105,
+      "learning_rate": 0.0001,
+      "loss": 1.8534,
+      "step": 7945
+    },
+    {
+      "epoch": 0.9231484170781296,
+      "grad_norm": 0.484602153301239,
+      "learning_rate": 0.0001,
+      "loss": 1.7658,
+      "step": 7946
+    },
+    {
+      "epoch": 0.92326459483009,
+      "grad_norm": 0.45211806893348694,
+      "learning_rate": 0.0001,
+      "loss": 1.3895,
+      "step": 7947
+    },
+    {
+      "epoch": 0.9233807725820505,
+      "grad_norm": 0.5238228440284729,
+      "learning_rate": 0.0001,
+      "loss": 1.7041,
+      "step": 7948
+    },
+    {
+      "epoch": 0.923496950334011,
+      "grad_norm": 0.4958471655845642,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 7949
+    },
+    {
+      "epoch": 0.9236131280859715,
+      "grad_norm": 0.4387954771518707,
+      "learning_rate": 0.0001,
+      "loss": 1.3532,
+      "step": 7950
+    },
+    {
+      "epoch": 0.923729305837932,
+      "grad_norm": 0.4402277171611786,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 7951
+    },
+    {
+      "epoch": 0.9238454835898925,
+      "grad_norm": 0.5100318193435669,
+      "learning_rate": 0.0001,
+      "loss": 1.5843,
+      "step": 7952
+    },
+    {
+      "epoch": 0.923961661341853,
+      "grad_norm": 0.4846263527870178,
+      "learning_rate": 0.0001,
+      "loss": 1.6603,
+      "step": 7953
+    },
+    {
+      "epoch": 0.9240778390938136,
+      "grad_norm": 0.4743453562259674,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 7954
+    },
+    {
+      "epoch": 0.9241940168457741,
+      "grad_norm": 0.5055150985717773,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 7955
+    },
+    {
+      "epoch": 0.9243101945977346,
+      "grad_norm": 0.46084272861480713,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 7956
+    },
+    {
+      "epoch": 0.924426372349695,
+      "grad_norm": 0.47752270102500916,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 7957
+    },
+    {
+      "epoch": 0.9245425501016555,
+      "grad_norm": 0.47318798303604126,
+      "learning_rate": 0.0001,
+      "loss": 1.6307,
+      "step": 7958
+    },
+    {
+      "epoch": 0.924658727853616,
+      "grad_norm": 0.4867374897003174,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 7959
+    },
+    {
+      "epoch": 0.9247749056055765,
+      "grad_norm": 0.48629096150398254,
+      "learning_rate": 0.0001,
+      "loss": 1.7713,
+      "step": 7960
+    },
+    {
+      "epoch": 0.924891083357537,
+      "grad_norm": 0.47836753726005554,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 7961
+    },
+    {
+      "epoch": 0.9250072611094975,
+      "grad_norm": 0.49589991569519043,
+      "learning_rate": 0.0001,
+      "loss": 1.6846,
+      "step": 7962
+    },
+    {
+      "epoch": 0.925123438861458,
+      "grad_norm": 0.46878737211227417,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 7963
+    },
+    {
+      "epoch": 0.9252396166134186,
+      "grad_norm": 0.4647105634212494,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 7964
+    },
+    {
+      "epoch": 0.9253557943653791,
+      "grad_norm": 0.46214747428894043,
+      "learning_rate": 0.0001,
+      "loss": 1.5577,
+      "step": 7965
+    },
+    {
+      "epoch": 0.9254719721173396,
+      "grad_norm": 0.5062007904052734,
+      "learning_rate": 0.0001,
+      "loss": 1.8203,
+      "step": 7966
+    },
+    {
+      "epoch": 0.9255881498693,
+      "grad_norm": 0.5121557116508484,
+      "learning_rate": 0.0001,
+      "loss": 1.7192,
+      "step": 7967
+    },
+    {
+      "epoch": 0.9257043276212605,
+      "grad_norm": 0.44176679849624634,
+      "learning_rate": 0.0001,
+      "loss": 1.4522,
+      "step": 7968
+    },
+    {
+      "epoch": 0.925820505373221,
+      "grad_norm": 0.47141146659851074,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 7969
+    },
+    {
+      "epoch": 0.9259366831251815,
+      "grad_norm": 0.48892006278038025,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 7970
+    },
+    {
+      "epoch": 0.926052860877142,
+      "grad_norm": 0.512973964214325,
+      "learning_rate": 0.0001,
+      "loss": 1.7151,
+      "step": 7971
+    },
+    {
+      "epoch": 0.9261690386291025,
+      "grad_norm": 0.4414759576320648,
+      "learning_rate": 0.0001,
+      "loss": 1.3828,
+      "step": 7972
+    },
+    {
+      "epoch": 0.926285216381063,
+      "grad_norm": 0.48887619376182556,
+      "learning_rate": 0.0001,
+      "loss": 1.6999,
+      "step": 7973
+    },
+    {
+      "epoch": 0.9264013941330236,
+      "grad_norm": 0.45762237906455994,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 7974
+    },
+    {
+      "epoch": 0.9265175718849841,
+      "grad_norm": 0.45847946405410767,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 7975
+    },
+    {
+      "epoch": 0.9266337496369446,
+      "grad_norm": 0.49288585782051086,
+      "learning_rate": 0.0001,
+      "loss": 1.762,
+      "step": 7976
+    },
+    {
+      "epoch": 0.926749927388905,
+      "grad_norm": 0.48561185598373413,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 7977
+    },
+    {
+      "epoch": 0.9268661051408655,
+      "grad_norm": 0.447745680809021,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 7978
+    },
+    {
+      "epoch": 0.926982282892826,
+      "grad_norm": 0.49611517786979675,
+      "learning_rate": 0.0001,
+      "loss": 1.6865,
+      "step": 7979
+    },
+    {
+      "epoch": 0.9270984606447865,
+      "grad_norm": 0.46112751960754395,
+      "learning_rate": 0.0001,
+      "loss": 1.6159,
+      "step": 7980
+    },
+    {
+      "epoch": 0.927214638396747,
+      "grad_norm": 0.4908156991004944,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 7981
+    },
+    {
+      "epoch": 0.9273308161487075,
+      "grad_norm": 0.4530208706855774,
+      "learning_rate": 0.0001,
+      "loss": 1.5535,
+      "step": 7982
+    },
+    {
+      "epoch": 0.927446993900668,
+      "grad_norm": 0.49181050062179565,
+      "learning_rate": 0.0001,
+      "loss": 1.7461,
+      "step": 7983
+    },
+    {
+      "epoch": 0.9275631716526285,
+      "grad_norm": 0.49187758564949036,
+      "learning_rate": 0.0001,
+      "loss": 1.7731,
+      "step": 7984
+    },
+    {
+      "epoch": 0.9276793494045891,
+      "grad_norm": 0.5021129846572876,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 7985
+    },
+    {
+      "epoch": 0.9277955271565496,
+      "grad_norm": 0.508470892906189,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 7986
+    },
+    {
+      "epoch": 0.92791170490851,
+      "grad_norm": 0.5105016231536865,
+      "learning_rate": 0.0001,
+      "loss": 1.7802,
+      "step": 7987
+    },
+    {
+      "epoch": 0.9280278826604705,
+      "grad_norm": 0.4812169671058655,
+      "learning_rate": 0.0001,
+      "loss": 1.6524,
+      "step": 7988
+    },
+    {
+      "epoch": 0.928144060412431,
+      "grad_norm": 0.4714551568031311,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 7989
+    },
+    {
+      "epoch": 0.9282602381643915,
+      "grad_norm": 0.4296482801437378,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 7990
+    },
+    {
+      "epoch": 0.928376415916352,
+      "grad_norm": 0.45855334401130676,
+      "learning_rate": 0.0001,
+      "loss": 1.574,
+      "step": 7991
+    },
+    {
+      "epoch": 0.9284925936683125,
+      "grad_norm": 0.4766031801700592,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 7992
+    },
+    {
+      "epoch": 0.928608771420273,
+      "grad_norm": 0.45450130105018616,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 7993
+    },
+    {
+      "epoch": 0.9287249491722335,
+      "grad_norm": 0.512444794178009,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 7994
+    },
+    {
+      "epoch": 0.9288411269241941,
+      "grad_norm": 0.49401575326919556,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 7995
+    },
+    {
+      "epoch": 0.9289573046761546,
+      "grad_norm": 0.5121031999588013,
+      "learning_rate": 0.0001,
+      "loss": 1.6722,
+      "step": 7996
+    },
+    {
+      "epoch": 0.929073482428115,
+      "grad_norm": 0.4602169692516327,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 7997
+    },
+    {
+      "epoch": 0.9291896601800755,
+      "grad_norm": 0.49888116121292114,
+      "learning_rate": 0.0001,
+      "loss": 1.728,
+      "step": 7998
+    },
+    {
+      "epoch": 0.929305837932036,
+      "grad_norm": 0.5206019878387451,
+      "learning_rate": 0.0001,
+      "loss": 1.6714,
+      "step": 7999
+    },
+    {
+      "epoch": 0.9294220156839965,
+      "grad_norm": 0.4638976454734802,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 8000
+    },
+    {
+      "epoch": 0.929538193435957,
+      "grad_norm": 0.48082998394966125,
+      "learning_rate": 0.0001,
+      "loss": 1.8074,
+      "step": 8001
+    },
+    {
+      "epoch": 0.9296543711879175,
+      "grad_norm": 0.4941411018371582,
+      "learning_rate": 0.0001,
+      "loss": 1.8171,
+      "step": 8002
+    },
+    {
+      "epoch": 0.929770548939878,
+      "grad_norm": 0.45711395144462585,
+      "learning_rate": 0.0001,
+      "loss": 1.6415,
+      "step": 8003
+    },
+    {
+      "epoch": 0.9298867266918385,
+      "grad_norm": 0.4651217758655548,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 8004
+    },
+    {
+      "epoch": 0.930002904443799,
+      "grad_norm": 0.45442432165145874,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 8005
+    },
+    {
+      "epoch": 0.9301190821957596,
+      "grad_norm": 0.4956342577934265,
+      "learning_rate": 0.0001,
+      "loss": 1.4252,
+      "step": 8006
+    },
+    {
+      "epoch": 0.93023525994772,
+      "grad_norm": 0.5071079730987549,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 8007
+    },
+    {
+      "epoch": 0.9303514376996805,
+      "grad_norm": 0.4929546117782593,
+      "learning_rate": 0.0001,
+      "loss": 1.6939,
+      "step": 8008
+    },
+    {
+      "epoch": 0.930467615451641,
+      "grad_norm": 0.5169224739074707,
+      "learning_rate": 0.0001,
+      "loss": 1.6255,
+      "step": 8009
+    },
+    {
+      "epoch": 0.9305837932036015,
+      "grad_norm": 0.5088503956794739,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 8010
+    },
+    {
+      "epoch": 0.930699970955562,
+      "grad_norm": 0.48332953453063965,
+      "learning_rate": 0.0001,
+      "loss": 1.712,
+      "step": 8011
+    },
+    {
+      "epoch": 0.9308161487075225,
+      "grad_norm": 0.49917006492614746,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 8012
+    },
+    {
+      "epoch": 0.930932326459483,
+      "grad_norm": 0.4794350564479828,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 8013
+    },
+    {
+      "epoch": 0.9310485042114435,
+      "grad_norm": 0.4891558885574341,
+      "learning_rate": 0.0001,
+      "loss": 1.6764,
+      "step": 8014
+    },
+    {
+      "epoch": 0.931164681963404,
+      "grad_norm": 0.4725422263145447,
+      "learning_rate": 0.0001,
+      "loss": 1.6397,
+      "step": 8015
+    },
+    {
+      "epoch": 0.9312808597153646,
+      "grad_norm": 0.48339328169822693,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 8016
+    },
+    {
+      "epoch": 0.931397037467325,
+      "grad_norm": 0.4951247572898865,
+      "learning_rate": 0.0001,
+      "loss": 1.6353,
+      "step": 8017
+    },
+    {
+      "epoch": 0.9315132152192855,
+      "grad_norm": 0.46006613969802856,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 8018
+    },
+    {
+      "epoch": 0.931629392971246,
+      "grad_norm": 0.4528697431087494,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 8019
+    },
+    {
+      "epoch": 0.9317455707232065,
+      "grad_norm": 0.46240419149398804,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 8020
+    },
+    {
+      "epoch": 0.931861748475167,
+      "grad_norm": 0.4489692449569702,
+      "learning_rate": 0.0001,
+      "loss": 1.5491,
+      "step": 8021
+    },
+    {
+      "epoch": 0.9319779262271275,
+      "grad_norm": 0.473308265209198,
+      "learning_rate": 0.0001,
+      "loss": 1.6718,
+      "step": 8022
+    },
+    {
+      "epoch": 0.932094103979088,
+      "grad_norm": 0.46006840467453003,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 8023
+    },
+    {
+      "epoch": 0.9322102817310485,
+      "grad_norm": 0.501011848449707,
+      "learning_rate": 0.0001,
+      "loss": 1.645,
+      "step": 8024
+    },
+    {
+      "epoch": 0.932326459483009,
+      "grad_norm": 0.46970438957214355,
+      "learning_rate": 0.0001,
+      "loss": 1.7605,
+      "step": 8025
+    },
+    {
+      "epoch": 0.9324426372349695,
+      "grad_norm": 0.45778888463974,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 8026
+    },
+    {
+      "epoch": 0.93255881498693,
+      "grad_norm": 0.46483826637268066,
+      "learning_rate": 0.0001,
+      "loss": 1.6317,
+      "step": 8027
+    },
+    {
+      "epoch": 0.9326749927388905,
+      "grad_norm": 0.4459386169910431,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 8028
+    },
+    {
+      "epoch": 0.932791170490851,
+      "grad_norm": 0.5162267088890076,
+      "learning_rate": 0.0001,
+      "loss": 1.7774,
+      "step": 8029
+    },
+    {
+      "epoch": 0.9329073482428115,
+      "grad_norm": 0.46646633744239807,
+      "learning_rate": 0.0001,
+      "loss": 1.443,
+      "step": 8030
+    },
+    {
+      "epoch": 0.933023525994772,
+      "grad_norm": 0.5139045119285583,
+      "learning_rate": 0.0001,
+      "loss": 1.7225,
+      "step": 8031
+    },
+    {
+      "epoch": 0.9331397037467325,
+      "grad_norm": 0.44446662068367004,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 8032
+    },
+    {
+      "epoch": 0.933255881498693,
+      "grad_norm": 0.46956273913383484,
+      "learning_rate": 0.0001,
+      "loss": 1.5063,
+      "step": 8033
+    },
+    {
+      "epoch": 0.9333720592506535,
+      "grad_norm": 0.4521263837814331,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 8034
+    },
+    {
+      "epoch": 0.933488237002614,
+      "grad_norm": 0.5222106575965881,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 8035
+    },
+    {
+      "epoch": 0.9336044147545745,
+      "grad_norm": 0.44853895902633667,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 8036
+    },
+    {
+      "epoch": 0.933720592506535,
+      "grad_norm": 0.48177504539489746,
+      "learning_rate": 0.0001,
+      "loss": 1.5831,
+      "step": 8037
+    },
+    {
+      "epoch": 0.9338367702584955,
+      "grad_norm": 0.4503477215766907,
+      "learning_rate": 0.0001,
+      "loss": 1.3327,
+      "step": 8038
+    },
+    {
+      "epoch": 0.933952948010456,
+      "grad_norm": 0.507217288017273,
+      "learning_rate": 0.0001,
+      "loss": 1.7489,
+      "step": 8039
+    },
+    {
+      "epoch": 0.9340691257624165,
+      "grad_norm": 0.4814956486225128,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 8040
+    },
+    {
+      "epoch": 0.934185303514377,
+      "grad_norm": 0.4742831885814667,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 8041
+    },
+    {
+      "epoch": 0.9343014812663375,
+      "grad_norm": 0.4570356607437134,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 8042
+    },
+    {
+      "epoch": 0.934417659018298,
+      "grad_norm": 0.46834975481033325,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 8043
+    },
+    {
+      "epoch": 0.9345338367702585,
+      "grad_norm": 0.5158945322036743,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 8044
+    },
+    {
+      "epoch": 0.934650014522219,
+      "grad_norm": 0.5279756188392639,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 8045
+    },
+    {
+      "epoch": 0.9347661922741795,
+      "grad_norm": 0.47560861706733704,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 8046
+    },
+    {
+      "epoch": 0.9348823700261399,
+      "grad_norm": 0.4888511300086975,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 8047
+    },
+    {
+      "epoch": 0.9349985477781005,
+      "grad_norm": 0.4677184820175171,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 8048
+    },
+    {
+      "epoch": 0.935114725530061,
+      "grad_norm": 0.4719541370868683,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 8049
+    },
+    {
+      "epoch": 0.9352309032820215,
+      "grad_norm": 0.477180540561676,
+      "learning_rate": 0.0001,
+      "loss": 1.6941,
+      "step": 8050
+    },
+    {
+      "epoch": 0.935347081033982,
+      "grad_norm": 0.46396854519844055,
+      "learning_rate": 0.0001,
+      "loss": 1.3669,
+      "step": 8051
+    },
+    {
+      "epoch": 0.9354632587859425,
+      "grad_norm": 0.4677002727985382,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 8052
+    },
+    {
+      "epoch": 0.935579436537903,
+      "grad_norm": 0.45009395480155945,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 8053
+    },
+    {
+      "epoch": 0.9356956142898635,
+      "grad_norm": 0.47856685519218445,
+      "learning_rate": 0.0001,
+      "loss": 1.613,
+      "step": 8054
+    },
+    {
+      "epoch": 0.935811792041824,
+      "grad_norm": 0.5348851084709167,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 8055
+    },
+    {
+      "epoch": 0.9359279697937845,
+      "grad_norm": 0.4984420835971832,
+      "learning_rate": 0.0001,
+      "loss": 1.6902,
+      "step": 8056
+    },
+    {
+      "epoch": 0.9360441475457449,
+      "grad_norm": 0.5199939608573914,
+      "learning_rate": 0.0001,
+      "loss": 1.6925,
+      "step": 8057
+    },
+    {
+      "epoch": 0.9361603252977055,
+      "grad_norm": 0.5202213525772095,
+      "learning_rate": 0.0001,
+      "loss": 1.6618,
+      "step": 8058
+    },
+    {
+      "epoch": 0.936276503049666,
+      "grad_norm": 0.5361291766166687,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 8059
+    },
+    {
+      "epoch": 0.9363926808016265,
+      "grad_norm": 0.49866026639938354,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 8060
+    },
+    {
+      "epoch": 0.936508858553587,
+      "grad_norm": 0.49306201934814453,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 8061
+    },
+    {
+      "epoch": 0.9366250363055475,
+      "grad_norm": 0.4813399910926819,
+      "learning_rate": 0.0001,
+      "loss": 1.6272,
+      "step": 8062
+    },
+    {
+      "epoch": 0.936741214057508,
+      "grad_norm": 0.47636452317237854,
+      "learning_rate": 0.0001,
+      "loss": 1.6794,
+      "step": 8063
+    },
+    {
+      "epoch": 0.9368573918094685,
+      "grad_norm": 0.4818163812160492,
+      "learning_rate": 0.0001,
+      "loss": 1.4483,
+      "step": 8064
+    },
+    {
+      "epoch": 0.936973569561429,
+      "grad_norm": 0.46799615025520325,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 8065
+    },
+    {
+      "epoch": 0.9370897473133895,
+      "grad_norm": 0.48917245864868164,
+      "learning_rate": 0.0001,
+      "loss": 1.6402,
+      "step": 8066
+    },
+    {
+      "epoch": 0.9372059250653499,
+      "grad_norm": 0.47310730814933777,
+      "learning_rate": 0.0001,
+      "loss": 1.6164,
+      "step": 8067
+    },
+    {
+      "epoch": 0.9373221028173104,
+      "grad_norm": 0.5221018195152283,
+      "learning_rate": 0.0001,
+      "loss": 1.8971,
+      "step": 8068
+    },
+    {
+      "epoch": 0.937438280569271,
+      "grad_norm": 0.48749321699142456,
+      "learning_rate": 0.0001,
+      "loss": 1.5722,
+      "step": 8069
+    },
+    {
+      "epoch": 0.9375544583212315,
+      "grad_norm": 0.4679155945777893,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 8070
+    },
+    {
+      "epoch": 0.937670636073192,
+      "grad_norm": 0.47157520055770874,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 8071
+    },
+    {
+      "epoch": 0.9377868138251525,
+      "grad_norm": 0.4769766628742218,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 8072
+    },
+    {
+      "epoch": 0.937902991577113,
+      "grad_norm": 0.47961485385894775,
+      "learning_rate": 0.0001,
+      "loss": 1.7161,
+      "step": 8073
+    },
+    {
+      "epoch": 0.9380191693290735,
+      "grad_norm": 0.4806070327758789,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 8074
+    },
+    {
+      "epoch": 0.938135347081034,
+      "grad_norm": 0.4962315261363983,
+      "learning_rate": 0.0001,
+      "loss": 1.658,
+      "step": 8075
+    },
+    {
+      "epoch": 0.9382515248329945,
+      "grad_norm": 0.5138833522796631,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 8076
+    },
+    {
+      "epoch": 0.9383677025849549,
+      "grad_norm": 0.4957931935787201,
+      "learning_rate": 0.0001,
+      "loss": 1.6425,
+      "step": 8077
+    },
+    {
+      "epoch": 0.9384838803369154,
+      "grad_norm": 0.49878206849098206,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 8078
+    },
+    {
+      "epoch": 0.938600058088876,
+      "grad_norm": 0.49658435583114624,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 8079
+    },
+    {
+      "epoch": 0.9387162358408365,
+      "grad_norm": 0.508560836315155,
+      "learning_rate": 0.0001,
+      "loss": 1.5825,
+      "step": 8080
+    },
+    {
+      "epoch": 0.938832413592797,
+      "grad_norm": 0.4853934943675995,
+      "learning_rate": 0.0001,
+      "loss": 1.5385,
+      "step": 8081
+    },
+    {
+      "epoch": 0.9389485913447575,
+      "grad_norm": 0.478221595287323,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 8082
+    },
+    {
+      "epoch": 0.939064769096718,
+      "grad_norm": 0.4989098012447357,
+      "learning_rate": 0.0001,
+      "loss": 1.647,
+      "step": 8083
+    },
+    {
+      "epoch": 0.9391809468486785,
+      "grad_norm": 0.4413130283355713,
+      "learning_rate": 0.0001,
+      "loss": 1.4044,
+      "step": 8084
+    },
+    {
+      "epoch": 0.939297124600639,
+      "grad_norm": 0.448485791683197,
+      "learning_rate": 0.0001,
+      "loss": 1.6516,
+      "step": 8085
+    },
+    {
+      "epoch": 0.9394133023525995,
+      "grad_norm": 0.4658021330833435,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 8086
+    },
+    {
+      "epoch": 0.9395294801045599,
+      "grad_norm": 0.48055848479270935,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 8087
+    },
+    {
+      "epoch": 0.9396456578565204,
+      "grad_norm": 0.47321316599845886,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 8088
+    },
+    {
+      "epoch": 0.9397618356084809,
+      "grad_norm": 0.44191011786460876,
+      "learning_rate": 0.0001,
+      "loss": 1.3483,
+      "step": 8089
+    },
+    {
+      "epoch": 0.9398780133604415,
+      "grad_norm": 0.5220785140991211,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 8090
+    },
+    {
+      "epoch": 0.939994191112402,
+      "grad_norm": 0.4720672070980072,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 8091
+    },
+    {
+      "epoch": 0.9401103688643625,
+      "grad_norm": 0.46814772486686707,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 8092
+    },
+    {
+      "epoch": 0.940226546616323,
+      "grad_norm": 0.5168691873550415,
+      "learning_rate": 0.0001,
+      "loss": 1.8196,
+      "step": 8093
+    },
+    {
+      "epoch": 0.9403427243682835,
+      "grad_norm": 0.4674499034881592,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 8094
+    },
+    {
+      "epoch": 0.940458902120244,
+      "grad_norm": 0.49029046297073364,
+      "learning_rate": 0.0001,
+      "loss": 1.7264,
+      "step": 8095
+    },
+    {
+      "epoch": 0.9405750798722045,
+      "grad_norm": 0.5169382691383362,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 8096
+    },
+    {
+      "epoch": 0.940691257624165,
+      "grad_norm": 0.48217153549194336,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 8097
+    },
+    {
+      "epoch": 0.9408074353761254,
+      "grad_norm": 0.49331241846084595,
+      "learning_rate": 0.0001,
+      "loss": 1.5199,
+      "step": 8098
+    },
+    {
+      "epoch": 0.9409236131280859,
+      "grad_norm": 0.46810704469680786,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 8099
+    },
+    {
+      "epoch": 0.9410397908800465,
+      "grad_norm": 0.5021561980247498,
+      "learning_rate": 0.0001,
+      "loss": 1.7799,
+      "step": 8100
+    },
+    {
+      "epoch": 0.941155968632007,
+      "grad_norm": 0.44212183356285095,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 8101
+    },
+    {
+      "epoch": 0.9412721463839675,
+      "grad_norm": 0.4898991882801056,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 8102
+    },
+    {
+      "epoch": 0.941388324135928,
+      "grad_norm": 0.45801275968551636,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 8103
+    },
+    {
+      "epoch": 0.9415045018878885,
+      "grad_norm": 0.45079344511032104,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 8104
+    },
+    {
+      "epoch": 0.941620679639849,
+      "grad_norm": 0.4727324843406677,
+      "learning_rate": 0.0001,
+      "loss": 1.6742,
+      "step": 8105
+    },
+    {
+      "epoch": 0.9417368573918095,
+      "grad_norm": 0.4644958972930908,
+      "learning_rate": 0.0001,
+      "loss": 1.4713,
+      "step": 8106
+    },
+    {
+      "epoch": 0.94185303514377,
+      "grad_norm": 0.5391181707382202,
+      "learning_rate": 0.0001,
+      "loss": 1.7521,
+      "step": 8107
+    },
+    {
+      "epoch": 0.9419692128957304,
+      "grad_norm": 0.5218836665153503,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 8108
+    },
+    {
+      "epoch": 0.9420853906476909,
+      "grad_norm": 0.486215740442276,
+      "learning_rate": 0.0001,
+      "loss": 1.6442,
+      "step": 8109
+    },
+    {
+      "epoch": 0.9422015683996514,
+      "grad_norm": 0.4889783263206482,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 8110
+    },
+    {
+      "epoch": 0.942317746151612,
+      "grad_norm": 0.44267192482948303,
+      "learning_rate": 0.0001,
+      "loss": 1.4231,
+      "step": 8111
+    },
+    {
+      "epoch": 0.9424339239035725,
+      "grad_norm": 0.49176719784736633,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 8112
+    },
+    {
+      "epoch": 0.942550101655533,
+      "grad_norm": 0.5220740437507629,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 8113
+    },
+    {
+      "epoch": 0.9426662794074935,
+      "grad_norm": 0.47036683559417725,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 8114
+    },
+    {
+      "epoch": 0.942782457159454,
+      "grad_norm": 0.4946513772010803,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 8115
+    },
+    {
+      "epoch": 0.9428986349114145,
+      "grad_norm": 0.48926377296447754,
+      "learning_rate": 0.0001,
+      "loss": 1.6243,
+      "step": 8116
+    },
+    {
+      "epoch": 0.943014812663375,
+      "grad_norm": 0.46401751041412354,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 8117
+    },
+    {
+      "epoch": 0.9431309904153354,
+      "grad_norm": 0.4936821162700653,
+      "learning_rate": 0.0001,
+      "loss": 1.5866,
+      "step": 8118
+    },
+    {
+      "epoch": 0.9432471681672959,
+      "grad_norm": 0.5162927508354187,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 8119
+    },
+    {
+      "epoch": 0.9433633459192564,
+      "grad_norm": 0.49203792214393616,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 8120
+    },
+    {
+      "epoch": 0.943479523671217,
+      "grad_norm": 0.48828327655792236,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 8121
+    },
+    {
+      "epoch": 0.9435957014231775,
+      "grad_norm": 0.4814288914203644,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 8122
+    },
+    {
+      "epoch": 0.943711879175138,
+      "grad_norm": 0.4744342267513275,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 8123
+    },
+    {
+      "epoch": 0.9438280569270985,
+      "grad_norm": 0.4669947922229767,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 8124
+    },
+    {
+      "epoch": 0.943944234679059,
+      "grad_norm": 0.4419528841972351,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 8125
+    },
+    {
+      "epoch": 0.9440604124310195,
+      "grad_norm": 0.47628259658813477,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 8126
+    },
+    {
+      "epoch": 0.94417659018298,
+      "grad_norm": 0.4776514768600464,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 8127
+    },
+    {
+      "epoch": 0.9442927679349404,
+      "grad_norm": 0.4446832537651062,
+      "learning_rate": 0.0001,
+      "loss": 1.5658,
+      "step": 8128
+    },
+    {
+      "epoch": 0.9444089456869009,
+      "grad_norm": 0.5115212202072144,
+      "learning_rate": 0.0001,
+      "loss": 1.6909,
+      "step": 8129
+    },
+    {
+      "epoch": 0.9445251234388614,
+      "grad_norm": 0.47382161021232605,
+      "learning_rate": 0.0001,
+      "loss": 1.409,
+      "step": 8130
+    },
+    {
+      "epoch": 0.9446413011908219,
+      "grad_norm": 0.4738979637622833,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 8131
+    },
+    {
+      "epoch": 0.9447574789427825,
+      "grad_norm": 0.4864945113658905,
+      "learning_rate": 0.0001,
+      "loss": 1.6893,
+      "step": 8132
+    },
+    {
+      "epoch": 0.944873656694743,
+      "grad_norm": 0.5098589658737183,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 8133
+    },
+    {
+      "epoch": 0.9449898344467035,
+      "grad_norm": 0.4890642464160919,
+      "learning_rate": 0.0001,
+      "loss": 1.7127,
+      "step": 8134
+    },
+    {
+      "epoch": 0.945106012198664,
+      "grad_norm": 0.499066025018692,
+      "learning_rate": 0.0001,
+      "loss": 1.722,
+      "step": 8135
+    },
+    {
+      "epoch": 0.9452221899506245,
+      "grad_norm": 0.4744119942188263,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 8136
+    },
+    {
+      "epoch": 0.945338367702585,
+      "grad_norm": 0.45313483476638794,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 8137
+    },
+    {
+      "epoch": 0.9454545454545454,
+      "grad_norm": 0.4957541227340698,
+      "learning_rate": 0.0001,
+      "loss": 1.7737,
+      "step": 8138
+    },
+    {
+      "epoch": 0.9455707232065059,
+      "grad_norm": 0.46510589122772217,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 8139
+    },
+    {
+      "epoch": 0.9456869009584664,
+      "grad_norm": 0.49345633387565613,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 8140
+    },
+    {
+      "epoch": 0.9458030787104269,
+      "grad_norm": 0.44916290044784546,
+      "learning_rate": 0.0001,
+      "loss": 1.5718,
+      "step": 8141
+    },
+    {
+      "epoch": 0.9459192564623875,
+      "grad_norm": 0.5268651247024536,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 8142
+    },
+    {
+      "epoch": 0.946035434214348,
+      "grad_norm": 0.4611073136329651,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 8143
+    },
+    {
+      "epoch": 0.9461516119663085,
+      "grad_norm": 0.4878610372543335,
+      "learning_rate": 0.0001,
+      "loss": 1.6375,
+      "step": 8144
+    },
+    {
+      "epoch": 0.946267789718269,
+      "grad_norm": 0.48265448212623596,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 8145
+    },
+    {
+      "epoch": 0.9463839674702295,
+      "grad_norm": 0.4883013963699341,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 8146
+    },
+    {
+      "epoch": 0.94650014522219,
+      "grad_norm": 0.4902404844760895,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 8147
+    },
+    {
+      "epoch": 0.9466163229741504,
+      "grad_norm": 0.48211053013801575,
+      "learning_rate": 0.0001,
+      "loss": 1.4926,
+      "step": 8148
+    },
+    {
+      "epoch": 0.9467325007261109,
+      "grad_norm": 0.4793367087841034,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 8149
+    },
+    {
+      "epoch": 0.9468486784780714,
+      "grad_norm": 0.47337889671325684,
+      "learning_rate": 0.0001,
+      "loss": 1.6667,
+      "step": 8150
+    },
+    {
+      "epoch": 0.9469648562300319,
+      "grad_norm": 0.45876795053482056,
+      "learning_rate": 0.0001,
+      "loss": 1.4361,
+      "step": 8151
+    },
+    {
+      "epoch": 0.9470810339819925,
+      "grad_norm": 0.5049482583999634,
+      "learning_rate": 0.0001,
+      "loss": 1.7341,
+      "step": 8152
+    },
+    {
+      "epoch": 0.947197211733953,
+      "grad_norm": 0.4585045874118805,
+      "learning_rate": 0.0001,
+      "loss": 1.6708,
+      "step": 8153
+    },
+    {
+      "epoch": 0.9473133894859135,
+      "grad_norm": 0.48491230607032776,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 8154
+    },
+    {
+      "epoch": 0.947429567237874,
+      "grad_norm": 0.48654916882514954,
+      "learning_rate": 0.0001,
+      "loss": 1.7573,
+      "step": 8155
+    },
+    {
+      "epoch": 0.9475457449898345,
+      "grad_norm": 0.4684811234474182,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 8156
+    },
+    {
+      "epoch": 0.947661922741795,
+      "grad_norm": 0.4712132513523102,
+      "learning_rate": 0.0001,
+      "loss": 1.6316,
+      "step": 8157
+    },
+    {
+      "epoch": 0.9477781004937554,
+      "grad_norm": 0.4901881217956543,
+      "learning_rate": 0.0001,
+      "loss": 1.8043,
+      "step": 8158
+    },
+    {
+      "epoch": 0.9478942782457159,
+      "grad_norm": 0.4755280017852783,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 8159
+    },
+    {
+      "epoch": 0.9480104559976764,
+      "grad_norm": 0.5146501064300537,
+      "learning_rate": 0.0001,
+      "loss": 1.5001,
+      "step": 8160
+    },
+    {
+      "epoch": 0.9481266337496369,
+      "grad_norm": 0.4786891043186188,
+      "learning_rate": 0.0001,
+      "loss": 1.6811,
+      "step": 8161
+    },
+    {
+      "epoch": 0.9482428115015974,
+      "grad_norm": 0.44507408142089844,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 8162
+    },
+    {
+      "epoch": 0.948358989253558,
+      "grad_norm": 0.4719574749469757,
+      "learning_rate": 0.0001,
+      "loss": 1.6408,
+      "step": 8163
+    },
+    {
+      "epoch": 0.9484751670055185,
+      "grad_norm": 0.48670482635498047,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 8164
+    },
+    {
+      "epoch": 0.948591344757479,
+      "grad_norm": 0.4823938012123108,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 8165
+    },
+    {
+      "epoch": 0.9487075225094395,
+      "grad_norm": 0.4881124794483185,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 8166
+    },
+    {
+      "epoch": 0.9488237002614,
+      "grad_norm": 0.4872508943080902,
+      "learning_rate": 0.0001,
+      "loss": 1.565,
+      "step": 8167
+    },
+    {
+      "epoch": 0.9489398780133604,
+      "grad_norm": 0.5031945109367371,
+      "learning_rate": 0.0001,
+      "loss": 1.6907,
+      "step": 8168
+    },
+    {
+      "epoch": 0.9490560557653209,
+      "grad_norm": 0.5044823884963989,
+      "learning_rate": 0.0001,
+      "loss": 1.6954,
+      "step": 8169
+    },
+    {
+      "epoch": 0.9491722335172814,
+      "grad_norm": 0.48246756196022034,
+      "learning_rate": 0.0001,
+      "loss": 1.6214,
+      "step": 8170
+    },
+    {
+      "epoch": 0.9492884112692419,
+      "grad_norm": 0.4963812232017517,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 8171
+    },
+    {
+      "epoch": 0.9494045890212024,
+      "grad_norm": 0.4816388785839081,
+      "learning_rate": 0.0001,
+      "loss": 1.5913,
+      "step": 8172
+    },
+    {
+      "epoch": 0.949520766773163,
+      "grad_norm": 0.4748592674732208,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 8173
+    },
+    {
+      "epoch": 0.9496369445251235,
+      "grad_norm": 0.46375009417533875,
+      "learning_rate": 0.0001,
+      "loss": 1.4255,
+      "step": 8174
+    },
+    {
+      "epoch": 0.949753122277084,
+      "grad_norm": 0.45289376378059387,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 8175
+    },
+    {
+      "epoch": 0.9498693000290445,
+      "grad_norm": 0.48634836077690125,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 8176
+    },
+    {
+      "epoch": 0.949985477781005,
+      "grad_norm": 0.49181076884269714,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 8177
+    },
+    {
+      "epoch": 0.9501016555329654,
+      "grad_norm": 0.45793280005455017,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 8178
+    },
+    {
+      "epoch": 0.9502178332849259,
+      "grad_norm": 0.4974653720855713,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 8179
+    },
+    {
+      "epoch": 0.9503340110368864,
+      "grad_norm": 0.4716089367866516,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 8180
+    },
+    {
+      "epoch": 0.9504501887888469,
+      "grad_norm": 0.45210596919059753,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 8181
+    },
+    {
+      "epoch": 0.9505663665408074,
+      "grad_norm": 0.5031945705413818,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 8182
+    },
+    {
+      "epoch": 0.9506825442927679,
+      "grad_norm": 0.46104729175567627,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 8183
+    },
+    {
+      "epoch": 0.9507987220447285,
+      "grad_norm": 0.45279985666275024,
+      "learning_rate": 0.0001,
+      "loss": 1.6108,
+      "step": 8184
+    },
+    {
+      "epoch": 0.950914899796689,
+      "grad_norm": 0.618247389793396,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 8185
+    },
+    {
+      "epoch": 0.9510310775486495,
+      "grad_norm": 0.47192826867103577,
+      "learning_rate": 0.0001,
+      "loss": 1.4801,
+      "step": 8186
+    },
+    {
+      "epoch": 0.95114725530061,
+      "grad_norm": 0.46734699606895447,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 8187
+    },
+    {
+      "epoch": 0.9512634330525704,
+      "grad_norm": 0.46649816632270813,
+      "learning_rate": 0.0001,
+      "loss": 1.3187,
+      "step": 8188
+    },
+    {
+      "epoch": 0.9513796108045309,
+      "grad_norm": 0.5192746520042419,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 8189
+    },
+    {
+      "epoch": 0.9514957885564914,
+      "grad_norm": 0.476370632648468,
+      "learning_rate": 0.0001,
+      "loss": 1.568,
+      "step": 8190
+    },
+    {
+      "epoch": 0.9516119663084519,
+      "grad_norm": 0.49778425693511963,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 8191
+    },
+    {
+      "epoch": 0.9517281440604124,
+      "grad_norm": 0.4732555150985718,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 8192
+    },
+    {
+      "epoch": 0.9518443218123729,
+      "grad_norm": 0.47929319739341736,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 8193
+    },
+    {
+      "epoch": 0.9519604995643335,
+      "grad_norm": 0.44977056980133057,
+      "learning_rate": 0.0001,
+      "loss": 1.4695,
+      "step": 8194
+    },
+    {
+      "epoch": 0.952076677316294,
+      "grad_norm": 0.5259702205657959,
+      "learning_rate": 0.0001,
+      "loss": 1.7958,
+      "step": 8195
+    },
+    {
+      "epoch": 0.9521928550682545,
+      "grad_norm": 0.4790816307067871,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 8196
+    },
+    {
+      "epoch": 0.952309032820215,
+      "grad_norm": 0.48329922556877136,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 8197
+    },
+    {
+      "epoch": 0.9524252105721754,
+      "grad_norm": 0.451102077960968,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 8198
+    },
+    {
+      "epoch": 0.9525413883241359,
+      "grad_norm": 0.4747765064239502,
+      "learning_rate": 0.0001,
+      "loss": 1.6745,
+      "step": 8199
+    },
+    {
+      "epoch": 0.9526575660760964,
+      "grad_norm": 0.5516743063926697,
+      "learning_rate": 0.0001,
+      "loss": 1.7372,
+      "step": 8200
+    },
+    {
+      "epoch": 0.9527737438280569,
+      "grad_norm": 0.4885726869106293,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 8201
+    },
+    {
+      "epoch": 0.9528899215800174,
+      "grad_norm": 0.4775313436985016,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 8202
+    },
+    {
+      "epoch": 0.9530060993319779,
+      "grad_norm": 0.4628119170665741,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 8203
+    },
+    {
+      "epoch": 0.9531222770839384,
+      "grad_norm": 0.4696086049079895,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 8204
+    },
+    {
+      "epoch": 0.953238454835899,
+      "grad_norm": 0.4801238179206848,
+      "learning_rate": 0.0001,
+      "loss": 1.6711,
+      "step": 8205
+    },
+    {
+      "epoch": 0.9533546325878595,
+      "grad_norm": 0.46630024909973145,
+      "learning_rate": 0.0001,
+      "loss": 1.6602,
+      "step": 8206
+    },
+    {
+      "epoch": 0.95347081033982,
+      "grad_norm": 0.5154693722724915,
+      "learning_rate": 0.0001,
+      "loss": 1.7074,
+      "step": 8207
+    },
+    {
+      "epoch": 0.9535869880917804,
+      "grad_norm": 0.456986665725708,
+      "learning_rate": 0.0001,
+      "loss": 1.4658,
+      "step": 8208
+    },
+    {
+      "epoch": 0.9537031658437409,
+      "grad_norm": 0.4490494430065155,
+      "learning_rate": 0.0001,
+      "loss": 1.4067,
+      "step": 8209
+    },
+    {
+      "epoch": 0.9538193435957014,
+      "grad_norm": 0.4910639226436615,
+      "learning_rate": 0.0001,
+      "loss": 1.6069,
+      "step": 8210
+    },
+    {
+      "epoch": 0.9539355213476619,
+      "grad_norm": 0.5007901787757874,
+      "learning_rate": 0.0001,
+      "loss": 1.6315,
+      "step": 8211
+    },
+    {
+      "epoch": 0.9540516990996224,
+      "grad_norm": 0.45628076791763306,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 8212
+    },
+    {
+      "epoch": 0.9541678768515829,
+      "grad_norm": 0.48491212725639343,
+      "learning_rate": 0.0001,
+      "loss": 1.605,
+      "step": 8213
+    },
+    {
+      "epoch": 0.9542840546035434,
+      "grad_norm": 0.48505932092666626,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 8214
+    },
+    {
+      "epoch": 0.954400232355504,
+      "grad_norm": 0.5037823915481567,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 8215
+    },
+    {
+      "epoch": 0.9545164101074645,
+      "grad_norm": 0.5135097503662109,
+      "learning_rate": 0.0001,
+      "loss": 1.7156,
+      "step": 8216
+    },
+    {
+      "epoch": 0.954632587859425,
+      "grad_norm": 0.48048651218414307,
+      "learning_rate": 0.0001,
+      "loss": 1.7173,
+      "step": 8217
+    },
+    {
+      "epoch": 0.9547487656113854,
+      "grad_norm": 0.48271870613098145,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 8218
+    },
+    {
+      "epoch": 0.9548649433633459,
+      "grad_norm": 0.48280853033065796,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 8219
+    },
+    {
+      "epoch": 0.9549811211153064,
+      "grad_norm": 0.4850626289844513,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 8220
+    },
+    {
+      "epoch": 0.9550972988672669,
+      "grad_norm": 0.48917168378829956,
+      "learning_rate": 0.0001,
+      "loss": 1.6632,
+      "step": 8221
+    },
+    {
+      "epoch": 0.9552134766192274,
+      "grad_norm": 0.4904051423072815,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 8222
+    },
+    {
+      "epoch": 0.9553296543711879,
+      "grad_norm": 0.47541365027427673,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 8223
+    },
+    {
+      "epoch": 0.9554458321231484,
+      "grad_norm": 0.49688780307769775,
+      "learning_rate": 0.0001,
+      "loss": 1.7412,
+      "step": 8224
+    },
+    {
+      "epoch": 0.9555620098751089,
+      "grad_norm": 0.5080480575561523,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 8225
+    },
+    {
+      "epoch": 0.9556781876270695,
+      "grad_norm": 0.48701146245002747,
+      "learning_rate": 0.0001,
+      "loss": 1.5155,
+      "step": 8226
+    },
+    {
+      "epoch": 0.95579436537903,
+      "grad_norm": 0.4538536071777344,
+      "learning_rate": 0.0001,
+      "loss": 1.5677,
+      "step": 8227
+    },
+    {
+      "epoch": 0.9559105431309904,
+      "grad_norm": 0.47831106185913086,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 8228
+    },
+    {
+      "epoch": 0.9560267208829509,
+      "grad_norm": 0.4827806055545807,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 8229
+    },
+    {
+      "epoch": 0.9561428986349114,
+      "grad_norm": 0.4946243166923523,
+      "learning_rate": 0.0001,
+      "loss": 1.6507,
+      "step": 8230
+    },
+    {
+      "epoch": 0.9562590763868719,
+      "grad_norm": 0.5115143060684204,
+      "learning_rate": 0.0001,
+      "loss": 1.6444,
+      "step": 8231
+    },
+    {
+      "epoch": 0.9563752541388324,
+      "grad_norm": 0.4955557584762573,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 8232
+    },
+    {
+      "epoch": 0.9564914318907929,
+      "grad_norm": 0.45829346776008606,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 8233
+    },
+    {
+      "epoch": 0.9566076096427534,
+      "grad_norm": 0.4455798864364624,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 8234
+    },
+    {
+      "epoch": 0.9567237873947139,
+      "grad_norm": 0.48835891485214233,
+      "learning_rate": 0.0001,
+      "loss": 1.6687,
+      "step": 8235
+    },
+    {
+      "epoch": 0.9568399651466745,
+      "grad_norm": 0.46689534187316895,
+      "learning_rate": 0.0001,
+      "loss": 1.6003,
+      "step": 8236
+    },
+    {
+      "epoch": 0.956956142898635,
+      "grad_norm": 0.5193210244178772,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 8237
+    },
+    {
+      "epoch": 0.9570723206505954,
+      "grad_norm": 0.46074503660202026,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 8238
+    },
+    {
+      "epoch": 0.9571884984025559,
+      "grad_norm": 0.48494353890419006,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 8239
+    },
+    {
+      "epoch": 0.9573046761545164,
+      "grad_norm": 0.45777463912963867,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 8240
+    },
+    {
+      "epoch": 0.9574208539064769,
+      "grad_norm": 0.49893805384635925,
+      "learning_rate": 0.0001,
+      "loss": 1.6708,
+      "step": 8241
+    },
+    {
+      "epoch": 0.9575370316584374,
+      "grad_norm": 0.500220537185669,
+      "learning_rate": 0.0001,
+      "loss": 1.6171,
+      "step": 8242
+    },
+    {
+      "epoch": 0.9576532094103979,
+      "grad_norm": 0.4701831638813019,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 8243
+    },
+    {
+      "epoch": 0.9577693871623584,
+      "grad_norm": 0.465260773897171,
+      "learning_rate": 0.0001,
+      "loss": 1.5155,
+      "step": 8244
+    },
+    {
+      "epoch": 0.9578855649143189,
+      "grad_norm": 0.4423646628856659,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 8245
+    },
+    {
+      "epoch": 0.9580017426662794,
+      "grad_norm": 0.49591514468193054,
+      "learning_rate": 0.0001,
+      "loss": 1.6665,
+      "step": 8246
+    },
+    {
+      "epoch": 0.95811792041824,
+      "grad_norm": 0.4624507427215576,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 8247
+    },
+    {
+      "epoch": 0.9582340981702004,
+      "grad_norm": 0.4537637531757355,
+      "learning_rate": 0.0001,
+      "loss": 1.3326,
+      "step": 8248
+    },
+    {
+      "epoch": 0.9583502759221609,
+      "grad_norm": 0.45610177516937256,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 8249
+    },
+    {
+      "epoch": 0.9584664536741214,
+      "grad_norm": 0.5058176517486572,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 8250
+    },
+    {
+      "epoch": 0.9585826314260819,
+      "grad_norm": 0.47740715742111206,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 8251
+    },
+    {
+      "epoch": 0.9586988091780424,
+      "grad_norm": 0.5462493300437927,
+      "learning_rate": 0.0001,
+      "loss": 1.643,
+      "step": 8252
+    },
+    {
+      "epoch": 0.9588149869300029,
+      "grad_norm": 0.4755356013774872,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 8253
+    },
+    {
+      "epoch": 0.9589311646819634,
+      "grad_norm": 0.4820975661277771,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 8254
+    },
+    {
+      "epoch": 0.9590473424339239,
+      "grad_norm": 0.4660854637622833,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 8255
+    },
+    {
+      "epoch": 0.9591635201858844,
+      "grad_norm": 0.5366430282592773,
+      "learning_rate": 0.0001,
+      "loss": 1.6443,
+      "step": 8256
+    },
+    {
+      "epoch": 0.959279697937845,
+      "grad_norm": 0.5088802576065063,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 8257
+    },
+    {
+      "epoch": 0.9593958756898054,
+      "grad_norm": 0.5079680681228638,
+      "learning_rate": 0.0001,
+      "loss": 1.753,
+      "step": 8258
+    },
+    {
+      "epoch": 0.9595120534417659,
+      "grad_norm": 0.4622834622859955,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 8259
+    },
+    {
+      "epoch": 0.9596282311937264,
+      "grad_norm": 0.5069734454154968,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 8260
+    },
+    {
+      "epoch": 0.9597444089456869,
+      "grad_norm": 0.493930459022522,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 8261
+    },
+    {
+      "epoch": 0.9598605866976474,
+      "grad_norm": 0.5060780644416809,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 8262
+    },
+    {
+      "epoch": 0.9599767644496079,
+      "grad_norm": 0.48548170924186707,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 8263
+    },
+    {
+      "epoch": 0.9600929422015684,
+      "grad_norm": 0.45788517594337463,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 8264
+    },
+    {
+      "epoch": 0.9602091199535289,
+      "grad_norm": 0.46271592378616333,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 8265
+    },
+    {
+      "epoch": 0.9603252977054894,
+      "grad_norm": 0.48142072558403015,
+      "learning_rate": 0.0001,
+      "loss": 1.627,
+      "step": 8266
+    },
+    {
+      "epoch": 0.9604414754574498,
+      "grad_norm": 0.47520777583122253,
+      "learning_rate": 0.0001,
+      "loss": 1.6576,
+      "step": 8267
+    },
+    {
+      "epoch": 0.9605576532094104,
+      "grad_norm": 0.4597654938697815,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 8268
+    },
+    {
+      "epoch": 0.9606738309613709,
+      "grad_norm": 0.4608597159385681,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 8269
+    },
+    {
+      "epoch": 0.9607900087133314,
+      "grad_norm": 0.4679330289363861,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 8270
+    },
+    {
+      "epoch": 0.9609061864652919,
+      "grad_norm": 0.49732252955436707,
+      "learning_rate": 0.0001,
+      "loss": 1.5015,
+      "step": 8271
+    },
+    {
+      "epoch": 0.9610223642172524,
+      "grad_norm": 0.4775616526603699,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 8272
+    },
+    {
+      "epoch": 0.9611385419692129,
+      "grad_norm": 0.5359445214271545,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 8273
+    },
+    {
+      "epoch": 0.9612547197211734,
+      "grad_norm": 0.4927753508090973,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 8274
+    },
+    {
+      "epoch": 0.9613708974731339,
+      "grad_norm": 0.48681047558784485,
+      "learning_rate": 0.0001,
+      "loss": 1.6162,
+      "step": 8275
+    },
+    {
+      "epoch": 0.9614870752250944,
+      "grad_norm": 0.49308332800865173,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 8276
+    },
+    {
+      "epoch": 0.9616032529770548,
+      "grad_norm": 0.48825904726982117,
+      "learning_rate": 0.0001,
+      "loss": 1.6006,
+      "step": 8277
+    },
+    {
+      "epoch": 0.9617194307290154,
+      "grad_norm": 0.5138652324676514,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 8278
+    },
+    {
+      "epoch": 0.9618356084809759,
+      "grad_norm": 0.5201594233512878,
+      "learning_rate": 0.0001,
+      "loss": 1.6361,
+      "step": 8279
+    },
+    {
+      "epoch": 0.9619517862329364,
+      "grad_norm": 0.5101884603500366,
+      "learning_rate": 0.0001,
+      "loss": 1.7062,
+      "step": 8280
+    },
+    {
+      "epoch": 0.9620679639848969,
+      "grad_norm": 0.45568764209747314,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 8281
+    },
+    {
+      "epoch": 0.9621841417368574,
+      "grad_norm": 0.46922385692596436,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 8282
+    },
+    {
+      "epoch": 0.9623003194888179,
+      "grad_norm": 0.47019103169441223,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 8283
+    },
+    {
+      "epoch": 0.9624164972407784,
+      "grad_norm": 0.5026654601097107,
+      "learning_rate": 0.0001,
+      "loss": 1.6182,
+      "step": 8284
+    },
+    {
+      "epoch": 0.9625326749927389,
+      "grad_norm": 0.5346139669418335,
+      "learning_rate": 0.0001,
+      "loss": 1.6782,
+      "step": 8285
+    },
+    {
+      "epoch": 0.9626488527446994,
+      "grad_norm": 0.4690621495246887,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 8286
+    },
+    {
+      "epoch": 0.9627650304966598,
+      "grad_norm": 0.5523574948310852,
+      "learning_rate": 0.0001,
+      "loss": 1.8614,
+      "step": 8287
+    },
+    {
+      "epoch": 0.9628812082486203,
+      "grad_norm": 0.5208613872528076,
+      "learning_rate": 0.0001,
+      "loss": 1.69,
+      "step": 8288
+    },
+    {
+      "epoch": 0.9629973860005809,
+      "grad_norm": 0.5181357264518738,
+      "learning_rate": 0.0001,
+      "loss": 1.6707,
+      "step": 8289
+    },
+    {
+      "epoch": 0.9631135637525414,
+      "grad_norm": 0.49998828768730164,
+      "learning_rate": 0.0001,
+      "loss": 1.57,
+      "step": 8290
+    },
+    {
+      "epoch": 0.9632297415045019,
+      "grad_norm": 0.47236523032188416,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 8291
+    },
+    {
+      "epoch": 0.9633459192564624,
+      "grad_norm": 0.49933215975761414,
+      "learning_rate": 0.0001,
+      "loss": 1.6872,
+      "step": 8292
+    },
+    {
+      "epoch": 0.9634620970084229,
+      "grad_norm": 0.48672953248023987,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 8293
+    },
+    {
+      "epoch": 0.9635782747603834,
+      "grad_norm": 0.45596054196357727,
+      "learning_rate": 0.0001,
+      "loss": 1.4982,
+      "step": 8294
+    },
+    {
+      "epoch": 0.9636944525123439,
+      "grad_norm": 0.47312960028648376,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 8295
+    },
+    {
+      "epoch": 0.9638106302643044,
+      "grad_norm": 0.5012199282646179,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 8296
+    },
+    {
+      "epoch": 0.9639268080162648,
+      "grad_norm": 0.516236424446106,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 8297
+    },
+    {
+      "epoch": 0.9640429857682253,
+      "grad_norm": 0.44847118854522705,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 8298
+    },
+    {
+      "epoch": 0.9641591635201859,
+      "grad_norm": 0.4940093457698822,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 8299
+    },
+    {
+      "epoch": 0.9642753412721464,
+      "grad_norm": 0.4744294285774231,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 8300
+    },
+    {
+      "epoch": 0.9643915190241069,
+      "grad_norm": 0.4880334436893463,
+      "learning_rate": 0.0001,
+      "loss": 1.5232,
+      "step": 8301
+    },
+    {
+      "epoch": 0.9645076967760674,
+      "grad_norm": 0.49763065576553345,
+      "learning_rate": 0.0001,
+      "loss": 1.6985,
+      "step": 8302
+    },
+    {
+      "epoch": 0.9646238745280279,
+      "grad_norm": 0.49351394176483154,
+      "learning_rate": 0.0001,
+      "loss": 1.7411,
+      "step": 8303
+    },
+    {
+      "epoch": 0.9647400522799884,
+      "grad_norm": 0.4660297930240631,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 8304
+    },
+    {
+      "epoch": 0.9648562300319489,
+      "grad_norm": 0.4791782796382904,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 8305
+    },
+    {
+      "epoch": 0.9649724077839094,
+      "grad_norm": 0.49400556087493896,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 8306
+    },
+    {
+      "epoch": 0.9650885855358698,
+      "grad_norm": 0.4803016483783722,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 8307
+    },
+    {
+      "epoch": 0.9652047632878303,
+      "grad_norm": 0.49062761664390564,
+      "learning_rate": 0.0001,
+      "loss": 1.6898,
+      "step": 8308
+    },
+    {
+      "epoch": 0.9653209410397908,
+      "grad_norm": 0.5109800696372986,
+      "learning_rate": 0.0001,
+      "loss": 1.631,
+      "step": 8309
+    },
+    {
+      "epoch": 0.9654371187917514,
+      "grad_norm": 0.503459632396698,
+      "learning_rate": 0.0001,
+      "loss": 1.6486,
+      "step": 8310
+    },
+    {
+      "epoch": 0.9655532965437119,
+      "grad_norm": 0.47792255878448486,
+      "learning_rate": 0.0001,
+      "loss": 1.5972,
+      "step": 8311
+    },
+    {
+      "epoch": 0.9656694742956724,
+      "grad_norm": 0.4999959468841553,
+      "learning_rate": 0.0001,
+      "loss": 1.6538,
+      "step": 8312
+    },
+    {
+      "epoch": 0.9657856520476329,
+      "grad_norm": 0.49376481771469116,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 8313
+    },
+    {
+      "epoch": 0.9659018297995934,
+      "grad_norm": 0.4859766662120819,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 8314
+    },
+    {
+      "epoch": 0.9660180075515539,
+      "grad_norm": 0.4656546413898468,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 8315
+    },
+    {
+      "epoch": 0.9661341853035144,
+      "grad_norm": 0.48763415217399597,
+      "learning_rate": 0.0001,
+      "loss": 1.5704,
+      "step": 8316
+    },
+    {
+      "epoch": 0.9662503630554748,
+      "grad_norm": 0.479206919670105,
+      "learning_rate": 0.0001,
+      "loss": 1.5964,
+      "step": 8317
+    },
+    {
+      "epoch": 0.9663665408074353,
+      "grad_norm": 0.48830246925354004,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 8318
+    },
+    {
+      "epoch": 0.9664827185593958,
+      "grad_norm": 0.49852919578552246,
+      "learning_rate": 0.0001,
+      "loss": 1.5722,
+      "step": 8319
+    },
+    {
+      "epoch": 0.9665988963113564,
+      "grad_norm": 0.45353084802627563,
+      "learning_rate": 0.0001,
+      "loss": 1.4958,
+      "step": 8320
+    },
+    {
+      "epoch": 0.9667150740633169,
+      "grad_norm": 0.49439936876296997,
+      "learning_rate": 0.0001,
+      "loss": 1.6234,
+      "step": 8321
+    },
+    {
+      "epoch": 0.9668312518152774,
+      "grad_norm": 0.48684781789779663,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 8322
+    },
+    {
+      "epoch": 0.9669474295672379,
+      "grad_norm": 0.47375577688217163,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 8323
+    },
+    {
+      "epoch": 0.9670636073191984,
+      "grad_norm": 0.45029014348983765,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 8324
+    },
+    {
+      "epoch": 0.9671797850711589,
+      "grad_norm": 0.49063849449157715,
+      "learning_rate": 0.0001,
+      "loss": 1.6855,
+      "step": 8325
+    },
+    {
+      "epoch": 0.9672959628231194,
+      "grad_norm": 0.5382382273674011,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 8326
+    },
+    {
+      "epoch": 0.9674121405750798,
+      "grad_norm": 0.49472776055336,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 8327
+    },
+    {
+      "epoch": 0.9675283183270403,
+      "grad_norm": 0.4821051359176636,
+      "learning_rate": 0.0001,
+      "loss": 1.6096,
+      "step": 8328
+    },
+    {
+      "epoch": 0.9676444960790008,
+      "grad_norm": 0.47823530435562134,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 8329
+    },
+    {
+      "epoch": 0.9677606738309613,
+      "grad_norm": 0.4891481399536133,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 8330
+    },
+    {
+      "epoch": 0.9678768515829219,
+      "grad_norm": 0.49164843559265137,
+      "learning_rate": 0.0001,
+      "loss": 1.6268,
+      "step": 8331
+    },
+    {
+      "epoch": 0.9679930293348824,
+      "grad_norm": 0.473734974861145,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 8332
+    },
+    {
+      "epoch": 0.9681092070868429,
+      "grad_norm": 0.47365325689315796,
+      "learning_rate": 0.0001,
+      "loss": 1.4787,
+      "step": 8333
+    },
+    {
+      "epoch": 0.9682253848388034,
+      "grad_norm": 0.4796622395515442,
+      "learning_rate": 0.0001,
+      "loss": 1.4766,
+      "step": 8334
+    },
+    {
+      "epoch": 0.9683415625907639,
+      "grad_norm": 0.5099849700927734,
+      "learning_rate": 0.0001,
+      "loss": 1.5258,
+      "step": 8335
+    },
+    {
+      "epoch": 0.9684577403427244,
+      "grad_norm": 0.4894762337207794,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 8336
+    },
+    {
+      "epoch": 0.9685739180946848,
+      "grad_norm": 0.5052227973937988,
+      "learning_rate": 0.0001,
+      "loss": 1.633,
+      "step": 8337
+    },
+    {
+      "epoch": 0.9686900958466453,
+      "grad_norm": 0.5599405169487,
+      "learning_rate": 0.0001,
+      "loss": 1.7586,
+      "step": 8338
+    },
+    {
+      "epoch": 0.9688062735986058,
+      "grad_norm": 0.5493488907814026,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 8339
+    },
+    {
+      "epoch": 0.9689224513505663,
+      "grad_norm": 0.47974830865859985,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 8340
+    },
+    {
+      "epoch": 0.9690386291025269,
+      "grad_norm": 0.46246853470802307,
+      "learning_rate": 0.0001,
+      "loss": 1.6754,
+      "step": 8341
+    },
+    {
+      "epoch": 0.9691548068544874,
+      "grad_norm": 0.5237904787063599,
+      "learning_rate": 0.0001,
+      "loss": 1.7271,
+      "step": 8342
+    },
+    {
+      "epoch": 0.9692709846064479,
+      "grad_norm": 0.49900978803634644,
+      "learning_rate": 0.0001,
+      "loss": 1.6926,
+      "step": 8343
+    },
+    {
+      "epoch": 0.9693871623584084,
+      "grad_norm": 0.45971980690956116,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 8344
+    },
+    {
+      "epoch": 0.9695033401103689,
+      "grad_norm": 0.46642741560935974,
+      "learning_rate": 0.0001,
+      "loss": 1.5411,
+      "step": 8345
+    },
+    {
+      "epoch": 0.9696195178623294,
+      "grad_norm": 0.4861880838871002,
+      "learning_rate": 0.0001,
+      "loss": 1.6039,
+      "step": 8346
+    },
+    {
+      "epoch": 0.9697356956142898,
+      "grad_norm": 0.49903160333633423,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 8347
+    },
+    {
+      "epoch": 0.9698518733662503,
+      "grad_norm": 0.47082969546318054,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 8348
+    },
+    {
+      "epoch": 0.9699680511182108,
+      "grad_norm": 0.48401105403900146,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 8349
+    },
+    {
+      "epoch": 0.9700842288701713,
+      "grad_norm": 0.4795251190662384,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 8350
+    },
+    {
+      "epoch": 0.9702004066221319,
+      "grad_norm": 0.4561566114425659,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 8351
+    },
+    {
+      "epoch": 0.9703165843740924,
+      "grad_norm": 0.48859137296676636,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 8352
+    },
+    {
+      "epoch": 0.9704327621260529,
+      "grad_norm": 0.447553426027298,
+      "learning_rate": 0.0001,
+      "loss": 1.4135,
+      "step": 8353
+    },
+    {
+      "epoch": 0.9705489398780134,
+      "grad_norm": 0.4990467429161072,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 8354
+    },
+    {
+      "epoch": 0.9706651176299739,
+      "grad_norm": 0.5028551816940308,
+      "learning_rate": 0.0001,
+      "loss": 1.7281,
+      "step": 8355
+    },
+    {
+      "epoch": 0.9707812953819344,
+      "grad_norm": 0.49683597683906555,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 8356
+    },
+    {
+      "epoch": 0.9708974731338949,
+      "grad_norm": 0.47198501229286194,
+      "learning_rate": 0.0001,
+      "loss": 1.6287,
+      "step": 8357
+    },
+    {
+      "epoch": 0.9710136508858553,
+      "grad_norm": 0.5202149152755737,
+      "learning_rate": 0.0001,
+      "loss": 1.6653,
+      "step": 8358
+    },
+    {
+      "epoch": 0.9711298286378158,
+      "grad_norm": 0.505222737789154,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 8359
+    },
+    {
+      "epoch": 0.9712460063897763,
+      "grad_norm": 0.5158810615539551,
+      "learning_rate": 0.0001,
+      "loss": 1.6364,
+      "step": 8360
+    },
+    {
+      "epoch": 0.9713621841417368,
+      "grad_norm": 0.5185474157333374,
+      "learning_rate": 0.0001,
+      "loss": 1.8292,
+      "step": 8361
+    },
+    {
+      "epoch": 0.9714783618936974,
+      "grad_norm": 0.479807049036026,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 8362
+    },
+    {
+      "epoch": 0.9715945396456579,
+      "grad_norm": 0.47190988063812256,
+      "learning_rate": 0.0001,
+      "loss": 1.6801,
+      "step": 8363
+    },
+    {
+      "epoch": 0.9717107173976184,
+      "grad_norm": 0.47049620747566223,
+      "learning_rate": 0.0001,
+      "loss": 1.5388,
+      "step": 8364
+    },
+    {
+      "epoch": 0.9718268951495789,
+      "grad_norm": 0.49020835757255554,
+      "learning_rate": 0.0001,
+      "loss": 1.6445,
+      "step": 8365
+    },
+    {
+      "epoch": 0.9719430729015394,
+      "grad_norm": 0.49755486845970154,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 8366
+    },
+    {
+      "epoch": 0.9720592506534999,
+      "grad_norm": 0.4959685504436493,
+      "learning_rate": 0.0001,
+      "loss": 1.6562,
+      "step": 8367
+    },
+    {
+      "epoch": 0.9721754284054603,
+      "grad_norm": 0.5201342701911926,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 8368
+    },
+    {
+      "epoch": 0.9722916061574208,
+      "grad_norm": 0.4995202422142029,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 8369
+    },
+    {
+      "epoch": 0.9724077839093813,
+      "grad_norm": 0.4514632225036621,
+      "learning_rate": 0.0001,
+      "loss": 1.3639,
+      "step": 8370
+    },
+    {
+      "epoch": 0.9725239616613418,
+      "grad_norm": 0.46928879618644714,
+      "learning_rate": 0.0001,
+      "loss": 1.4085,
+      "step": 8371
+    },
+    {
+      "epoch": 0.9726401394133024,
+      "grad_norm": 0.4802158772945404,
+      "learning_rate": 0.0001,
+      "loss": 1.5368,
+      "step": 8372
+    },
+    {
+      "epoch": 0.9727563171652629,
+      "grad_norm": 0.4651092290878296,
+      "learning_rate": 0.0001,
+      "loss": 1.5232,
+      "step": 8373
+    },
+    {
+      "epoch": 0.9728724949172234,
+      "grad_norm": 0.46335577964782715,
+      "learning_rate": 0.0001,
+      "loss": 1.4136,
+      "step": 8374
+    },
+    {
+      "epoch": 0.9729886726691839,
+      "grad_norm": 0.48499464988708496,
+      "learning_rate": 0.0001,
+      "loss": 1.7017,
+      "step": 8375
+    },
+    {
+      "epoch": 0.9731048504211444,
+      "grad_norm": 0.4987207055091858,
+      "learning_rate": 0.0001,
+      "loss": 1.6231,
+      "step": 8376
+    },
+    {
+      "epoch": 0.9732210281731049,
+      "grad_norm": 0.505811333656311,
+      "learning_rate": 0.0001,
+      "loss": 1.7212,
+      "step": 8377
+    },
+    {
+      "epoch": 0.9733372059250653,
+      "grad_norm": 0.49306151270866394,
+      "learning_rate": 0.0001,
+      "loss": 1.6688,
+      "step": 8378
+    },
+    {
+      "epoch": 0.9734533836770258,
+      "grad_norm": 0.4923427700996399,
+      "learning_rate": 0.0001,
+      "loss": 1.6344,
+      "step": 8379
+    },
+    {
+      "epoch": 0.9735695614289863,
+      "grad_norm": 0.4896833002567291,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 8380
+    },
+    {
+      "epoch": 0.9736857391809468,
+      "grad_norm": 0.52273029088974,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 8381
+    },
+    {
+      "epoch": 0.9738019169329073,
+      "grad_norm": 0.48786941170692444,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 8382
+    },
+    {
+      "epoch": 0.9739180946848679,
+      "grad_norm": 0.49467676877975464,
+      "learning_rate": 0.0001,
+      "loss": 1.7052,
+      "step": 8383
+    },
+    {
+      "epoch": 0.9740342724368284,
+      "grad_norm": 0.49212512373924255,
+      "learning_rate": 0.0001,
+      "loss": 1.6339,
+      "step": 8384
+    },
+    {
+      "epoch": 0.9741504501887889,
+      "grad_norm": 0.5036439299583435,
+      "learning_rate": 0.0001,
+      "loss": 1.8137,
+      "step": 8385
+    },
+    {
+      "epoch": 0.9742666279407494,
+      "grad_norm": 0.5129532814025879,
+      "learning_rate": 0.0001,
+      "loss": 1.7383,
+      "step": 8386
+    },
+    {
+      "epoch": 0.9743828056927099,
+      "grad_norm": 0.5009684562683105,
+      "learning_rate": 0.0001,
+      "loss": 1.6897,
+      "step": 8387
+    },
+    {
+      "epoch": 0.9744989834446703,
+      "grad_norm": 0.47049927711486816,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 8388
+    },
+    {
+      "epoch": 0.9746151611966308,
+      "grad_norm": 0.47478753328323364,
+      "learning_rate": 0.0001,
+      "loss": 1.6045,
+      "step": 8389
+    },
+    {
+      "epoch": 0.9747313389485913,
+      "grad_norm": 0.46627920866012573,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 8390
+    },
+    {
+      "epoch": 0.9748475167005518,
+      "grad_norm": 0.47100529074668884,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 8391
+    },
+    {
+      "epoch": 0.9749636944525123,
+      "grad_norm": 0.4588838815689087,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 8392
+    },
+    {
+      "epoch": 0.9750798722044729,
+      "grad_norm": 0.5259383916854858,
+      "learning_rate": 0.0001,
+      "loss": 1.6828,
+      "step": 8393
+    },
+    {
+      "epoch": 0.9751960499564334,
+      "grad_norm": 0.46649569272994995,
+      "learning_rate": 0.0001,
+      "loss": 1.5351,
+      "step": 8394
+    },
+    {
+      "epoch": 0.9753122277083939,
+      "grad_norm": 0.48280128836631775,
+      "learning_rate": 0.0001,
+      "loss": 1.8095,
+      "step": 8395
+    },
+    {
+      "epoch": 0.9754284054603544,
+      "grad_norm": 0.510216474533081,
+      "learning_rate": 0.0001,
+      "loss": 1.5507,
+      "step": 8396
+    },
+    {
+      "epoch": 0.9755445832123149,
+      "grad_norm": 0.48745256662368774,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 8397
+    },
+    {
+      "epoch": 0.9756607609642753,
+      "grad_norm": 0.5059899091720581,
+      "learning_rate": 0.0001,
+      "loss": 1.6724,
+      "step": 8398
+    },
+    {
+      "epoch": 0.9757769387162358,
+      "grad_norm": 0.4656660854816437,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 8399
+    },
+    {
+      "epoch": 0.9758931164681963,
+      "grad_norm": 0.5184774398803711,
+      "learning_rate": 0.0001,
+      "loss": 1.8107,
+      "step": 8400
+    },
+    {
+      "epoch": 0.9760092942201568,
+      "grad_norm": 0.4714641273021698,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 8401
+    },
+    {
+      "epoch": 0.9761254719721173,
+      "grad_norm": 0.4753645062446594,
+      "learning_rate": 0.0001,
+      "loss": 1.7,
+      "step": 8402
+    },
+    {
+      "epoch": 0.9762416497240778,
+      "grad_norm": 0.5242862701416016,
+      "learning_rate": 0.0001,
+      "loss": 1.8109,
+      "step": 8403
+    },
+    {
+      "epoch": 0.9763578274760384,
+      "grad_norm": 0.5038365721702576,
+      "learning_rate": 0.0001,
+      "loss": 1.8278,
+      "step": 8404
+    },
+    {
+      "epoch": 0.9764740052279989,
+      "grad_norm": 0.5028054714202881,
+      "learning_rate": 0.0001,
+      "loss": 1.7647,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9765901829799594,
+      "grad_norm": 0.4805586636066437,
+      "learning_rate": 0.0001,
+      "loss": 1.5734,
+      "step": 8406
+    },
+    {
+      "epoch": 0.9767063607319199,
+      "grad_norm": 0.4857881963253021,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 8407
+    },
+    {
+      "epoch": 0.9768225384838803,
+      "grad_norm": 0.47471508383750916,
+      "learning_rate": 0.0001,
+      "loss": 1.5072,
+      "step": 8408
+    },
+    {
+      "epoch": 0.9769387162358408,
+      "grad_norm": 0.4627721309661865,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 8409
+    },
+    {
+      "epoch": 0.9770548939878013,
+      "grad_norm": 0.4718911647796631,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9771710717397618,
+      "grad_norm": 0.5181329846382141,
+      "learning_rate": 0.0001,
+      "loss": 1.6377,
+      "step": 8411
+    },
+    {
+      "epoch": 0.9772872494917223,
+      "grad_norm": 0.4619660973548889,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 8412
+    },
+    {
+      "epoch": 0.9774034272436828,
+      "grad_norm": 0.5440229773521423,
+      "learning_rate": 0.0001,
+      "loss": 1.7901,
+      "step": 8413
+    },
+    {
+      "epoch": 0.9775196049956434,
+      "grad_norm": 0.4952663481235504,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 8414
+    },
+    {
+      "epoch": 0.9776357827476039,
+      "grad_norm": 0.4936034083366394,
+      "learning_rate": 0.0001,
+      "loss": 1.6011,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9777519604995644,
+      "grad_norm": 0.5232906341552734,
+      "learning_rate": 0.0001,
+      "loss": 1.5554,
+      "step": 8416
+    },
+    {
+      "epoch": 0.9778681382515249,
+      "grad_norm": 0.5204361081123352,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 8417
+    },
+    {
+      "epoch": 0.9779843160034853,
+      "grad_norm": 0.4749402701854706,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 8418
+    },
+    {
+      "epoch": 0.9781004937554458,
+      "grad_norm": 0.4584287106990814,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 8419
+    },
+    {
+      "epoch": 0.9782166715074063,
+      "grad_norm": 0.4839183986186981,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9783328492593668,
+      "grad_norm": 0.4972332715988159,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 8421
+    },
+    {
+      "epoch": 0.9784490270113273,
+      "grad_norm": 0.5022913813591003,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 8422
+    },
+    {
+      "epoch": 0.9785652047632878,
+      "grad_norm": 0.5144555568695068,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 8423
+    },
+    {
+      "epoch": 0.9786813825152483,
+      "grad_norm": 0.5038486123085022,
+      "learning_rate": 0.0001,
+      "loss": 1.6212,
+      "step": 8424
+    },
+    {
+      "epoch": 0.9787975602672089,
+      "grad_norm": 0.45622316002845764,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 8425
+    },
+    {
+      "epoch": 0.9789137380191694,
+      "grad_norm": 0.48347601294517517,
+      "learning_rate": 0.0001,
+      "loss": 1.5393,
+      "step": 8426
+    },
+    {
+      "epoch": 0.9790299157711299,
+      "grad_norm": 0.5117625594139099,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 8427
+    },
+    {
+      "epoch": 0.9791460935230903,
+      "grad_norm": 0.47401314973831177,
+      "learning_rate": 0.0001,
+      "loss": 1.6127,
+      "step": 8428
+    },
+    {
+      "epoch": 0.9792622712750508,
+      "grad_norm": 0.4935537278652191,
+      "learning_rate": 0.0001,
+      "loss": 1.6038,
+      "step": 8429
+    },
+    {
+      "epoch": 0.9793784490270113,
+      "grad_norm": 0.47367969155311584,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9794946267789718,
+      "grad_norm": 0.4980182647705078,
+      "learning_rate": 0.0001,
+      "loss": 1.6661,
+      "step": 8431
+    },
+    {
+      "epoch": 0.9796108045309323,
+      "grad_norm": 0.5033068656921387,
+      "learning_rate": 0.0001,
+      "loss": 1.6927,
+      "step": 8432
+    },
+    {
+      "epoch": 0.9797269822828928,
+      "grad_norm": 0.49150750041007996,
+      "learning_rate": 0.0001,
+      "loss": 1.7359,
+      "step": 8433
+    },
+    {
+      "epoch": 0.9798431600348533,
+      "grad_norm": 0.4907017648220062,
+      "learning_rate": 0.0001,
+      "loss": 1.6193,
+      "step": 8434
+    },
+    {
+      "epoch": 0.9799593377868139,
+      "grad_norm": 0.46434399485588074,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 8435
+    },
+    {
+      "epoch": 0.9800755155387744,
+      "grad_norm": 0.4952322840690613,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 8436
+    },
+    {
+      "epoch": 0.9801916932907349,
+      "grad_norm": 0.4482409358024597,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 8437
+    },
+    {
+      "epoch": 0.9803078710426953,
+      "grad_norm": 0.4462791979312897,
+      "learning_rate": 0.0001,
+      "loss": 1.4427,
+      "step": 8438
+    },
+    {
+      "epoch": 0.9804240487946558,
+      "grad_norm": 0.5374612808227539,
+      "learning_rate": 0.0001,
+      "loss": 1.8228,
+      "step": 8439
+    },
+    {
+      "epoch": 0.9805402265466163,
+      "grad_norm": 0.5060397386550903,
+      "learning_rate": 0.0001,
+      "loss": 1.693,
+      "step": 8440
+    },
+    {
+      "epoch": 0.9806564042985768,
+      "grad_norm": 0.4491170048713684,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 8441
+    },
+    {
+      "epoch": 0.9807725820505373,
+      "grad_norm": 0.49389106035232544,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 8442
+    },
+    {
+      "epoch": 0.9808887598024978,
+      "grad_norm": 0.49476566910743713,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 8443
+    },
+    {
+      "epoch": 0.9810049375544583,
+      "grad_norm": 0.45577380061149597,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 8444
+    },
+    {
+      "epoch": 0.9811211153064188,
+      "grad_norm": 0.48937976360321045,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 8445
+    },
+    {
+      "epoch": 0.9812372930583794,
+      "grad_norm": 0.4766305983066559,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 8446
+    },
+    {
+      "epoch": 0.9813534708103399,
+      "grad_norm": 0.4743248224258423,
+      "learning_rate": 0.0001,
+      "loss": 1.6103,
+      "step": 8447
+    },
+    {
+      "epoch": 0.9814696485623003,
+      "grad_norm": 0.504391610622406,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 8448
+    },
+    {
+      "epoch": 0.9815858263142608,
+      "grad_norm": 0.46172428131103516,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 8449
+    },
+    {
+      "epoch": 0.9817020040662213,
+      "grad_norm": 0.4999525249004364,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 8450
+    },
+    {
+      "epoch": 0.9818181818181818,
+      "grad_norm": 0.49794474244117737,
+      "learning_rate": 0.0001,
+      "loss": 1.7356,
+      "step": 8451
+    },
+    {
+      "epoch": 0.9819343595701423,
+      "grad_norm": 0.47993290424346924,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 8452
+    },
+    {
+      "epoch": 0.9820505373221028,
+      "grad_norm": 0.5240808725357056,
+      "learning_rate": 0.0001,
+      "loss": 1.7495,
+      "step": 8453
+    },
+    {
+      "epoch": 0.9821667150740633,
+      "grad_norm": 0.486801415681839,
+      "learning_rate": 0.0001,
+      "loss": 1.6978,
+      "step": 8454
+    },
+    {
+      "epoch": 0.9822828928260238,
+      "grad_norm": 0.5116367936134338,
+      "learning_rate": 0.0001,
+      "loss": 1.6763,
+      "step": 8455
+    },
+    {
+      "epoch": 0.9823990705779844,
+      "grad_norm": 0.4694158732891083,
+      "learning_rate": 0.0001,
+      "loss": 1.638,
+      "step": 8456
+    },
+    {
+      "epoch": 0.9825152483299449,
+      "grad_norm": 0.4877493679523468,
+      "learning_rate": 0.0001,
+      "loss": 1.5703,
+      "step": 8457
+    },
+    {
+      "epoch": 0.9826314260819053,
+      "grad_norm": 0.5048372745513916,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 8458
+    },
+    {
+      "epoch": 0.9827476038338658,
+      "grad_norm": 0.4648038148880005,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 8459
+    },
+    {
+      "epoch": 0.9828637815858263,
+      "grad_norm": 0.48997968435287476,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 8460
+    },
+    {
+      "epoch": 0.9829799593377868,
+      "grad_norm": 0.4944762885570526,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 8461
+    },
+    {
+      "epoch": 0.9830961370897473,
+      "grad_norm": 0.4919930398464203,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 8462
+    },
+    {
+      "epoch": 0.9832123148417078,
+      "grad_norm": 0.49641650915145874,
+      "learning_rate": 0.0001,
+      "loss": 1.5649,
+      "step": 8463
+    },
+    {
+      "epoch": 0.9833284925936683,
+      "grad_norm": 0.4891132414340973,
+      "learning_rate": 0.0001,
+      "loss": 1.6833,
+      "step": 8464
+    },
+    {
+      "epoch": 0.9834446703456288,
+      "grad_norm": 0.4601823687553406,
+      "learning_rate": 0.0001,
+      "loss": 1.531,
+      "step": 8465
+    },
+    {
+      "epoch": 0.9835608480975893,
+      "grad_norm": 0.4825168550014496,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 8466
+    },
+    {
+      "epoch": 0.9836770258495499,
+      "grad_norm": 0.5182327032089233,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 8467
+    },
+    {
+      "epoch": 0.9837932036015103,
+      "grad_norm": 0.47299832105636597,
+      "learning_rate": 0.0001,
+      "loss": 1.6853,
+      "step": 8468
+    },
+    {
+      "epoch": 0.9839093813534708,
+      "grad_norm": 0.505413293838501,
+      "learning_rate": 0.0001,
+      "loss": 1.664,
+      "step": 8469
+    },
+    {
+      "epoch": 0.9840255591054313,
+      "grad_norm": 0.4697488844394684,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 8470
+    },
+    {
+      "epoch": 0.9841417368573918,
+      "grad_norm": 0.4823164641857147,
+      "learning_rate": 0.0001,
+      "loss": 1.6351,
+      "step": 8471
+    },
+    {
+      "epoch": 0.9842579146093523,
+      "grad_norm": 0.49613675475120544,
+      "learning_rate": 0.0001,
+      "loss": 1.6637,
+      "step": 8472
+    },
+    {
+      "epoch": 0.9843740923613128,
+      "grad_norm": 0.49066096544265747,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 8473
+    },
+    {
+      "epoch": 0.9844902701132733,
+      "grad_norm": 0.4947955012321472,
+      "learning_rate": 0.0001,
+      "loss": 1.6372,
+      "step": 8474
+    },
+    {
+      "epoch": 0.9846064478652338,
+      "grad_norm": 0.4665144681930542,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 8475
+    },
+    {
+      "epoch": 0.9847226256171943,
+      "grad_norm": 0.5288201570510864,
+      "learning_rate": 0.0001,
+      "loss": 1.8544,
+      "step": 8476
+    },
+    {
+      "epoch": 0.9848388033691549,
+      "grad_norm": 0.4513753354549408,
+      "learning_rate": 0.0001,
+      "loss": 1.3675,
+      "step": 8477
+    },
+    {
+      "epoch": 0.9849549811211153,
+      "grad_norm": 0.49009808897972107,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 8478
+    },
+    {
+      "epoch": 0.9850711588730758,
+      "grad_norm": 0.5423754453659058,
+      "learning_rate": 0.0001,
+      "loss": 1.7736,
+      "step": 8479
+    },
+    {
+      "epoch": 0.9851873366250363,
+      "grad_norm": 0.5365590453147888,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 8480
+    },
+    {
+      "epoch": 0.9853035143769968,
+      "grad_norm": 0.44759395718574524,
+      "learning_rate": 0.0001,
+      "loss": 1.5299,
+      "step": 8481
+    },
+    {
+      "epoch": 0.9854196921289573,
+      "grad_norm": 0.5050853490829468,
+      "learning_rate": 0.0001,
+      "loss": 1.5673,
+      "step": 8482
+    },
+    {
+      "epoch": 0.9855358698809178,
+      "grad_norm": 0.4968460202217102,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 8483
+    },
+    {
+      "epoch": 0.9856520476328783,
+      "grad_norm": 0.4942067265510559,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 8484
+    },
+    {
+      "epoch": 0.9857682253848388,
+      "grad_norm": 0.5043184161186218,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 8485
+    },
+    {
+      "epoch": 0.9858844031367993,
+      "grad_norm": 0.4849433898925781,
+      "learning_rate": 0.0001,
+      "loss": 1.6144,
+      "step": 8486
+    },
+    {
+      "epoch": 0.9860005808887597,
+      "grad_norm": 0.5094505548477173,
+      "learning_rate": 0.0001,
+      "loss": 1.6876,
+      "step": 8487
+    },
+    {
+      "epoch": 0.9861167586407203,
+      "grad_norm": 0.5001594424247742,
+      "learning_rate": 0.0001,
+      "loss": 1.6043,
+      "step": 8488
+    },
+    {
+      "epoch": 0.9862329363926808,
+      "grad_norm": 0.47105178236961365,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 8489
+    },
+    {
+      "epoch": 0.9863491141446413,
+      "grad_norm": 0.4758048951625824,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 8490
+    },
+    {
+      "epoch": 0.9864652918966018,
+      "grad_norm": 0.4678146541118622,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 8491
+    },
+    {
+      "epoch": 0.9865814696485623,
+      "grad_norm": 0.4687998294830322,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 8492
+    },
+    {
+      "epoch": 0.9866976474005228,
+      "grad_norm": 0.5320387482643127,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 8493
+    },
+    {
+      "epoch": 0.9868138251524833,
+      "grad_norm": 0.49444085359573364,
+      "learning_rate": 0.0001,
+      "loss": 1.64,
+      "step": 8494
+    },
+    {
+      "epoch": 0.9869300029044438,
+      "grad_norm": 0.4462701678276062,
+      "learning_rate": 0.0001,
+      "loss": 1.491,
+      "step": 8495
+    },
+    {
+      "epoch": 0.9870461806564043,
+      "grad_norm": 0.46784061193466187,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 8496
+    },
+    {
+      "epoch": 0.9871623584083647,
+      "grad_norm": 0.4919097423553467,
+      "learning_rate": 0.0001,
+      "loss": 1.6808,
+      "step": 8497
+    },
+    {
+      "epoch": 0.9872785361603253,
+      "grad_norm": 0.48356717824935913,
+      "learning_rate": 0.0001,
+      "loss": 1.7458,
+      "step": 8498
+    },
+    {
+      "epoch": 0.9873947139122858,
+      "grad_norm": 0.4529293477535248,
+      "learning_rate": 0.0001,
+      "loss": 1.6302,
+      "step": 8499
+    },
+    {
+      "epoch": 0.9875108916642463,
+      "grad_norm": 0.48209211230278015,
+      "learning_rate": 0.0001,
+      "loss": 1.63,
+      "step": 8500
+    },
+    {
+      "epoch": 0.9876270694162068,
+      "grad_norm": 0.4654987156391144,
+      "learning_rate": 0.0001,
+      "loss": 1.6053,
+      "step": 8501
+    },
+    {
+      "epoch": 0.9877432471681673,
+      "grad_norm": 0.4693775177001953,
+      "learning_rate": 0.0001,
+      "loss": 1.4176,
+      "step": 8502
+    },
+    {
+      "epoch": 0.9878594249201278,
+      "grad_norm": 0.4983534514904022,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 8503
+    },
+    {
+      "epoch": 0.9879756026720883,
+      "grad_norm": 0.4841345250606537,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 8504
+    },
+    {
+      "epoch": 0.9880917804240488,
+      "grad_norm": 0.5288339853286743,
+      "learning_rate": 0.0001,
+      "loss": 1.7306,
+      "step": 8505
+    },
+    {
+      "epoch": 0.9882079581760093,
+      "grad_norm": 0.482524037361145,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 8506
+    },
+    {
+      "epoch": 0.9883241359279697,
+      "grad_norm": 0.48337680101394653,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 8507
+    },
+    {
+      "epoch": 0.9884403136799302,
+      "grad_norm": 0.4871644079685211,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 8508
+    },
+    {
+      "epoch": 0.9885564914318908,
+      "grad_norm": 0.47832533717155457,
+      "learning_rate": 0.0001,
+      "loss": 1.6587,
+      "step": 8509
+    },
+    {
+      "epoch": 0.9886726691838513,
+      "grad_norm": 0.4998105764389038,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 8510
+    },
+    {
+      "epoch": 0.9887888469358118,
+      "grad_norm": 0.4856521189212799,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 8511
+    },
+    {
+      "epoch": 0.9889050246877723,
+      "grad_norm": 0.4702942967414856,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 8512
+    },
+    {
+      "epoch": 0.9890212024397328,
+      "grad_norm": 0.4836369752883911,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 8513
+    },
+    {
+      "epoch": 0.9891373801916933,
+      "grad_norm": 0.4650129973888397,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 8514
+    },
+    {
+      "epoch": 0.9892535579436538,
+      "grad_norm": 0.47725415229797363,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 8515
+    },
+    {
+      "epoch": 0.9893697356956143,
+      "grad_norm": 0.4913800358772278,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 8516
+    },
+    {
+      "epoch": 0.9894859134475747,
+      "grad_norm": 0.5163478851318359,
+      "learning_rate": 0.0001,
+      "loss": 1.6778,
+      "step": 8517
+    },
+    {
+      "epoch": 0.9896020911995352,
+      "grad_norm": 0.4827187657356262,
+      "learning_rate": 0.0001,
+      "loss": 1.6692,
+      "step": 8518
+    },
+    {
+      "epoch": 0.9897182689514958,
+      "grad_norm": 0.5060868859291077,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 8519
+    },
+    {
+      "epoch": 0.9898344467034563,
+      "grad_norm": 0.5102714896202087,
+      "learning_rate": 0.0001,
+      "loss": 1.6579,
+      "step": 8520
+    },
+    {
+      "epoch": 0.9899506244554168,
+      "grad_norm": 0.4683992862701416,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 8521
+    },
+    {
+      "epoch": 0.9900668022073773,
+      "grad_norm": 0.4988732635974884,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 8522
+    },
+    {
+      "epoch": 0.9901829799593378,
+      "grad_norm": 0.48030632734298706,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 8523
+    },
+    {
+      "epoch": 0.9902991577112983,
+      "grad_norm": 0.48717671632766724,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 8524
+    },
+    {
+      "epoch": 0.9904153354632588,
+      "grad_norm": 0.4652939736843109,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 8525
+    },
+    {
+      "epoch": 0.9905315132152193,
+      "grad_norm": 0.4697415232658386,
+      "learning_rate": 0.0001,
+      "loss": 1.5953,
+      "step": 8526
+    },
+    {
+      "epoch": 0.9906476909671798,
+      "grad_norm": 0.4873330593109131,
+      "learning_rate": 0.0001,
+      "loss": 1.4776,
+      "step": 8527
+    },
+    {
+      "epoch": 0.9907638687191402,
+      "grad_norm": 0.5050543546676636,
+      "learning_rate": 0.0001,
+      "loss": 1.6988,
+      "step": 8528
+    },
+    {
+      "epoch": 0.9908800464711008,
+      "grad_norm": 0.499187171459198,
+      "learning_rate": 0.0001,
+      "loss": 1.7552,
+      "step": 8529
+    },
+    {
+      "epoch": 0.9909962242230613,
+      "grad_norm": 0.5250015258789062,
+      "learning_rate": 0.0001,
+      "loss": 1.7133,
+      "step": 8530
+    },
+    {
+      "epoch": 0.9911124019750218,
+      "grad_norm": 0.48826849460601807,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 8531
+    },
+    {
+      "epoch": 0.9912285797269823,
+      "grad_norm": 0.45894527435302734,
+      "learning_rate": 0.0001,
+      "loss": 1.4252,
+      "step": 8532
+    },
+    {
+      "epoch": 0.9913447574789428,
+      "grad_norm": 0.45296138525009155,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 8533
+    },
+    {
+      "epoch": 0.9914609352309033,
+      "grad_norm": 0.5120823383331299,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 8534
+    },
+    {
+      "epoch": 0.9915771129828638,
+      "grad_norm": 0.5219103693962097,
+      "learning_rate": 0.0001,
+      "loss": 1.7421,
+      "step": 8535
+    },
+    {
+      "epoch": 0.9916932907348243,
+      "grad_norm": 0.41183364391326904,
+      "learning_rate": 0.0001,
+      "loss": 1.2066,
+      "step": 8536
+    },
+    {
+      "epoch": 0.9918094684867848,
+      "grad_norm": 0.49815478920936584,
+      "learning_rate": 0.0001,
+      "loss": 1.671,
+      "step": 8537
+    },
+    {
+      "epoch": 0.9919256462387452,
+      "grad_norm": 0.4828040301799774,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 8538
+    },
+    {
+      "epoch": 0.9920418239907057,
+      "grad_norm": 0.48174577951431274,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 8539
+    },
+    {
+      "epoch": 0.9921580017426663,
+      "grad_norm": 0.4945439100265503,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 8540
+    },
+    {
+      "epoch": 0.9922741794946268,
+      "grad_norm": 0.5396121144294739,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 8541
+    },
+    {
+      "epoch": 0.9923903572465873,
+      "grad_norm": 0.4893397390842438,
+      "learning_rate": 0.0001,
+      "loss": 1.6438,
+      "step": 8542
+    },
+    {
+      "epoch": 0.9925065349985478,
+      "grad_norm": 0.4512374699115753,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 8543
+    },
+    {
+      "epoch": 0.9926227127505083,
+      "grad_norm": 0.5038131475448608,
+      "learning_rate": 0.0001,
+      "loss": 1.6511,
+      "step": 8544
+    },
+    {
+      "epoch": 0.9927388905024688,
+      "grad_norm": 0.47863298654556274,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 8545
+    },
+    {
+      "epoch": 0.9928550682544293,
+      "grad_norm": 0.4746008515357971,
+      "learning_rate": 0.0001,
+      "loss": 1.4573,
+      "step": 8546
+    },
+    {
+      "epoch": 0.9929712460063898,
+      "grad_norm": 0.5020937323570251,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 8547
+    },
+    {
+      "epoch": 0.9930874237583502,
+      "grad_norm": 0.5293759703636169,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 8548
+    },
+    {
+      "epoch": 0.9932036015103107,
+      "grad_norm": 0.49491435289382935,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 8549
+    },
+    {
+      "epoch": 0.9933197792622713,
+      "grad_norm": 0.4743204414844513,
+      "learning_rate": 0.0001,
+      "loss": 1.6571,
+      "step": 8550
+    },
+    {
+      "epoch": 0.9934359570142318,
+      "grad_norm": 0.47521528601646423,
+      "learning_rate": 0.0001,
+      "loss": 1.527,
+      "step": 8551
+    },
+    {
+      "epoch": 0.9935521347661923,
+      "grad_norm": 0.5073684453964233,
+      "learning_rate": 0.0001,
+      "loss": 1.7728,
+      "step": 8552
+    },
+    {
+      "epoch": 0.9936683125181528,
+      "grad_norm": 0.5444392561912537,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 8553
+    },
+    {
+      "epoch": 0.9937844902701133,
+      "grad_norm": 0.47420382499694824,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 8554
+    },
+    {
+      "epoch": 0.9939006680220738,
+      "grad_norm": 0.4889985918998718,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9940168457740343,
+      "grad_norm": 0.49144667387008667,
+      "learning_rate": 0.0001,
+      "loss": 1.6487,
+      "step": 8556
+    },
+    {
+      "epoch": 0.9941330235259948,
+      "grad_norm": 0.5177491307258606,
+      "learning_rate": 0.0001,
+      "loss": 1.6854,
+      "step": 8557
+    },
+    {
+      "epoch": 0.9942492012779552,
+      "grad_norm": 0.4622519612312317,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 8558
+    },
+    {
+      "epoch": 0.9943653790299157,
+      "grad_norm": 0.463818222284317,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 8559
+    },
+    {
+      "epoch": 0.9944815567818762,
+      "grad_norm": 0.46386709809303284,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9945977345338368,
+      "grad_norm": 0.4772911071777344,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 8561
+    },
+    {
+      "epoch": 0.9947139122857973,
+      "grad_norm": 0.47931206226348877,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 8562
+    },
+    {
+      "epoch": 0.9948300900377578,
+      "grad_norm": 0.5052580237388611,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 8563
+    },
+    {
+      "epoch": 0.9949462677897183,
+      "grad_norm": 0.46733200550079346,
+      "learning_rate": 0.0001,
+      "loss": 1.3345,
+      "step": 8564
+    },
+    {
+      "epoch": 0.9950624455416788,
+      "grad_norm": 0.5169939994812012,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9951786232936393,
+      "grad_norm": 0.5090740919113159,
+      "learning_rate": 0.0001,
+      "loss": 1.7412,
+      "step": 8566
+    },
+    {
+      "epoch": 0.9952948010455998,
+      "grad_norm": 0.4987412691116333,
+      "learning_rate": 0.0001,
+      "loss": 1.4844,
+      "step": 8567
+    },
+    {
+      "epoch": 0.9954109787975602,
+      "grad_norm": 0.44403526186943054,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 8568
+    },
+    {
+      "epoch": 0.9955271565495207,
+      "grad_norm": 0.5056188702583313,
+      "learning_rate": 0.0001,
+      "loss": 1.659,
+      "step": 8569
+    },
+    {
+      "epoch": 0.9956433343014812,
+      "grad_norm": 0.4768769145011902,
+      "learning_rate": 0.0001,
+      "loss": 1.4187,
+      "step": 8570
+    },
+    {
+      "epoch": 0.9957595120534418,
+      "grad_norm": 0.5366693735122681,
+      "learning_rate": 0.0001,
+      "loss": 1.6996,
+      "step": 8571
+    },
+    {
+      "epoch": 0.9958756898054023,
+      "grad_norm": 0.49038922786712646,
+      "learning_rate": 0.0001,
+      "loss": 1.6859,
+      "step": 8572
+    },
+    {
+      "epoch": 0.9959918675573628,
+      "grad_norm": 0.4839872419834137,
+      "learning_rate": 0.0001,
+      "loss": 1.7012,
+      "step": 8573
+    },
+    {
+      "epoch": 0.9961080453093233,
+      "grad_norm": 0.4965837001800537,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 8574
+    },
+    {
+      "epoch": 0.9962242230612838,
+      "grad_norm": 0.4981982111930847,
+      "learning_rate": 0.0001,
+      "loss": 1.7764,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9963404008132443,
+      "grad_norm": 0.47916489839553833,
+      "learning_rate": 0.0001,
+      "loss": 1.4184,
+      "step": 8576
+    },
+    {
+      "epoch": 0.9964565785652048,
+      "grad_norm": 0.5100079774856567,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 8577
+    },
+    {
+      "epoch": 0.9965727563171652,
+      "grad_norm": 0.46579378843307495,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 8578
+    },
+    {
+      "epoch": 0.9966889340691257,
+      "grad_norm": 0.4757480323314667,
+      "learning_rate": 0.0001,
+      "loss": 1.5739,
+      "step": 8579
+    },
+    {
+      "epoch": 0.9968051118210862,
+      "grad_norm": 0.49027368426322937,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9969212895730467,
+      "grad_norm": 0.4813475012779236,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 8581
+    },
+    {
+      "epoch": 0.9970374673250073,
+      "grad_norm": 0.4723621606826782,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 8582
+    },
+    {
+      "epoch": 0.9971536450769678,
+      "grad_norm": 0.476460337638855,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 8583
+    },
+    {
+      "epoch": 0.9972698228289283,
+      "grad_norm": 0.4537820518016815,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 8584
+    },
+    {
+      "epoch": 0.9973860005808888,
+      "grad_norm": 0.4895983636379242,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9975021783328493,
+      "grad_norm": 0.4980904161930084,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 8586
+    },
+    {
+      "epoch": 0.9976183560848098,
+      "grad_norm": 0.4741244912147522,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 8587
+    },
+    {
+      "epoch": 0.9977345338367702,
+      "grad_norm": 0.4763984680175781,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 8588
+    },
+    {
+      "epoch": 0.9978507115887307,
+      "grad_norm": 0.49783235788345337,
+      "learning_rate": 0.0001,
+      "loss": 1.6471,
+      "step": 8589
+    },
+    {
+      "epoch": 0.9979668893406912,
+      "grad_norm": 0.5360357165336609,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9980830670926517,
+      "grad_norm": 0.5079225301742554,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 8591
+    },
+    {
+      "epoch": 0.9981992448446123,
+      "grad_norm": 0.5056329965591431,
+      "learning_rate": 0.0001,
+      "loss": 1.6868,
+      "step": 8592
+    },
+    {
+      "epoch": 0.9983154225965728,
+      "grad_norm": 0.46236658096313477,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 8593
+    },
+    {
+      "epoch": 0.9984316003485333,
+      "grad_norm": 0.5396600365638733,
+      "learning_rate": 0.0001,
+      "loss": 1.807,
+      "step": 8594
+    },
+    {
+      "epoch": 0.9985477781004938,
+      "grad_norm": 0.46349719166755676,
+      "learning_rate": 0.0001,
+      "loss": 1.6027,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9986639558524543,
+      "grad_norm": 0.5006660223007202,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 8596
+    },
+    {
+      "epoch": 0.9987801336044148,
+      "grad_norm": 0.4902550280094147,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 8597
+    },
+    {
+      "epoch": 0.9988963113563752,
+      "grad_norm": 0.48567691445350647,
+      "learning_rate": 0.0001,
+      "loss": 1.561,
+      "step": 8598
+    },
+    {
+      "epoch": 0.9990124891083357,
+      "grad_norm": 0.47548311948776245,
+      "learning_rate": 0.0001,
+      "loss": 1.7234,
+      "step": 8599
+    },
+    {
+      "epoch": 0.9991286668602962,
+      "grad_norm": 0.47243523597717285,
+      "learning_rate": 0.0001,
+      "loss": 1.5788,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9992448446122567,
+      "grad_norm": 0.4965050518512726,
+      "learning_rate": 0.0001,
+      "loss": 1.6962,
+      "step": 8601
+    },
+    {
+      "epoch": 0.9993610223642172,
+      "grad_norm": 0.5007203817367554,
+      "learning_rate": 0.0001,
+      "loss": 1.707,
+      "step": 8602
+    },
+    {
+      "epoch": 0.9994772001161778,
+      "grad_norm": 0.48062342405319214,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 8603
+    },
+    {
+      "epoch": 0.9995933778681383,
+      "grad_norm": 0.5184653401374817,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 8604
+    },
+    {
+      "epoch": 0.9997095556200988,
+      "grad_norm": 0.4720516502857208,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 8605
+    },
+    {
+      "epoch": 0.9998257333720593,
+      "grad_norm": 0.4640704393386841,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 8606
+    },
+    {
+      "epoch": 0.9999419111240198,
+      "grad_norm": 0.4718928933143616,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 8607
+    },
+    {
+      "epoch": 1.0000580888759802,
+      "grad_norm": 0.4901491701602936,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 8608
+    },
+    {
+      "epoch": 1.0001742666279407,
+      "grad_norm": 0.5047919154167175,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 8609
+    },
+    {
+      "epoch": 1.0002904443799012,
+      "grad_norm": 0.5121291279792786,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 8610
+    },
+    {
+      "epoch": 1.0004066221318617,
+      "grad_norm": 0.46516066789627075,
+      "learning_rate": 0.0001,
+      "loss": 1.3715,
+      "step": 8611
+    },
+    {
+      "epoch": 1.0005227998838222,
+      "grad_norm": 0.4786158800125122,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 8612
+    },
+    {
+      "epoch": 1.0006389776357827,
+      "grad_norm": 0.507445752620697,
+      "learning_rate": 0.0001,
+      "loss": 1.4281,
+      "step": 8613
+    },
+    {
+      "epoch": 1.0007551553877432,
+      "grad_norm": 0.5036141872406006,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 8614
+    },
+    {
+      "epoch": 1.0008713331397037,
+      "grad_norm": 0.46509867906570435,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 8615
+    },
+    {
+      "epoch": 1.0009875108916642,
+      "grad_norm": 0.5023954510688782,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 8616
+    },
+    {
+      "epoch": 1.0011036886436246,
+      "grad_norm": 0.4837322235107422,
+      "learning_rate": 0.0001,
+      "loss": 1.3964,
+      "step": 8617
+    },
+    {
+      "epoch": 1.0012198663955854,
+      "grad_norm": 0.5714748501777649,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 8618
+    },
+    {
+      "epoch": 1.0013360441475458,
+      "grad_norm": 0.4612061381340027,
+      "learning_rate": 0.0001,
+      "loss": 1.3439,
+      "step": 8619
+    },
+    {
+      "epoch": 1.0014522218995063,
+      "grad_norm": 0.5135992765426636,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 8620
+    },
+    {
+      "epoch": 1.0015683996514668,
+      "grad_norm": 0.5336510539054871,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 8621
+    },
+    {
+      "epoch": 1.0016845774034273,
+      "grad_norm": 0.531232476234436,
+      "learning_rate": 0.0001,
+      "loss": 1.3025,
+      "step": 8622
+    },
+    {
+      "epoch": 1.0018007551553878,
+      "grad_norm": 0.5298891663551331,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 8623
+    },
+    {
+      "epoch": 1.0019169329073483,
+      "grad_norm": 0.5009737014770508,
+      "learning_rate": 0.0001,
+      "loss": 1.3511,
+      "step": 8624
+    },
+    {
+      "epoch": 1.0020331106593088,
+      "grad_norm": 0.5414840579032898,
+      "learning_rate": 0.0001,
+      "loss": 1.3387,
+      "step": 8625
+    },
+    {
+      "epoch": 1.0021492884112693,
+      "grad_norm": 0.5156089067459106,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 8626
+    },
+    {
+      "epoch": 1.0022654661632298,
+      "grad_norm": 0.48899945616722107,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 8627
+    },
+    {
+      "epoch": 1.0023816439151902,
+      "grad_norm": 0.5076591968536377,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 8628
+    },
+    {
+      "epoch": 1.0024978216671507,
+      "grad_norm": 0.5568975806236267,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 8629
+    },
+    {
+      "epoch": 1.0026139994191112,
+      "grad_norm": 0.5163382887840271,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 8630
+    },
+    {
+      "epoch": 1.0027301771710717,
+      "grad_norm": 0.5191821455955505,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 8631
+    },
+    {
+      "epoch": 1.0028463549230322,
+      "grad_norm": 0.5488196015357971,
+      "learning_rate": 0.0001,
+      "loss": 1.5607,
+      "step": 8632
+    },
+    {
+      "epoch": 1.0029625326749927,
+      "grad_norm": 0.49666348099708557,
+      "learning_rate": 0.0001,
+      "loss": 1.3315,
+      "step": 8633
+    },
+    {
+      "epoch": 1.0030787104269532,
+      "grad_norm": 0.5014142394065857,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 8634
+    },
+    {
+      "epoch": 1.0031948881789137,
+      "grad_norm": 0.48046401143074036,
+      "learning_rate": 0.0001,
+      "loss": 1.4761,
+      "step": 8635
+    },
+    {
+      "epoch": 1.0033110659308742,
+      "grad_norm": 0.6161479353904724,
+      "learning_rate": 0.0001,
+      "loss": 1.364,
+      "step": 8636
+    },
+    {
+      "epoch": 1.0034272436828346,
+      "grad_norm": 0.5767279267311096,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 8637
+    },
+    {
+      "epoch": 1.0035434214347951,
+      "grad_norm": 0.517012357711792,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 8638
+    },
+    {
+      "epoch": 1.0036595991867558,
+      "grad_norm": 0.5002815127372742,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 8639
+    },
+    {
+      "epoch": 1.0037757769387163,
+      "grad_norm": 0.5128475427627563,
+      "learning_rate": 0.0001,
+      "loss": 1.3022,
+      "step": 8640
+    },
+    {
+      "epoch": 1.0038919546906768,
+      "grad_norm": 0.5304717421531677,
+      "learning_rate": 0.0001,
+      "loss": 1.657,
+      "step": 8641
+    },
+    {
+      "epoch": 1.0040081324426373,
+      "grad_norm": 0.4929862320423126,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 8642
+    },
+    {
+      "epoch": 1.0041243101945978,
+      "grad_norm": 0.5075370669364929,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 8643
+    },
+    {
+      "epoch": 1.0042404879465583,
+      "grad_norm": 0.5167858004570007,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 8644
+    },
+    {
+      "epoch": 1.0043566656985188,
+      "grad_norm": 0.45497381687164307,
+      "learning_rate": 0.0001,
+      "loss": 1.2107,
+      "step": 8645
+    },
+    {
+      "epoch": 1.0044728434504793,
+      "grad_norm": 0.5639305710792542,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 8646
+    },
+    {
+      "epoch": 1.0045890212024398,
+      "grad_norm": 0.558202862739563,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 8647
+    },
+    {
+      "epoch": 1.0047051989544002,
+      "grad_norm": 0.5788266658782959,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 8648
+    },
+    {
+      "epoch": 1.0048213767063607,
+      "grad_norm": 0.5463119745254517,
+      "learning_rate": 0.0001,
+      "loss": 1.5891,
+      "step": 8649
+    },
+    {
+      "epoch": 1.0049375544583212,
+      "grad_norm": 0.5739423632621765,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 8650
+    },
+    {
+      "epoch": 1.0050537322102817,
+      "grad_norm": 0.5138026475906372,
+      "learning_rate": 0.0001,
+      "loss": 1.366,
+      "step": 8651
+    },
+    {
+      "epoch": 1.0051699099622422,
+      "grad_norm": 0.5081404447555542,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 8652
+    },
+    {
+      "epoch": 1.0052860877142027,
+      "grad_norm": 0.5575280785560608,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 8653
+    },
+    {
+      "epoch": 1.0054022654661632,
+      "grad_norm": 0.49208971858024597,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 8654
+    },
+    {
+      "epoch": 1.0055184432181237,
+      "grad_norm": 0.4909859001636505,
+      "learning_rate": 0.0001,
+      "loss": 1.3758,
+      "step": 8655
+    },
+    {
+      "epoch": 1.0056346209700842,
+      "grad_norm": 0.5197198987007141,
+      "learning_rate": 0.0001,
+      "loss": 1.6733,
+      "step": 8656
+    },
+    {
+      "epoch": 1.0057507987220446,
+      "grad_norm": 0.4799864888191223,
+      "learning_rate": 0.0001,
+      "loss": 1.4464,
+      "step": 8657
+    },
+    {
+      "epoch": 1.0058669764740051,
+      "grad_norm": 0.4831238389015198,
+      "learning_rate": 0.0001,
+      "loss": 1.3808,
+      "step": 8658
+    },
+    {
+      "epoch": 1.0059831542259656,
+      "grad_norm": 0.47417446970939636,
+      "learning_rate": 0.0001,
+      "loss": 1.3522,
+      "step": 8659
+    },
+    {
+      "epoch": 1.0060993319779263,
+      "grad_norm": 0.48708805441856384,
+      "learning_rate": 0.0001,
+      "loss": 1.4166,
+      "step": 8660
+    },
+    {
+      "epoch": 1.0062155097298868,
+      "grad_norm": 0.5076435208320618,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 8661
+    },
+    {
+      "epoch": 1.0063316874818473,
+      "grad_norm": 0.4805087745189667,
+      "learning_rate": 0.0001,
+      "loss": 1.3286,
+      "step": 8662
+    },
+    {
+      "epoch": 1.0064478652338078,
+      "grad_norm": 0.5486977696418762,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 8663
+    },
+    {
+      "epoch": 1.0065640429857683,
+      "grad_norm": 0.5629132390022278,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 8664
+    },
+    {
+      "epoch": 1.0066802207377288,
+      "grad_norm": 0.5157844424247742,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 8665
+    },
+    {
+      "epoch": 1.0067963984896893,
+      "grad_norm": 0.5176374316215515,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 8666
+    },
+    {
+      "epoch": 1.0069125762416498,
+      "grad_norm": 0.5113467574119568,
+      "learning_rate": 0.0001,
+      "loss": 1.3973,
+      "step": 8667
+    },
+    {
+      "epoch": 1.0070287539936102,
+      "grad_norm": 0.5078527331352234,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 8668
+    },
+    {
+      "epoch": 1.0071449317455707,
+      "grad_norm": 0.5093343257904053,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 8669
+    },
+    {
+      "epoch": 1.0072611094975312,
+      "grad_norm": 0.5552733540534973,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 8670
+    },
+    {
+      "epoch": 1.0073772872494917,
+      "grad_norm": 0.5340274572372437,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 8671
+    },
+    {
+      "epoch": 1.0074934650014522,
+      "grad_norm": 0.5053258538246155,
+      "learning_rate": 0.0001,
+      "loss": 1.3473,
+      "step": 8672
+    },
+    {
+      "epoch": 1.0076096427534127,
+      "grad_norm": 0.5026618838310242,
+      "learning_rate": 0.0001,
+      "loss": 1.3966,
+      "step": 8673
+    },
+    {
+      "epoch": 1.0077258205053732,
+      "grad_norm": 0.6316371560096741,
+      "learning_rate": 0.0001,
+      "loss": 1.6261,
+      "step": 8674
+    },
+    {
+      "epoch": 1.0078419982573337,
+      "grad_norm": 0.48417550325393677,
+      "learning_rate": 0.0001,
+      "loss": 1.3751,
+      "step": 8675
+    },
+    {
+      "epoch": 1.0079581760092942,
+      "grad_norm": 0.4940205514431,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 8676
+    },
+    {
+      "epoch": 1.0080743537612546,
+      "grad_norm": 0.5635389089584351,
+      "learning_rate": 0.0001,
+      "loss": 1.4426,
+      "step": 8677
+    },
+    {
+      "epoch": 1.0081905315132151,
+      "grad_norm": 0.5025110840797424,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 8678
+    },
+    {
+      "epoch": 1.0083067092651756,
+      "grad_norm": 0.5255551338195801,
+      "learning_rate": 0.0001,
+      "loss": 1.347,
+      "step": 8679
+    },
+    {
+      "epoch": 1.0084228870171361,
+      "grad_norm": 0.5457799434661865,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 8680
+    },
+    {
+      "epoch": 1.0085390647690968,
+      "grad_norm": 0.5311607122421265,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 8681
+    },
+    {
+      "epoch": 1.0086552425210573,
+      "grad_norm": 0.5294631719589233,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 8682
+    },
+    {
+      "epoch": 1.0087714202730178,
+      "grad_norm": 0.5186182856559753,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 8683
+    },
+    {
+      "epoch": 1.0088875980249783,
+      "grad_norm": 0.502144455909729,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 8684
+    },
+    {
+      "epoch": 1.0090037757769388,
+      "grad_norm": 0.49139168858528137,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 8685
+    },
+    {
+      "epoch": 1.0091199535288993,
+      "grad_norm": 0.5063096284866333,
+      "learning_rate": 0.0001,
+      "loss": 1.3168,
+      "step": 8686
+    },
+    {
+      "epoch": 1.0092361312808598,
+      "grad_norm": 0.539588212966919,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 8687
+    },
+    {
+      "epoch": 1.0093523090328202,
+      "grad_norm": 0.5028448104858398,
+      "learning_rate": 0.0001,
+      "loss": 1.3836,
+      "step": 8688
+    },
+    {
+      "epoch": 1.0094684867847807,
+      "grad_norm": 0.5191101431846619,
+      "learning_rate": 0.0001,
+      "loss": 1.4249,
+      "step": 8689
+    },
+    {
+      "epoch": 1.0095846645367412,
+      "grad_norm": 0.5024988055229187,
+      "learning_rate": 0.0001,
+      "loss": 1.3312,
+      "step": 8690
+    },
+    {
+      "epoch": 1.0097008422887017,
+      "grad_norm": 0.5392231941223145,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 8691
+    },
+    {
+      "epoch": 1.0098170200406622,
+      "grad_norm": 0.5703687071800232,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 8692
+    },
+    {
+      "epoch": 1.0099331977926227,
+      "grad_norm": 0.5280400514602661,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 8693
+    },
+    {
+      "epoch": 1.0100493755445832,
+      "grad_norm": 0.5094478726387024,
+      "learning_rate": 0.0001,
+      "loss": 1.4625,
+      "step": 8694
+    },
+    {
+      "epoch": 1.0101655532965437,
+      "grad_norm": 0.5534915328025818,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 8695
+    },
+    {
+      "epoch": 1.0102817310485042,
+      "grad_norm": 0.49775946140289307,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 8696
+    },
+    {
+      "epoch": 1.0103979088004647,
+      "grad_norm": 0.5414407253265381,
+      "learning_rate": 0.0001,
+      "loss": 1.3478,
+      "step": 8697
+    },
+    {
+      "epoch": 1.0105140865524251,
+      "grad_norm": 0.5187419652938843,
+      "learning_rate": 0.0001,
+      "loss": 1.3347,
+      "step": 8698
+    },
+    {
+      "epoch": 1.0106302643043856,
+      "grad_norm": 0.5660606026649475,
+      "learning_rate": 0.0001,
+      "loss": 1.4956,
+      "step": 8699
+    },
+    {
+      "epoch": 1.0107464420563461,
+      "grad_norm": 0.5465399026870728,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 8700
+    },
+    {
+      "epoch": 1.0108626198083066,
+      "grad_norm": 0.5141976475715637,
+      "learning_rate": 0.0001,
+      "loss": 1.3697,
+      "step": 8701
+    },
+    {
+      "epoch": 1.0109787975602673,
+      "grad_norm": 0.5536482930183411,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 8702
+    },
+    {
+      "epoch": 1.0110949753122278,
+      "grad_norm": 0.5275995135307312,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 8703
+    },
+    {
+      "epoch": 1.0112111530641883,
+      "grad_norm": 0.5400773286819458,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 8704
+    },
+    {
+      "epoch": 1.0113273308161488,
+      "grad_norm": 0.5452625751495361,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 8705
+    },
+    {
+      "epoch": 1.0114435085681093,
+      "grad_norm": 0.5686435699462891,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 8706
+    },
+    {
+      "epoch": 1.0115596863200698,
+      "grad_norm": 0.5052348971366882,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 8707
+    },
+    {
+      "epoch": 1.0116758640720303,
+      "grad_norm": 0.5007266998291016,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 8708
+    },
+    {
+      "epoch": 1.0117920418239907,
+      "grad_norm": 0.5304220914840698,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 8709
+    },
+    {
+      "epoch": 1.0119082195759512,
+      "grad_norm": 0.5678626298904419,
+      "learning_rate": 0.0001,
+      "loss": 1.5826,
+      "step": 8710
+    },
+    {
+      "epoch": 1.0120243973279117,
+      "grad_norm": 0.5126900672912598,
+      "learning_rate": 0.0001,
+      "loss": 1.3308,
+      "step": 8711
+    },
+    {
+      "epoch": 1.0121405750798722,
+      "grad_norm": 0.543702244758606,
+      "learning_rate": 0.0001,
+      "loss": 1.404,
+      "step": 8712
+    },
+    {
+      "epoch": 1.0122567528318327,
+      "grad_norm": 0.5045793056488037,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 8713
+    },
+    {
+      "epoch": 1.0123729305837932,
+      "grad_norm": 0.496380478143692,
+      "learning_rate": 0.0001,
+      "loss": 1.3823,
+      "step": 8714
+    },
+    {
+      "epoch": 1.0124891083357537,
+      "grad_norm": 0.5149445533752441,
+      "learning_rate": 0.0001,
+      "loss": 1.2453,
+      "step": 8715
+    },
+    {
+      "epoch": 1.0126052860877142,
+      "grad_norm": 0.5445921421051025,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 8716
+    },
+    {
+      "epoch": 1.0127214638396747,
+      "grad_norm": 0.536666214466095,
+      "learning_rate": 0.0001,
+      "loss": 1.6137,
+      "step": 8717
+    },
+    {
+      "epoch": 1.0128376415916351,
+      "grad_norm": 0.49998173117637634,
+      "learning_rate": 0.0001,
+      "loss": 1.4359,
+      "step": 8718
+    },
+    {
+      "epoch": 1.0129538193435956,
+      "grad_norm": 0.5400097966194153,
+      "learning_rate": 0.0001,
+      "loss": 1.5228,
+      "step": 8719
+    },
+    {
+      "epoch": 1.0130699970955561,
+      "grad_norm": 0.5498408079147339,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 8720
+    },
+    {
+      "epoch": 1.0131861748475166,
+      "grad_norm": 0.4792373478412628,
+      "learning_rate": 0.0001,
+      "loss": 1.3712,
+      "step": 8721
+    },
+    {
+      "epoch": 1.013302352599477,
+      "grad_norm": 0.49711140990257263,
+      "learning_rate": 0.0001,
+      "loss": 1.4074,
+      "step": 8722
+    },
+    {
+      "epoch": 1.0134185303514378,
+      "grad_norm": 0.5019938349723816,
+      "learning_rate": 0.0001,
+      "loss": 1.3793,
+      "step": 8723
+    },
+    {
+      "epoch": 1.0135347081033983,
+      "grad_norm": 0.4867321252822876,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 8724
+    },
+    {
+      "epoch": 1.0136508858553588,
+      "grad_norm": 0.5257205367088318,
+      "learning_rate": 0.0001,
+      "loss": 1.4509,
+      "step": 8725
+    },
+    {
+      "epoch": 1.0137670636073193,
+      "grad_norm": 0.5636011362075806,
+      "learning_rate": 0.0001,
+      "loss": 1.5382,
+      "step": 8726
+    },
+    {
+      "epoch": 1.0138832413592798,
+      "grad_norm": 0.5135155916213989,
+      "learning_rate": 0.0001,
+      "loss": 1.2232,
+      "step": 8727
+    },
+    {
+      "epoch": 1.0139994191112403,
+      "grad_norm": 0.524466872215271,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 8728
+    },
+    {
+      "epoch": 1.0141155968632007,
+      "grad_norm": 0.5000982284545898,
+      "learning_rate": 0.0001,
+      "loss": 1.3986,
+      "step": 8729
+    },
+    {
+      "epoch": 1.0142317746151612,
+      "grad_norm": 0.5312855243682861,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 8730
+    },
+    {
+      "epoch": 1.0143479523671217,
+      "grad_norm": 0.530696451663971,
+      "learning_rate": 0.0001,
+      "loss": 1.5603,
+      "step": 8731
+    },
+    {
+      "epoch": 1.0144641301190822,
+      "grad_norm": 0.5051412582397461,
+      "learning_rate": 0.0001,
+      "loss": 1.2977,
+      "step": 8732
+    },
+    {
+      "epoch": 1.0145803078710427,
+      "grad_norm": 0.5274531245231628,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 8733
+    },
+    {
+      "epoch": 1.0146964856230032,
+      "grad_norm": 0.5309851169586182,
+      "learning_rate": 0.0001,
+      "loss": 1.47,
+      "step": 8734
+    },
+    {
+      "epoch": 1.0148126633749637,
+      "grad_norm": 0.575518012046814,
+      "learning_rate": 0.0001,
+      "loss": 1.6739,
+      "step": 8735
+    },
+    {
+      "epoch": 1.0149288411269242,
+      "grad_norm": 0.5956768989562988,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 8736
+    },
+    {
+      "epoch": 1.0150450188788847,
+      "grad_norm": 0.532591700553894,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 8737
+    },
+    {
+      "epoch": 1.0151611966308451,
+      "grad_norm": 0.534724235534668,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 8738
+    },
+    {
+      "epoch": 1.0152773743828056,
+      "grad_norm": 0.5360231995582581,
+      "learning_rate": 0.0001,
+      "loss": 1.2831,
+      "step": 8739
+    },
+    {
+      "epoch": 1.0153935521347661,
+      "grad_norm": 0.557018518447876,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 8740
+    },
+    {
+      "epoch": 1.0155097298867266,
+      "grad_norm": 0.6012376546859741,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 8741
+    },
+    {
+      "epoch": 1.015625907638687,
+      "grad_norm": 0.4950985908508301,
+      "learning_rate": 0.0001,
+      "loss": 1.3743,
+      "step": 8742
+    },
+    {
+      "epoch": 1.0157420853906476,
+      "grad_norm": 0.5329170227050781,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 8743
+    },
+    {
+      "epoch": 1.0158582631426083,
+      "grad_norm": 0.5413834452629089,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 8744
+    },
+    {
+      "epoch": 1.0159744408945688,
+      "grad_norm": 0.5318317413330078,
+      "learning_rate": 0.0001,
+      "loss": 1.516,
+      "step": 8745
+    },
+    {
+      "epoch": 1.0160906186465293,
+      "grad_norm": 0.49791741371154785,
+      "learning_rate": 0.0001,
+      "loss": 1.3419,
+      "step": 8746
+    },
+    {
+      "epoch": 1.0162067963984898,
+      "grad_norm": 0.5165941119194031,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 8747
+    },
+    {
+      "epoch": 1.0163229741504503,
+      "grad_norm": 0.5537638068199158,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 8748
+    },
+    {
+      "epoch": 1.0164391519024107,
+      "grad_norm": 0.5026431083679199,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 8749
+    },
+    {
+      "epoch": 1.0165553296543712,
+      "grad_norm": 0.5347225069999695,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 8750
+    },
+    {
+      "epoch": 1.0166715074063317,
+      "grad_norm": 0.5551308393478394,
+      "learning_rate": 0.0001,
+      "loss": 1.5918,
+      "step": 8751
+    },
+    {
+      "epoch": 1.0167876851582922,
+      "grad_norm": 0.5152553915977478,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 8752
+    },
+    {
+      "epoch": 1.0169038629102527,
+      "grad_norm": 0.5854957699775696,
+      "learning_rate": 0.0001,
+      "loss": 1.7042,
+      "step": 8753
+    },
+    {
+      "epoch": 1.0170200406622132,
+      "grad_norm": 0.5672193169593811,
+      "learning_rate": 0.0001,
+      "loss": 1.6815,
+      "step": 8754
+    },
+    {
+      "epoch": 1.0171362184141737,
+      "grad_norm": 0.5420659184455872,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 8755
+    },
+    {
+      "epoch": 1.0172523961661342,
+      "grad_norm": 0.5565413236618042,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 8756
+    },
+    {
+      "epoch": 1.0173685739180947,
+      "grad_norm": 0.49962347745895386,
+      "learning_rate": 0.0001,
+      "loss": 1.4468,
+      "step": 8757
+    },
+    {
+      "epoch": 1.0174847516700551,
+      "grad_norm": 0.5187664031982422,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 8758
+    },
+    {
+      "epoch": 1.0176009294220156,
+      "grad_norm": 0.5551736354827881,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 8759
+    },
+    {
+      "epoch": 1.0177171071739761,
+      "grad_norm": 0.5217881202697754,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 8760
+    },
+    {
+      "epoch": 1.0178332849259366,
+      "grad_norm": 0.5211256742477417,
+      "learning_rate": 0.0001,
+      "loss": 1.4881,
+      "step": 8761
+    },
+    {
+      "epoch": 1.017949462677897,
+      "grad_norm": 0.5577038526535034,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 8762
+    },
+    {
+      "epoch": 1.0180656404298576,
+      "grad_norm": 0.5176113247871399,
+      "learning_rate": 0.0001,
+      "loss": 1.3198,
+      "step": 8763
+    },
+    {
+      "epoch": 1.018181818181818,
+      "grad_norm": 0.5276638865470886,
+      "learning_rate": 0.0001,
+      "loss": 1.4202,
+      "step": 8764
+    },
+    {
+      "epoch": 1.0182979959337788,
+      "grad_norm": 0.5093098282814026,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 8765
+    },
+    {
+      "epoch": 1.0184141736857393,
+      "grad_norm": 0.5525769591331482,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 8766
+    },
+    {
+      "epoch": 1.0185303514376998,
+      "grad_norm": 0.4998970329761505,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 8767
+    },
+    {
+      "epoch": 1.0186465291896603,
+      "grad_norm": 0.5910646915435791,
+      "learning_rate": 0.0001,
+      "loss": 1.7389,
+      "step": 8768
+    },
+    {
+      "epoch": 1.0187627069416207,
+      "grad_norm": 0.5479773879051208,
+      "learning_rate": 0.0001,
+      "loss": 1.6655,
+      "step": 8769
+    },
+    {
+      "epoch": 1.0188788846935812,
+      "grad_norm": 0.5090250968933105,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 8770
+    },
+    {
+      "epoch": 1.0189950624455417,
+      "grad_norm": 0.5310608148574829,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 8771
+    },
+    {
+      "epoch": 1.0191112401975022,
+      "grad_norm": 0.4970023036003113,
+      "learning_rate": 0.0001,
+      "loss": 1.3572,
+      "step": 8772
+    },
+    {
+      "epoch": 1.0192274179494627,
+      "grad_norm": 0.504270613193512,
+      "learning_rate": 0.0001,
+      "loss": 1.373,
+      "step": 8773
+    },
+    {
+      "epoch": 1.0193435957014232,
+      "grad_norm": 0.49610963463783264,
+      "learning_rate": 0.0001,
+      "loss": 1.3157,
+      "step": 8774
+    },
+    {
+      "epoch": 1.0194597734533837,
+      "grad_norm": 0.5968664884567261,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 8775
+    },
+    {
+      "epoch": 1.0195759512053442,
+      "grad_norm": 0.4930863082408905,
+      "learning_rate": 0.0001,
+      "loss": 1.4478,
+      "step": 8776
+    },
+    {
+      "epoch": 1.0196921289573047,
+      "grad_norm": 0.5297977924346924,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 8777
+    },
+    {
+      "epoch": 1.0198083067092651,
+      "grad_norm": 0.5384182333946228,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 8778
+    },
+    {
+      "epoch": 1.0199244844612256,
+      "grad_norm": 0.5289637446403503,
+      "learning_rate": 0.0001,
+      "loss": 1.3421,
+      "step": 8779
+    },
+    {
+      "epoch": 1.0200406622131861,
+      "grad_norm": 0.4995555877685547,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 8780
+    },
+    {
+      "epoch": 1.0201568399651466,
+      "grad_norm": 0.5270565748214722,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 8781
+    },
+    {
+      "epoch": 1.020273017717107,
+      "grad_norm": 0.5530638694763184,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 8782
+    },
+    {
+      "epoch": 1.0203891954690676,
+      "grad_norm": 0.519996702671051,
+      "learning_rate": 0.0001,
+      "loss": 1.4601,
+      "step": 8783
+    },
+    {
+      "epoch": 1.020505373221028,
+      "grad_norm": 0.5551260113716125,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 8784
+    },
+    {
+      "epoch": 1.0206215509729886,
+      "grad_norm": 0.5324636697769165,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 8785
+    },
+    {
+      "epoch": 1.0207377287249493,
+      "grad_norm": 0.5127756595611572,
+      "learning_rate": 0.0001,
+      "loss": 1.4244,
+      "step": 8786
+    },
+    {
+      "epoch": 1.0208539064769098,
+      "grad_norm": 0.5320945978164673,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 8787
+    },
+    {
+      "epoch": 1.0209700842288703,
+      "grad_norm": 0.5293431282043457,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 8788
+    },
+    {
+      "epoch": 1.0210862619808307,
+      "grad_norm": 0.5258355140686035,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 8789
+    },
+    {
+      "epoch": 1.0212024397327912,
+      "grad_norm": 0.6581287980079651,
+      "learning_rate": 0.0001,
+      "loss": 1.3883,
+      "step": 8790
+    },
+    {
+      "epoch": 1.0213186174847517,
+      "grad_norm": 0.5154805779457092,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 8791
+    },
+    {
+      "epoch": 1.0214347952367122,
+      "grad_norm": 0.48728328943252563,
+      "learning_rate": 0.0001,
+      "loss": 1.3606,
+      "step": 8792
+    },
+    {
+      "epoch": 1.0215509729886727,
+      "grad_norm": 0.5164970755577087,
+      "learning_rate": 0.0001,
+      "loss": 1.6262,
+      "step": 8793
+    },
+    {
+      "epoch": 1.0216671507406332,
+      "grad_norm": 0.5162052512168884,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 8794
+    },
+    {
+      "epoch": 1.0217833284925937,
+      "grad_norm": 0.5044278502464294,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 8795
+    },
+    {
+      "epoch": 1.0218995062445542,
+      "grad_norm": 0.5441913604736328,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 8796
+    },
+    {
+      "epoch": 1.0220156839965147,
+      "grad_norm": 0.4957961440086365,
+      "learning_rate": 0.0001,
+      "loss": 1.3308,
+      "step": 8797
+    },
+    {
+      "epoch": 1.0221318617484751,
+      "grad_norm": 0.49223870038986206,
+      "learning_rate": 0.0001,
+      "loss": 1.2865,
+      "step": 8798
+    },
+    {
+      "epoch": 1.0222480395004356,
+      "grad_norm": 0.5398341417312622,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 8799
+    },
+    {
+      "epoch": 1.0223642172523961,
+      "grad_norm": 0.5927699208259583,
+      "learning_rate": 0.0001,
+      "loss": 1.6615,
+      "step": 8800
+    },
+    {
+      "epoch": 1.0224803950043566,
+      "grad_norm": 0.5420244932174683,
+      "learning_rate": 0.0001,
+      "loss": 1.3955,
+      "step": 8801
+    },
+    {
+      "epoch": 1.022596572756317,
+      "grad_norm": 0.5336340069770813,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 8802
+    },
+    {
+      "epoch": 1.0227127505082776,
+      "grad_norm": 0.5515368580818176,
+      "learning_rate": 0.0001,
+      "loss": 1.4957,
+      "step": 8803
+    },
+    {
+      "epoch": 1.022828928260238,
+      "grad_norm": 0.5734601020812988,
+      "learning_rate": 0.0001,
+      "loss": 1.5024,
+      "step": 8804
+    },
+    {
+      "epoch": 1.0229451060121986,
+      "grad_norm": 0.4861997067928314,
+      "learning_rate": 0.0001,
+      "loss": 1.346,
+      "step": 8805
+    },
+    {
+      "epoch": 1.023061283764159,
+      "grad_norm": 0.501109778881073,
+      "learning_rate": 0.0001,
+      "loss": 1.3825,
+      "step": 8806
+    },
+    {
+      "epoch": 1.0231774615161198,
+      "grad_norm": 0.5470065474510193,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 8807
+    },
+    {
+      "epoch": 1.0232936392680803,
+      "grad_norm": 0.5535728335380554,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 8808
+    },
+    {
+      "epoch": 1.0234098170200407,
+      "grad_norm": 0.555833101272583,
+      "learning_rate": 0.0001,
+      "loss": 1.6016,
+      "step": 8809
+    },
+    {
+      "epoch": 1.0235259947720012,
+      "grad_norm": 0.48432549834251404,
+      "learning_rate": 0.0001,
+      "loss": 1.2699,
+      "step": 8810
+    },
+    {
+      "epoch": 1.0236421725239617,
+      "grad_norm": 0.5172594785690308,
+      "learning_rate": 0.0001,
+      "loss": 1.4708,
+      "step": 8811
+    },
+    {
+      "epoch": 1.0237583502759222,
+      "grad_norm": 0.5628485083580017,
+      "learning_rate": 0.0001,
+      "loss": 1.6034,
+      "step": 8812
+    },
+    {
+      "epoch": 1.0238745280278827,
+      "grad_norm": 0.6112015247344971,
+      "learning_rate": 0.0001,
+      "loss": 1.6817,
+      "step": 8813
+    },
+    {
+      "epoch": 1.0239907057798432,
+      "grad_norm": 0.5038037896156311,
+      "learning_rate": 0.0001,
+      "loss": 1.4484,
+      "step": 8814
+    },
+    {
+      "epoch": 1.0241068835318037,
+      "grad_norm": 0.5269456505775452,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 8815
+    },
+    {
+      "epoch": 1.0242230612837642,
+      "grad_norm": 0.48238101601600647,
+      "learning_rate": 0.0001,
+      "loss": 1.3186,
+      "step": 8816
+    },
+    {
+      "epoch": 1.0243392390357247,
+      "grad_norm": 0.5162169337272644,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 8817
+    },
+    {
+      "epoch": 1.0244554167876851,
+      "grad_norm": 0.5213837027549744,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 8818
+    },
+    {
+      "epoch": 1.0245715945396456,
+      "grad_norm": 0.5037201642990112,
+      "learning_rate": 0.0001,
+      "loss": 1.2898,
+      "step": 8819
+    },
+    {
+      "epoch": 1.0246877722916061,
+      "grad_norm": 0.517815113067627,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 8820
+    },
+    {
+      "epoch": 1.0248039500435666,
+      "grad_norm": 0.5647580027580261,
+      "learning_rate": 0.0001,
+      "loss": 1.5806,
+      "step": 8821
+    },
+    {
+      "epoch": 1.024920127795527,
+      "grad_norm": 0.549393355846405,
+      "learning_rate": 0.0001,
+      "loss": 1.5951,
+      "step": 8822
+    },
+    {
+      "epoch": 1.0250363055474876,
+      "grad_norm": 0.5423721075057983,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 8823
+    },
+    {
+      "epoch": 1.025152483299448,
+      "grad_norm": 0.5369595289230347,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 8824
+    },
+    {
+      "epoch": 1.0252686610514086,
+      "grad_norm": 0.6080834269523621,
+      "learning_rate": 0.0001,
+      "loss": 1.7438,
+      "step": 8825
+    },
+    {
+      "epoch": 1.025384838803369,
+      "grad_norm": 0.5140199065208435,
+      "learning_rate": 0.0001,
+      "loss": 1.3756,
+      "step": 8826
+    },
+    {
+      "epoch": 1.0255010165553295,
+      "grad_norm": 0.5014476776123047,
+      "learning_rate": 0.0001,
+      "loss": 1.5028,
+      "step": 8827
+    },
+    {
+      "epoch": 1.0256171943072903,
+      "grad_norm": 0.5367702841758728,
+      "learning_rate": 0.0001,
+      "loss": 1.3651,
+      "step": 8828
+    },
+    {
+      "epoch": 1.0257333720592507,
+      "grad_norm": 0.5167638659477234,
+      "learning_rate": 0.0001,
+      "loss": 1.329,
+      "step": 8829
+    },
+    {
+      "epoch": 1.0258495498112112,
+      "grad_norm": 0.5390440225601196,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 8830
+    },
+    {
+      "epoch": 1.0259657275631717,
+      "grad_norm": 0.5453746914863586,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 8831
+    },
+    {
+      "epoch": 1.0260819053151322,
+      "grad_norm": 0.5189757943153381,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 8832
+    },
+    {
+      "epoch": 1.0261980830670927,
+      "grad_norm": 0.5460232496261597,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 8833
+    },
+    {
+      "epoch": 1.0263142608190532,
+      "grad_norm": 0.4965575337409973,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 8834
+    },
+    {
+      "epoch": 1.0264304385710137,
+      "grad_norm": 0.5070030689239502,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 8835
+    },
+    {
+      "epoch": 1.0265466163229742,
+      "grad_norm": 0.5291348099708557,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 8836
+    },
+    {
+      "epoch": 1.0266627940749347,
+      "grad_norm": 0.49816176295280457,
+      "learning_rate": 0.0001,
+      "loss": 1.3712,
+      "step": 8837
+    },
+    {
+      "epoch": 1.0267789718268951,
+      "grad_norm": 0.5235051512718201,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 8838
+    },
+    {
+      "epoch": 1.0268951495788556,
+      "grad_norm": 0.5131754875183105,
+      "learning_rate": 0.0001,
+      "loss": 1.336,
+      "step": 8839
+    },
+    {
+      "epoch": 1.0270113273308161,
+      "grad_norm": 0.5183280110359192,
+      "learning_rate": 0.0001,
+      "loss": 1.4141,
+      "step": 8840
+    },
+    {
+      "epoch": 1.0271275050827766,
+      "grad_norm": 0.5118065476417542,
+      "learning_rate": 0.0001,
+      "loss": 1.4272,
+      "step": 8841
+    },
+    {
+      "epoch": 1.027243682834737,
+      "grad_norm": 0.5284550189971924,
+      "learning_rate": 0.0001,
+      "loss": 1.438,
+      "step": 8842
+    },
+    {
+      "epoch": 1.0273598605866976,
+      "grad_norm": 0.5649248957633972,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 8843
+    },
+    {
+      "epoch": 1.027476038338658,
+      "grad_norm": 0.5072284936904907,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 8844
+    },
+    {
+      "epoch": 1.0275922160906186,
+      "grad_norm": 0.5180877447128296,
+      "learning_rate": 0.0001,
+      "loss": 1.3545,
+      "step": 8845
+    },
+    {
+      "epoch": 1.027708393842579,
+      "grad_norm": 0.5194306373596191,
+      "learning_rate": 0.0001,
+      "loss": 1.4195,
+      "step": 8846
+    },
+    {
+      "epoch": 1.0278245715945395,
+      "grad_norm": 0.5667744278907776,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 8847
+    },
+    {
+      "epoch": 1.0279407493465,
+      "grad_norm": 0.5082692503929138,
+      "learning_rate": 0.0001,
+      "loss": 1.5024,
+      "step": 8848
+    },
+    {
+      "epoch": 1.0280569270984607,
+      "grad_norm": 0.6028507947921753,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 8849
+    },
+    {
+      "epoch": 1.0281731048504212,
+      "grad_norm": 0.5387246012687683,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 8850
+    },
+    {
+      "epoch": 1.0282892826023817,
+      "grad_norm": 0.4912740886211395,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 8851
+    },
+    {
+      "epoch": 1.0284054603543422,
+      "grad_norm": 0.5244132876396179,
+      "learning_rate": 0.0001,
+      "loss": 1.3191,
+      "step": 8852
+    },
+    {
+      "epoch": 1.0285216381063027,
+      "grad_norm": 0.5031388998031616,
+      "learning_rate": 0.0001,
+      "loss": 1.3729,
+      "step": 8853
+    },
+    {
+      "epoch": 1.0286378158582632,
+      "grad_norm": 0.544710636138916,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 8854
+    },
+    {
+      "epoch": 1.0287539936102237,
+      "grad_norm": 0.54068523645401,
+      "learning_rate": 0.0001,
+      "loss": 1.3927,
+      "step": 8855
+    },
+    {
+      "epoch": 1.0288701713621842,
+      "grad_norm": 0.5172038078308105,
+      "learning_rate": 0.0001,
+      "loss": 1.4375,
+      "step": 8856
+    },
+    {
+      "epoch": 1.0289863491141447,
+      "grad_norm": 0.49291446805000305,
+      "learning_rate": 0.0001,
+      "loss": 1.3361,
+      "step": 8857
+    },
+    {
+      "epoch": 1.0291025268661051,
+      "grad_norm": 0.5305606722831726,
+      "learning_rate": 0.0001,
+      "loss": 1.4012,
+      "step": 8858
+    },
+    {
+      "epoch": 1.0292187046180656,
+      "grad_norm": 0.4981054365634918,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 8859
+    },
+    {
+      "epoch": 1.0293348823700261,
+      "grad_norm": 0.510947585105896,
+      "learning_rate": 0.0001,
+      "loss": 1.3286,
+      "step": 8860
+    },
+    {
+      "epoch": 1.0294510601219866,
+      "grad_norm": 0.5207487940788269,
+      "learning_rate": 0.0001,
+      "loss": 1.4023,
+      "step": 8861
+    },
+    {
+      "epoch": 1.029567237873947,
+      "grad_norm": 0.5454730987548828,
+      "learning_rate": 0.0001,
+      "loss": 1.5136,
+      "step": 8862
+    },
+    {
+      "epoch": 1.0296834156259076,
+      "grad_norm": 0.5361663699150085,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 8863
+    },
+    {
+      "epoch": 1.029799593377868,
+      "grad_norm": 0.5152962803840637,
+      "learning_rate": 0.0001,
+      "loss": 1.3848,
+      "step": 8864
+    },
+    {
+      "epoch": 1.0299157711298286,
+      "grad_norm": 0.5426158905029297,
+      "learning_rate": 0.0001,
+      "loss": 1.2973,
+      "step": 8865
+    },
+    {
+      "epoch": 1.030031948881789,
+      "grad_norm": 0.5150105357170105,
+      "learning_rate": 0.0001,
+      "loss": 1.4212,
+      "step": 8866
+    },
+    {
+      "epoch": 1.0301481266337496,
+      "grad_norm": 0.5968256592750549,
+      "learning_rate": 0.0001,
+      "loss": 1.6619,
+      "step": 8867
+    },
+    {
+      "epoch": 1.03026430438571,
+      "grad_norm": 0.4991264343261719,
+      "learning_rate": 0.0001,
+      "loss": 1.3305,
+      "step": 8868
+    },
+    {
+      "epoch": 1.0303804821376705,
+      "grad_norm": 0.5225552916526794,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 8869
+    },
+    {
+      "epoch": 1.0304966598896312,
+      "grad_norm": 0.5236073732376099,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 8870
+    },
+    {
+      "epoch": 1.0306128376415917,
+      "grad_norm": 0.5573896169662476,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 8871
+    },
+    {
+      "epoch": 1.0307290153935522,
+      "grad_norm": 0.47595614194869995,
+      "learning_rate": 0.0001,
+      "loss": 1.312,
+      "step": 8872
+    },
+    {
+      "epoch": 1.0308451931455127,
+      "grad_norm": 0.5099783539772034,
+      "learning_rate": 0.0001,
+      "loss": 1.3819,
+      "step": 8873
+    },
+    {
+      "epoch": 1.0309613708974732,
+      "grad_norm": 0.5189265608787537,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 8874
+    },
+    {
+      "epoch": 1.0310775486494337,
+      "grad_norm": 0.567966103553772,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 8875
+    },
+    {
+      "epoch": 1.0311937264013942,
+      "grad_norm": 0.5577686429023743,
+      "learning_rate": 0.0001,
+      "loss": 1.674,
+      "step": 8876
+    },
+    {
+      "epoch": 1.0313099041533547,
+      "grad_norm": 0.5853254795074463,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 8877
+    },
+    {
+      "epoch": 1.0314260819053152,
+      "grad_norm": 0.5423609018325806,
+      "learning_rate": 0.0001,
+      "loss": 1.4577,
+      "step": 8878
+    },
+    {
+      "epoch": 1.0315422596572756,
+      "grad_norm": 0.5053103566169739,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 8879
+    },
+    {
+      "epoch": 1.0316584374092361,
+      "grad_norm": 0.5344913601875305,
+      "learning_rate": 0.0001,
+      "loss": 1.7323,
+      "step": 8880
+    },
+    {
+      "epoch": 1.0317746151611966,
+      "grad_norm": 0.5342625379562378,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 8881
+    },
+    {
+      "epoch": 1.031890792913157,
+      "grad_norm": 0.48883843421936035,
+      "learning_rate": 0.0001,
+      "loss": 1.2327,
+      "step": 8882
+    },
+    {
+      "epoch": 1.0320069706651176,
+      "grad_norm": 0.49293336272239685,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 8883
+    },
+    {
+      "epoch": 1.032123148417078,
+      "grad_norm": 0.4973764717578888,
+      "learning_rate": 0.0001,
+      "loss": 1.1467,
+      "step": 8884
+    },
+    {
+      "epoch": 1.0322393261690386,
+      "grad_norm": 0.49885573983192444,
+      "learning_rate": 0.0001,
+      "loss": 1.4539,
+      "step": 8885
+    },
+    {
+      "epoch": 1.032355503920999,
+      "grad_norm": 0.5181958079338074,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 8886
+    },
+    {
+      "epoch": 1.0324716816729596,
+      "grad_norm": 0.5104524493217468,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 8887
+    },
+    {
+      "epoch": 1.03258785942492,
+      "grad_norm": 0.5194922089576721,
+      "learning_rate": 0.0001,
+      "loss": 1.4931,
+      "step": 8888
+    },
+    {
+      "epoch": 1.0327040371768805,
+      "grad_norm": 0.5585922002792358,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 8889
+    },
+    {
+      "epoch": 1.032820214928841,
+      "grad_norm": 0.49765944480895996,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 8890
+    },
+    {
+      "epoch": 1.0329363926808017,
+      "grad_norm": 0.5335497260093689,
+      "learning_rate": 0.0001,
+      "loss": 1.3586,
+      "step": 8891
+    },
+    {
+      "epoch": 1.0330525704327622,
+      "grad_norm": 0.5823356509208679,
+      "learning_rate": 0.0001,
+      "loss": 1.4327,
+      "step": 8892
+    },
+    {
+      "epoch": 1.0331687481847227,
+      "grad_norm": 0.5257899165153503,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 8893
+    },
+    {
+      "epoch": 1.0332849259366832,
+      "grad_norm": 0.5384433269500732,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 8894
+    },
+    {
+      "epoch": 1.0334011036886437,
+      "grad_norm": 0.5588715672492981,
+      "learning_rate": 0.0001,
+      "loss": 1.3805,
+      "step": 8895
+    },
+    {
+      "epoch": 1.0335172814406042,
+      "grad_norm": 0.5330429673194885,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 8896
+    },
+    {
+      "epoch": 1.0336334591925647,
+      "grad_norm": 0.515548050403595,
+      "learning_rate": 0.0001,
+      "loss": 1.447,
+      "step": 8897
+    },
+    {
+      "epoch": 1.0337496369445252,
+      "grad_norm": 0.5171681046485901,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 8898
+    },
+    {
+      "epoch": 1.0338658146964856,
+      "grad_norm": 0.5828492045402527,
+      "learning_rate": 0.0001,
+      "loss": 1.6547,
+      "step": 8899
+    },
+    {
+      "epoch": 1.0339819924484461,
+      "grad_norm": 0.4846728444099426,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 8900
+    },
+    {
+      "epoch": 1.0340981702004066,
+      "grad_norm": 0.5182422995567322,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 8901
+    },
+    {
+      "epoch": 1.034214347952367,
+      "grad_norm": 0.5431578755378723,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 8902
+    },
+    {
+      "epoch": 1.0343305257043276,
+      "grad_norm": 0.5584523677825928,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 8903
+    },
+    {
+      "epoch": 1.034446703456288,
+      "grad_norm": 0.5252351760864258,
+      "learning_rate": 0.0001,
+      "loss": 1.269,
+      "step": 8904
+    },
+    {
+      "epoch": 1.0345628812082486,
+      "grad_norm": 0.575616717338562,
+      "learning_rate": 0.0001,
+      "loss": 1.6217,
+      "step": 8905
+    },
+    {
+      "epoch": 1.034679058960209,
+      "grad_norm": 0.6257792115211487,
+      "learning_rate": 0.0001,
+      "loss": 1.7119,
+      "step": 8906
+    },
+    {
+      "epoch": 1.0347952367121696,
+      "grad_norm": 0.532507598400116,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 8907
+    },
+    {
+      "epoch": 1.03491141446413,
+      "grad_norm": 0.5430809855461121,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 8908
+    },
+    {
+      "epoch": 1.0350275922160905,
+      "grad_norm": 0.5150136351585388,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 8909
+    },
+    {
+      "epoch": 1.035143769968051,
+      "grad_norm": 0.5215170383453369,
+      "learning_rate": 0.0001,
+      "loss": 1.4457,
+      "step": 8910
+    },
+    {
+      "epoch": 1.0352599477200117,
+      "grad_norm": 0.5841111540794373,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 8911
+    },
+    {
+      "epoch": 1.0353761254719722,
+      "grad_norm": 0.5070576667785645,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 8912
+    },
+    {
+      "epoch": 1.0354923032239327,
+      "grad_norm": 0.5849988460540771,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 8913
+    },
+    {
+      "epoch": 1.0356084809758932,
+      "grad_norm": 0.5653141736984253,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 8914
+    },
+    {
+      "epoch": 1.0357246587278537,
+      "grad_norm": 0.5517808198928833,
+      "learning_rate": 0.0001,
+      "loss": 1.5693,
+      "step": 8915
+    },
+    {
+      "epoch": 1.0358408364798142,
+      "grad_norm": 0.5292582511901855,
+      "learning_rate": 0.0001,
+      "loss": 1.3979,
+      "step": 8916
+    },
+    {
+      "epoch": 1.0359570142317747,
+      "grad_norm": 0.5182827711105347,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 8917
+    },
+    {
+      "epoch": 1.0360731919837352,
+      "grad_norm": 0.5552460551261902,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 8918
+    },
+    {
+      "epoch": 1.0361893697356956,
+      "grad_norm": 0.5514026880264282,
+      "learning_rate": 0.0001,
+      "loss": 1.3159,
+      "step": 8919
+    },
+    {
+      "epoch": 1.0363055474876561,
+      "grad_norm": 0.5646169185638428,
+      "learning_rate": 0.0001,
+      "loss": 1.4833,
+      "step": 8920
+    },
+    {
+      "epoch": 1.0364217252396166,
+      "grad_norm": 0.5406813621520996,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 8921
+    },
+    {
+      "epoch": 1.036537902991577,
+      "grad_norm": 0.5190636515617371,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 8922
+    },
+    {
+      "epoch": 1.0366540807435376,
+      "grad_norm": 0.538845956325531,
+      "learning_rate": 0.0001,
+      "loss": 1.6992,
+      "step": 8923
+    },
+    {
+      "epoch": 1.036770258495498,
+      "grad_norm": 0.5209506154060364,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 8924
+    },
+    {
+      "epoch": 1.0368864362474586,
+      "grad_norm": 0.5400375127792358,
+      "learning_rate": 0.0001,
+      "loss": 1.332,
+      "step": 8925
+    },
+    {
+      "epoch": 1.037002613999419,
+      "grad_norm": 0.5286649465560913,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 8926
+    },
+    {
+      "epoch": 1.0371187917513796,
+      "grad_norm": 0.5743568539619446,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 8927
+    },
+    {
+      "epoch": 1.03723496950334,
+      "grad_norm": 0.5251564979553223,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 8928
+    },
+    {
+      "epoch": 1.0373511472553005,
+      "grad_norm": 0.5093726515769958,
+      "learning_rate": 0.0001,
+      "loss": 1.4133,
+      "step": 8929
+    },
+    {
+      "epoch": 1.037467325007261,
+      "grad_norm": 0.5467512011528015,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 8930
+    },
+    {
+      "epoch": 1.0375835027592215,
+      "grad_norm": 0.5583242774009705,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 8931
+    },
+    {
+      "epoch": 1.037699680511182,
+      "grad_norm": 0.5613623857498169,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 8932
+    },
+    {
+      "epoch": 1.0378158582631427,
+      "grad_norm": 0.5468906760215759,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 8933
+    },
+    {
+      "epoch": 1.0379320360151032,
+      "grad_norm": 0.5442972183227539,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 8934
+    },
+    {
+      "epoch": 1.0380482137670637,
+      "grad_norm": 0.5066025853157043,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 8935
+    },
+    {
+      "epoch": 1.0381643915190242,
+      "grad_norm": 0.548570990562439,
+      "learning_rate": 0.0001,
+      "loss": 1.5706,
+      "step": 8936
+    },
+    {
+      "epoch": 1.0382805692709847,
+      "grad_norm": 0.5386464595794678,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 8937
+    },
+    {
+      "epoch": 1.0383967470229452,
+      "grad_norm": 0.6019245386123657,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 8938
+    },
+    {
+      "epoch": 1.0385129247749056,
+      "grad_norm": 0.5137844085693359,
+      "learning_rate": 0.0001,
+      "loss": 1.3028,
+      "step": 8939
+    },
+    {
+      "epoch": 1.0386291025268661,
+      "grad_norm": 0.5185123085975647,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 8940
+    },
+    {
+      "epoch": 1.0387452802788266,
+      "grad_norm": 0.5555708408355713,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 8941
+    },
+    {
+      "epoch": 1.038861458030787,
+      "grad_norm": 0.49165698885917664,
+      "learning_rate": 0.0001,
+      "loss": 1.3102,
+      "step": 8942
+    },
+    {
+      "epoch": 1.0389776357827476,
+      "grad_norm": 0.568077564239502,
+      "learning_rate": 0.0001,
+      "loss": 1.6435,
+      "step": 8943
+    },
+    {
+      "epoch": 1.039093813534708,
+      "grad_norm": 0.5384320020675659,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 8944
+    },
+    {
+      "epoch": 1.0392099912866686,
+      "grad_norm": 0.5384372472763062,
+      "learning_rate": 0.0001,
+      "loss": 1.3606,
+      "step": 8945
+    },
+    {
+      "epoch": 1.039326169038629,
+      "grad_norm": 0.5323529839515686,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 8946
+    },
+    {
+      "epoch": 1.0394423467905896,
+      "grad_norm": 0.5654020309448242,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 8947
+    },
+    {
+      "epoch": 1.03955852454255,
+      "grad_norm": 0.5455036163330078,
+      "learning_rate": 0.0001,
+      "loss": 1.2164,
+      "step": 8948
+    },
+    {
+      "epoch": 1.0396747022945105,
+      "grad_norm": 0.5437000393867493,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 8949
+    },
+    {
+      "epoch": 1.039790880046471,
+      "grad_norm": 0.5596999526023865,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 8950
+    },
+    {
+      "epoch": 1.0399070577984315,
+      "grad_norm": 0.5869429707527161,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 8951
+    },
+    {
+      "epoch": 1.040023235550392,
+      "grad_norm": 0.5478681921958923,
+      "learning_rate": 0.0001,
+      "loss": 1.34,
+      "step": 8952
+    },
+    {
+      "epoch": 1.0401394133023527,
+      "grad_norm": 0.5384049415588379,
+      "learning_rate": 0.0001,
+      "loss": 1.4315,
+      "step": 8953
+    },
+    {
+      "epoch": 1.0402555910543132,
+      "grad_norm": 0.5583698749542236,
+      "learning_rate": 0.0001,
+      "loss": 1.5661,
+      "step": 8954
+    },
+    {
+      "epoch": 1.0403717688062737,
+      "grad_norm": 0.5423256754875183,
+      "learning_rate": 0.0001,
+      "loss": 1.4291,
+      "step": 8955
+    },
+    {
+      "epoch": 1.0404879465582342,
+      "grad_norm": 0.5298671722412109,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 8956
+    },
+    {
+      "epoch": 1.0406041243101947,
+      "grad_norm": 0.5103389024734497,
+      "learning_rate": 0.0001,
+      "loss": 1.2942,
+      "step": 8957
+    },
+    {
+      "epoch": 1.0407203020621552,
+      "grad_norm": 0.5083001852035522,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 8958
+    },
+    {
+      "epoch": 1.0408364798141156,
+      "grad_norm": 0.5489501953125,
+      "learning_rate": 0.0001,
+      "loss": 1.5624,
+      "step": 8959
+    },
+    {
+      "epoch": 1.0409526575660761,
+      "grad_norm": 0.5339622497558594,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 8960
+    },
+    {
+      "epoch": 1.0410688353180366,
+      "grad_norm": 0.5265421271324158,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 8961
+    },
+    {
+      "epoch": 1.0411850130699971,
+      "grad_norm": 0.539472222328186,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 8962
+    },
+    {
+      "epoch": 1.0413011908219576,
+      "grad_norm": 0.5840896964073181,
+      "learning_rate": 0.0001,
+      "loss": 1.4922,
+      "step": 8963
+    },
+    {
+      "epoch": 1.041417368573918,
+      "grad_norm": 0.5356101393699646,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 8964
+    },
+    {
+      "epoch": 1.0415335463258786,
+      "grad_norm": 0.5115817785263062,
+      "learning_rate": 0.0001,
+      "loss": 1.4016,
+      "step": 8965
+    },
+    {
+      "epoch": 1.041649724077839,
+      "grad_norm": 0.5461002588272095,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 8966
+    },
+    {
+      "epoch": 1.0417659018297996,
+      "grad_norm": 0.515606701374054,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 8967
+    },
+    {
+      "epoch": 1.04188207958176,
+      "grad_norm": 0.5375041365623474,
+      "learning_rate": 0.0001,
+      "loss": 1.6334,
+      "step": 8968
+    },
+    {
+      "epoch": 1.0419982573337205,
+      "grad_norm": 0.5649837851524353,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 8969
+    },
+    {
+      "epoch": 1.042114435085681,
+      "grad_norm": 0.5564467310905457,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 8970
+    },
+    {
+      "epoch": 1.0422306128376415,
+      "grad_norm": 0.5804434418678284,
+      "learning_rate": 0.0001,
+      "loss": 1.6147,
+      "step": 8971
+    },
+    {
+      "epoch": 1.042346790589602,
+      "grad_norm": 0.5008756518363953,
+      "learning_rate": 0.0001,
+      "loss": 1.3786,
+      "step": 8972
+    },
+    {
+      "epoch": 1.0424629683415625,
+      "grad_norm": 0.4874401092529297,
+      "learning_rate": 0.0001,
+      "loss": 1.4006,
+      "step": 8973
+    },
+    {
+      "epoch": 1.042579146093523,
+      "grad_norm": 0.5213797688484192,
+      "learning_rate": 0.0001,
+      "loss": 1.5882,
+      "step": 8974
+    },
+    {
+      "epoch": 1.0426953238454837,
+      "grad_norm": 0.5059088468551636,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 8975
+    },
+    {
+      "epoch": 1.0428115015974442,
+      "grad_norm": 0.5125690698623657,
+      "learning_rate": 0.0001,
+      "loss": 1.3756,
+      "step": 8976
+    },
+    {
+      "epoch": 1.0429276793494047,
+      "grad_norm": 0.4973143935203552,
+      "learning_rate": 0.0001,
+      "loss": 1.4351,
+      "step": 8977
+    },
+    {
+      "epoch": 1.0430438571013652,
+      "grad_norm": 0.5088719129562378,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 8978
+    },
+    {
+      "epoch": 1.0431600348533256,
+      "grad_norm": 0.5613342523574829,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 8979
+    },
+    {
+      "epoch": 1.0432762126052861,
+      "grad_norm": 0.5364037752151489,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 8980
+    },
+    {
+      "epoch": 1.0433923903572466,
+      "grad_norm": 0.5083293914794922,
+      "learning_rate": 0.0001,
+      "loss": 1.4213,
+      "step": 8981
+    },
+    {
+      "epoch": 1.0435085681092071,
+      "grad_norm": 0.5561593174934387,
+      "learning_rate": 0.0001,
+      "loss": 1.4165,
+      "step": 8982
+    },
+    {
+      "epoch": 1.0436247458611676,
+      "grad_norm": 0.49919378757476807,
+      "learning_rate": 0.0001,
+      "loss": 1.4196,
+      "step": 8983
+    },
+    {
+      "epoch": 1.043740923613128,
+      "grad_norm": 0.5361339449882507,
+      "learning_rate": 0.0001,
+      "loss": 1.3825,
+      "step": 8984
+    },
+    {
+      "epoch": 1.0438571013650886,
+      "grad_norm": 0.5248799920082092,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 8985
+    },
+    {
+      "epoch": 1.043973279117049,
+      "grad_norm": 0.5760166049003601,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 8986
+    },
+    {
+      "epoch": 1.0440894568690096,
+      "grad_norm": 0.580518364906311,
+      "learning_rate": 0.0001,
+      "loss": 1.5113,
+      "step": 8987
+    },
+    {
+      "epoch": 1.04420563462097,
+      "grad_norm": 0.5657472610473633,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 8988
+    },
+    {
+      "epoch": 1.0443218123729305,
+      "grad_norm": 0.5147629380226135,
+      "learning_rate": 0.0001,
+      "loss": 1.3792,
+      "step": 8989
+    },
+    {
+      "epoch": 1.044437990124891,
+      "grad_norm": 0.5364604592323303,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 8990
+    },
+    {
+      "epoch": 1.0445541678768515,
+      "grad_norm": 0.4889623820781708,
+      "learning_rate": 0.0001,
+      "loss": 1.3463,
+      "step": 8991
+    },
+    {
+      "epoch": 1.044670345628812,
+      "grad_norm": 0.5112910270690918,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 8992
+    },
+    {
+      "epoch": 1.0447865233807725,
+      "grad_norm": 0.589764416217804,
+      "learning_rate": 0.0001,
+      "loss": 1.5348,
+      "step": 8993
+    },
+    {
+      "epoch": 1.044902701132733,
+      "grad_norm": 0.5439825057983398,
+      "learning_rate": 0.0001,
+      "loss": 1.5048,
+      "step": 8994
+    },
+    {
+      "epoch": 1.0450188788846937,
+      "grad_norm": 0.5245374441146851,
+      "learning_rate": 0.0001,
+      "loss": 1.4242,
+      "step": 8995
+    },
+    {
+      "epoch": 1.0451350566366542,
+      "grad_norm": 0.5435076951980591,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 8996
+    },
+    {
+      "epoch": 1.0452512343886147,
+      "grad_norm": 0.6293114423751831,
+      "learning_rate": 0.0001,
+      "loss": 1.6639,
+      "step": 8997
+    },
+    {
+      "epoch": 1.0453674121405752,
+      "grad_norm": 0.5348904728889465,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 8998
+    },
+    {
+      "epoch": 1.0454835898925356,
+      "grad_norm": 0.5541291832923889,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 8999
+    },
+    {
+      "epoch": 1.0455997676444961,
+      "grad_norm": 0.6162405610084534,
+      "learning_rate": 0.0001,
+      "loss": 1.8332,
+      "step": 9000
+    },
+    {
+      "epoch": 1.0457159453964566,
+      "grad_norm": 0.5342404842376709,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 9001
+    },
+    {
+      "epoch": 1.0458321231484171,
+      "grad_norm": 0.5075395107269287,
+      "learning_rate": 0.0001,
+      "loss": 1.2412,
+      "step": 9002
+    },
+    {
+      "epoch": 1.0459483009003776,
+      "grad_norm": 0.529941737651825,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 9003
+    },
+    {
+      "epoch": 1.046064478652338,
+      "grad_norm": 0.5113204717636108,
+      "learning_rate": 0.0001,
+      "loss": 1.4907,
+      "step": 9004
+    },
+    {
+      "epoch": 1.0461806564042986,
+      "grad_norm": 0.5787484049797058,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 9005
+    },
+    {
+      "epoch": 1.046296834156259,
+      "grad_norm": 0.5223624110221863,
+      "learning_rate": 0.0001,
+      "loss": 1.4266,
+      "step": 9006
+    },
+    {
+      "epoch": 1.0464130119082196,
+      "grad_norm": 0.5437747836112976,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 9007
+    },
+    {
+      "epoch": 1.04652918966018,
+      "grad_norm": 0.5204196572303772,
+      "learning_rate": 0.0001,
+      "loss": 1.43,
+      "step": 9008
+    },
+    {
+      "epoch": 1.0466453674121405,
+      "grad_norm": 0.5550678372383118,
+      "learning_rate": 0.0001,
+      "loss": 1.5454,
+      "step": 9009
+    },
+    {
+      "epoch": 1.046761545164101,
+      "grad_norm": 0.5623777508735657,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 9010
+    },
+    {
+      "epoch": 1.0468777229160615,
+      "grad_norm": 0.5182955861091614,
+      "learning_rate": 0.0001,
+      "loss": 1.5477,
+      "step": 9011
+    },
+    {
+      "epoch": 1.046993900668022,
+      "grad_norm": 0.5236296653747559,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 9012
+    },
+    {
+      "epoch": 1.0471100784199825,
+      "grad_norm": 0.5745497345924377,
+      "learning_rate": 0.0001,
+      "loss": 1.6299,
+      "step": 9013
+    },
+    {
+      "epoch": 1.047226256171943,
+      "grad_norm": 0.5346052646636963,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 9014
+    },
+    {
+      "epoch": 1.0473424339239035,
+      "grad_norm": 0.523341178894043,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 9015
+    },
+    {
+      "epoch": 1.047458611675864,
+      "grad_norm": 0.5185964703559875,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 9016
+    },
+    {
+      "epoch": 1.0475747894278247,
+      "grad_norm": 0.5521393418312073,
+      "learning_rate": 0.0001,
+      "loss": 1.7436,
+      "step": 9017
+    },
+    {
+      "epoch": 1.0476909671797852,
+      "grad_norm": 0.566999077796936,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 9018
+    },
+    {
+      "epoch": 1.0478071449317456,
+      "grad_norm": 0.5497294068336487,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 9019
+    },
+    {
+      "epoch": 1.0479233226837061,
+      "grad_norm": 0.5116751194000244,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 9020
+    },
+    {
+      "epoch": 1.0480395004356666,
+      "grad_norm": 0.5744738578796387,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 9021
+    },
+    {
+      "epoch": 1.0481556781876271,
+      "grad_norm": 0.4972259998321533,
+      "learning_rate": 0.0001,
+      "loss": 1.3011,
+      "step": 9022
+    },
+    {
+      "epoch": 1.0482718559395876,
+      "grad_norm": 0.5275663137435913,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 9023
+    },
+    {
+      "epoch": 1.048388033691548,
+      "grad_norm": 0.4991401135921478,
+      "learning_rate": 0.0001,
+      "loss": 1.2919,
+      "step": 9024
+    },
+    {
+      "epoch": 1.0485042114435086,
+      "grad_norm": 0.507801353931427,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 9025
+    },
+    {
+      "epoch": 1.048620389195469,
+      "grad_norm": 0.528578519821167,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 9026
+    },
+    {
+      "epoch": 1.0487365669474296,
+      "grad_norm": 0.5243181586265564,
+      "learning_rate": 0.0001,
+      "loss": 1.466,
+      "step": 9027
+    },
+    {
+      "epoch": 1.04885274469939,
+      "grad_norm": 0.5625602602958679,
+      "learning_rate": 0.0001,
+      "loss": 1.4427,
+      "step": 9028
+    },
+    {
+      "epoch": 1.0489689224513505,
+      "grad_norm": 0.539786159992218,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 9029
+    },
+    {
+      "epoch": 1.049085100203311,
+      "grad_norm": 0.5501077175140381,
+      "learning_rate": 0.0001,
+      "loss": 1.269,
+      "step": 9030
+    },
+    {
+      "epoch": 1.0492012779552715,
+      "grad_norm": 0.5288931131362915,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 9031
+    },
+    {
+      "epoch": 1.049317455707232,
+      "grad_norm": 0.5670494437217712,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 9032
+    },
+    {
+      "epoch": 1.0494336334591925,
+      "grad_norm": 0.5463694930076599,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 9033
+    },
+    {
+      "epoch": 1.049549811211153,
+      "grad_norm": 0.5625814199447632,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 9034
+    },
+    {
+      "epoch": 1.0496659889631135,
+      "grad_norm": 0.5289028286933899,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 9035
+    },
+    {
+      "epoch": 1.049782166715074,
+      "grad_norm": 0.5531147122383118,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 9036
+    },
+    {
+      "epoch": 1.0498983444670347,
+      "grad_norm": 0.5495567917823792,
+      "learning_rate": 0.0001,
+      "loss": 1.3533,
+      "step": 9037
+    },
+    {
+      "epoch": 1.0500145222189952,
+      "grad_norm": 0.5757434368133545,
+      "learning_rate": 0.0001,
+      "loss": 1.6329,
+      "step": 9038
+    },
+    {
+      "epoch": 1.0501306999709557,
+      "grad_norm": 0.5772126317024231,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 9039
+    },
+    {
+      "epoch": 1.0502468777229161,
+      "grad_norm": 0.5771605968475342,
+      "learning_rate": 0.0001,
+      "loss": 1.3436,
+      "step": 9040
+    },
+    {
+      "epoch": 1.0503630554748766,
+      "grad_norm": 0.5607427358627319,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 9041
+    },
+    {
+      "epoch": 1.0504792332268371,
+      "grad_norm": 0.5379540920257568,
+      "learning_rate": 0.0001,
+      "loss": 1.4105,
+      "step": 9042
+    },
+    {
+      "epoch": 1.0505954109787976,
+      "grad_norm": 0.5043210387229919,
+      "learning_rate": 0.0001,
+      "loss": 1.4431,
+      "step": 9043
+    },
+    {
+      "epoch": 1.050711588730758,
+      "grad_norm": 0.5489790439605713,
+      "learning_rate": 0.0001,
+      "loss": 1.4465,
+      "step": 9044
+    },
+    {
+      "epoch": 1.0508277664827186,
+      "grad_norm": 0.5171908736228943,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 9045
+    },
+    {
+      "epoch": 1.050943944234679,
+      "grad_norm": 0.5216082334518433,
+      "learning_rate": 0.0001,
+      "loss": 1.3013,
+      "step": 9046
+    },
+    {
+      "epoch": 1.0510601219866396,
+      "grad_norm": 0.5552801489830017,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 9047
+    },
+    {
+      "epoch": 1.0511762997386,
+      "grad_norm": 0.5355478525161743,
+      "learning_rate": 0.0001,
+      "loss": 1.3406,
+      "step": 9048
+    },
+    {
+      "epoch": 1.0512924774905605,
+      "grad_norm": 0.5976464748382568,
+      "learning_rate": 0.0001,
+      "loss": 1.6491,
+      "step": 9049
+    },
+    {
+      "epoch": 1.051408655242521,
+      "grad_norm": 0.5584911704063416,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 9050
+    },
+    {
+      "epoch": 1.0515248329944815,
+      "grad_norm": 0.5827659964561462,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 9051
+    },
+    {
+      "epoch": 1.051641010746442,
+      "grad_norm": 0.5411894917488098,
+      "learning_rate": 0.0001,
+      "loss": 1.397,
+      "step": 9052
+    },
+    {
+      "epoch": 1.0517571884984025,
+      "grad_norm": 0.507165789604187,
+      "learning_rate": 0.0001,
+      "loss": 1.3847,
+      "step": 9053
+    },
+    {
+      "epoch": 1.051873366250363,
+      "grad_norm": 0.5528204441070557,
+      "learning_rate": 0.0001,
+      "loss": 1.3638,
+      "step": 9054
+    },
+    {
+      "epoch": 1.0519895440023235,
+      "grad_norm": 0.5597609877586365,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 9055
+    },
+    {
+      "epoch": 1.052105721754284,
+      "grad_norm": 0.5310388207435608,
+      "learning_rate": 0.0001,
+      "loss": 1.4972,
+      "step": 9056
+    },
+    {
+      "epoch": 1.0522218995062445,
+      "grad_norm": 0.5567108392715454,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 9057
+    },
+    {
+      "epoch": 1.052338077258205,
+      "grad_norm": 0.539630651473999,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 9058
+    },
+    {
+      "epoch": 1.0524542550101657,
+      "grad_norm": 0.5464910864830017,
+      "learning_rate": 0.0001,
+      "loss": 1.4159,
+      "step": 9059
+    },
+    {
+      "epoch": 1.0525704327621261,
+      "grad_norm": 0.5596455931663513,
+      "learning_rate": 0.0001,
+      "loss": 1.4302,
+      "step": 9060
+    },
+    {
+      "epoch": 1.0526866105140866,
+      "grad_norm": 0.5715380311012268,
+      "learning_rate": 0.0001,
+      "loss": 1.3316,
+      "step": 9061
+    },
+    {
+      "epoch": 1.0528027882660471,
+      "grad_norm": 0.5546104907989502,
+      "learning_rate": 0.0001,
+      "loss": 1.3358,
+      "step": 9062
+    },
+    {
+      "epoch": 1.0529189660180076,
+      "grad_norm": 0.7018982768058777,
+      "learning_rate": 0.0001,
+      "loss": 1.3523,
+      "step": 9063
+    },
+    {
+      "epoch": 1.053035143769968,
+      "grad_norm": 0.5631259679794312,
+      "learning_rate": 0.0001,
+      "loss": 1.3676,
+      "step": 9064
+    },
+    {
+      "epoch": 1.0531513215219286,
+      "grad_norm": 0.5323575735092163,
+      "learning_rate": 0.0001,
+      "loss": 1.3987,
+      "step": 9065
+    },
+    {
+      "epoch": 1.053267499273889,
+      "grad_norm": 0.5142088532447815,
+      "learning_rate": 0.0001,
+      "loss": 1.4036,
+      "step": 9066
+    },
+    {
+      "epoch": 1.0533836770258496,
+      "grad_norm": 0.5592623353004456,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 9067
+    },
+    {
+      "epoch": 1.05349985477781,
+      "grad_norm": 0.5052597522735596,
+      "learning_rate": 0.0001,
+      "loss": 1.4116,
+      "step": 9068
+    },
+    {
+      "epoch": 1.0536160325297705,
+      "grad_norm": 0.517086923122406,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 9069
+    },
+    {
+      "epoch": 1.053732210281731,
+      "grad_norm": 0.5268409252166748,
+      "learning_rate": 0.0001,
+      "loss": 1.4975,
+      "step": 9070
+    },
+    {
+      "epoch": 1.0538483880336915,
+      "grad_norm": 0.536361038684845,
+      "learning_rate": 0.0001,
+      "loss": 1.5102,
+      "step": 9071
+    },
+    {
+      "epoch": 1.053964565785652,
+      "grad_norm": 0.5307349562644958,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 9072
+    },
+    {
+      "epoch": 1.0540807435376125,
+      "grad_norm": 0.5793299078941345,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 9073
+    },
+    {
+      "epoch": 1.054196921289573,
+      "grad_norm": 0.5216554999351501,
+      "learning_rate": 0.0001,
+      "loss": 1.4039,
+      "step": 9074
+    },
+    {
+      "epoch": 1.0543130990415335,
+      "grad_norm": 0.49403879046440125,
+      "learning_rate": 0.0001,
+      "loss": 1.2771,
+      "step": 9075
+    },
+    {
+      "epoch": 1.054429276793494,
+      "grad_norm": 0.5500255823135376,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 9076
+    },
+    {
+      "epoch": 1.0545454545454545,
+      "grad_norm": 0.5305203795433044,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 9077
+    },
+    {
+      "epoch": 1.054661632297415,
+      "grad_norm": 0.5569433569908142,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 9078
+    },
+    {
+      "epoch": 1.0547778100493757,
+      "grad_norm": 0.4997643232345581,
+      "learning_rate": 0.0001,
+      "loss": 1.3793,
+      "step": 9079
+    },
+    {
+      "epoch": 1.0548939878013361,
+      "grad_norm": 0.5268417596817017,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 9080
+    },
+    {
+      "epoch": 1.0550101655532966,
+      "grad_norm": 0.5548385381698608,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 9081
+    },
+    {
+      "epoch": 1.0551263433052571,
+      "grad_norm": 0.5406033992767334,
+      "learning_rate": 0.0001,
+      "loss": 1.5279,
+      "step": 9082
+    },
+    {
+      "epoch": 1.0552425210572176,
+      "grad_norm": 0.5218345522880554,
+      "learning_rate": 0.0001,
+      "loss": 1.4086,
+      "step": 9083
+    },
+    {
+      "epoch": 1.055358698809178,
+      "grad_norm": 0.5393182635307312,
+      "learning_rate": 0.0001,
+      "loss": 1.3487,
+      "step": 9084
+    },
+    {
+      "epoch": 1.0554748765611386,
+      "grad_norm": 0.5356934070587158,
+      "learning_rate": 0.0001,
+      "loss": 1.3364,
+      "step": 9085
+    },
+    {
+      "epoch": 1.055591054313099,
+      "grad_norm": 0.5603475570678711,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 9086
+    },
+    {
+      "epoch": 1.0557072320650596,
+      "grad_norm": 0.5328572988510132,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 9087
+    },
+    {
+      "epoch": 1.05582340981702,
+      "grad_norm": 0.5267220139503479,
+      "learning_rate": 0.0001,
+      "loss": 1.383,
+      "step": 9088
+    },
+    {
+      "epoch": 1.0559395875689805,
+      "grad_norm": 0.5259425640106201,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 9089
+    },
+    {
+      "epoch": 1.056055765320941,
+      "grad_norm": 0.5833221673965454,
+      "learning_rate": 0.0001,
+      "loss": 1.6627,
+      "step": 9090
+    },
+    {
+      "epoch": 1.0561719430729015,
+      "grad_norm": 0.5569556951522827,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 9091
+    },
+    {
+      "epoch": 1.056288120824862,
+      "grad_norm": 0.5657941699028015,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 9092
+    },
+    {
+      "epoch": 1.0564042985768225,
+      "grad_norm": 0.5313854813575745,
+      "learning_rate": 0.0001,
+      "loss": 1.4161,
+      "step": 9093
+    },
+    {
+      "epoch": 1.056520476328783,
+      "grad_norm": 0.530001699924469,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 9094
+    },
+    {
+      "epoch": 1.0566366540807435,
+      "grad_norm": 0.5262467265129089,
+      "learning_rate": 0.0001,
+      "loss": 1.4129,
+      "step": 9095
+    },
+    {
+      "epoch": 1.056752831832704,
+      "grad_norm": 0.567335844039917,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 9096
+    },
+    {
+      "epoch": 1.0568690095846645,
+      "grad_norm": 0.5013755559921265,
+      "learning_rate": 0.0001,
+      "loss": 1.4375,
+      "step": 9097
+    },
+    {
+      "epoch": 1.056985187336625,
+      "grad_norm": 0.494312047958374,
+      "learning_rate": 0.0001,
+      "loss": 1.445,
+      "step": 9098
+    },
+    {
+      "epoch": 1.0571013650885854,
+      "grad_norm": 0.5399910807609558,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 9099
+    },
+    {
+      "epoch": 1.0572175428405461,
+      "grad_norm": 0.5516524314880371,
+      "learning_rate": 0.0001,
+      "loss": 1.3564,
+      "step": 9100
+    },
+    {
+      "epoch": 1.0573337205925066,
+      "grad_norm": 0.5304887294769287,
+      "learning_rate": 0.0001,
+      "loss": 1.3785,
+      "step": 9101
+    },
+    {
+      "epoch": 1.0574498983444671,
+      "grad_norm": 0.591498851776123,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 9102
+    },
+    {
+      "epoch": 1.0575660760964276,
+      "grad_norm": 0.5390315651893616,
+      "learning_rate": 0.0001,
+      "loss": 1.3806,
+      "step": 9103
+    },
+    {
+      "epoch": 1.057682253848388,
+      "grad_norm": 0.5507586598396301,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 9104
+    },
+    {
+      "epoch": 1.0577984316003486,
+      "grad_norm": 0.5363011956214905,
+      "learning_rate": 0.0001,
+      "loss": 1.3498,
+      "step": 9105
+    },
+    {
+      "epoch": 1.057914609352309,
+      "grad_norm": 0.5194579362869263,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 9106
+    },
+    {
+      "epoch": 1.0580307871042696,
+      "grad_norm": 0.5246658325195312,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 9107
+    },
+    {
+      "epoch": 1.05814696485623,
+      "grad_norm": 0.5200863480567932,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 9108
+    },
+    {
+      "epoch": 1.0582631426081905,
+      "grad_norm": 0.5483285188674927,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 9109
+    },
+    {
+      "epoch": 1.058379320360151,
+      "grad_norm": 0.49789682030677795,
+      "learning_rate": 0.0001,
+      "loss": 1.3264,
+      "step": 9110
+    },
+    {
+      "epoch": 1.0584954981121115,
+      "grad_norm": 0.5474060773849487,
+      "learning_rate": 0.0001,
+      "loss": 1.4245,
+      "step": 9111
+    },
+    {
+      "epoch": 1.058611675864072,
+      "grad_norm": 0.6262152791023254,
+      "learning_rate": 0.0001,
+      "loss": 1.6001,
+      "step": 9112
+    },
+    {
+      "epoch": 1.0587278536160325,
+      "grad_norm": 0.522517740726471,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 9113
+    },
+    {
+      "epoch": 1.058844031367993,
+      "grad_norm": 0.5017138123512268,
+      "learning_rate": 0.0001,
+      "loss": 1.413,
+      "step": 9114
+    },
+    {
+      "epoch": 1.0589602091199535,
+      "grad_norm": 0.5721921920776367,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 9115
+    },
+    {
+      "epoch": 1.059076386871914,
+      "grad_norm": 0.5460672378540039,
+      "learning_rate": 0.0001,
+      "loss": 1.3668,
+      "step": 9116
+    },
+    {
+      "epoch": 1.0591925646238745,
+      "grad_norm": 0.5231673121452332,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 9117
+    },
+    {
+      "epoch": 1.059308742375835,
+      "grad_norm": 0.5345351099967957,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 9118
+    },
+    {
+      "epoch": 1.0594249201277954,
+      "grad_norm": 0.5771772861480713,
+      "learning_rate": 0.0001,
+      "loss": 1.708,
+      "step": 9119
+    },
+    {
+      "epoch": 1.059541097879756,
+      "grad_norm": 0.5495395660400391,
+      "learning_rate": 0.0001,
+      "loss": 1.4496,
+      "step": 9120
+    },
+    {
+      "epoch": 1.0596572756317166,
+      "grad_norm": 0.5115154981613159,
+      "learning_rate": 0.0001,
+      "loss": 1.2167,
+      "step": 9121
+    },
+    {
+      "epoch": 1.0597734533836771,
+      "grad_norm": 0.5737767219543457,
+      "learning_rate": 0.0001,
+      "loss": 1.6987,
+      "step": 9122
+    },
+    {
+      "epoch": 1.0598896311356376,
+      "grad_norm": 0.5117427706718445,
+      "learning_rate": 0.0001,
+      "loss": 1.3973,
+      "step": 9123
+    },
+    {
+      "epoch": 1.060005808887598,
+      "grad_norm": 0.5852498412132263,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 9124
+    },
+    {
+      "epoch": 1.0601219866395586,
+      "grad_norm": 0.5291174054145813,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 9125
+    },
+    {
+      "epoch": 1.060238164391519,
+      "grad_norm": 0.5075642466545105,
+      "learning_rate": 0.0001,
+      "loss": 1.414,
+      "step": 9126
+    },
+    {
+      "epoch": 1.0603543421434796,
+      "grad_norm": 0.47634080052375793,
+      "learning_rate": 0.0001,
+      "loss": 1.3369,
+      "step": 9127
+    },
+    {
+      "epoch": 1.06047051989544,
+      "grad_norm": 0.5397040843963623,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 9128
+    },
+    {
+      "epoch": 1.0605866976474005,
+      "grad_norm": 0.5401850938796997,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 9129
+    },
+    {
+      "epoch": 1.060702875399361,
+      "grad_norm": 0.5514304041862488,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 9130
+    },
+    {
+      "epoch": 1.0608190531513215,
+      "grad_norm": 0.5192503929138184,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 9131
+    },
+    {
+      "epoch": 1.060935230903282,
+      "grad_norm": 0.5016432404518127,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 9132
+    },
+    {
+      "epoch": 1.0610514086552425,
+      "grad_norm": 0.5191652178764343,
+      "learning_rate": 0.0001,
+      "loss": 1.3155,
+      "step": 9133
+    },
+    {
+      "epoch": 1.061167586407203,
+      "grad_norm": 0.5235947966575623,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 9134
+    },
+    {
+      "epoch": 1.0612837641591635,
+      "grad_norm": 0.538102924823761,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 9135
+    },
+    {
+      "epoch": 1.061399941911124,
+      "grad_norm": 0.5283055901527405,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 9136
+    },
+    {
+      "epoch": 1.0615161196630845,
+      "grad_norm": 0.5631850361824036,
+      "learning_rate": 0.0001,
+      "loss": 1.662,
+      "step": 9137
+    },
+    {
+      "epoch": 1.061632297415045,
+      "grad_norm": 0.5427637100219727,
+      "learning_rate": 0.0001,
+      "loss": 1.4435,
+      "step": 9138
+    },
+    {
+      "epoch": 1.0617484751670054,
+      "grad_norm": 0.5537014603614807,
+      "learning_rate": 0.0001,
+      "loss": 1.5422,
+      "step": 9139
+    },
+    {
+      "epoch": 1.061864652918966,
+      "grad_norm": 0.5393050312995911,
+      "learning_rate": 0.0001,
+      "loss": 1.6201,
+      "step": 9140
+    },
+    {
+      "epoch": 1.0619808306709264,
+      "grad_norm": 0.5826035141944885,
+      "learning_rate": 0.0001,
+      "loss": 1.6369,
+      "step": 9141
+    },
+    {
+      "epoch": 1.0620970084228871,
+      "grad_norm": 0.5376782417297363,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 9142
+    },
+    {
+      "epoch": 1.0622131861748476,
+      "grad_norm": 0.5235744714736938,
+      "learning_rate": 0.0001,
+      "loss": 1.3827,
+      "step": 9143
+    },
+    {
+      "epoch": 1.062329363926808,
+      "grad_norm": 0.5324127674102783,
+      "learning_rate": 0.0001,
+      "loss": 1.4812,
+      "step": 9144
+    },
+    {
+      "epoch": 1.0624455416787686,
+      "grad_norm": 0.5413472652435303,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 9145
+    },
+    {
+      "epoch": 1.062561719430729,
+      "grad_norm": 0.5227320790290833,
+      "learning_rate": 0.0001,
+      "loss": 1.5676,
+      "step": 9146
+    },
+    {
+      "epoch": 1.0626778971826896,
+      "grad_norm": 0.5471850037574768,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 9147
+    },
+    {
+      "epoch": 1.06279407493465,
+      "grad_norm": 0.5257613658905029,
+      "learning_rate": 0.0001,
+      "loss": 1.4821,
+      "step": 9148
+    },
+    {
+      "epoch": 1.0629102526866105,
+      "grad_norm": 0.5216736793518066,
+      "learning_rate": 0.0001,
+      "loss": 1.4997,
+      "step": 9149
+    },
+    {
+      "epoch": 1.063026430438571,
+      "grad_norm": 0.5594833493232727,
+      "learning_rate": 0.0001,
+      "loss": 1.6263,
+      "step": 9150
+    },
+    {
+      "epoch": 1.0631426081905315,
+      "grad_norm": 0.49669763445854187,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 9151
+    },
+    {
+      "epoch": 1.063258785942492,
+      "grad_norm": 0.5265761613845825,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 9152
+    },
+    {
+      "epoch": 1.0633749636944525,
+      "grad_norm": 0.5300150513648987,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 9153
+    },
+    {
+      "epoch": 1.063491141446413,
+      "grad_norm": 0.5201780796051025,
+      "learning_rate": 0.0001,
+      "loss": 1.3244,
+      "step": 9154
+    },
+    {
+      "epoch": 1.0636073191983735,
+      "grad_norm": 0.6005831360816956,
+      "learning_rate": 0.0001,
+      "loss": 1.5095,
+      "step": 9155
+    },
+    {
+      "epoch": 1.063723496950334,
+      "grad_norm": 0.5342938303947449,
+      "learning_rate": 0.0001,
+      "loss": 1.2851,
+      "step": 9156
+    },
+    {
+      "epoch": 1.0638396747022945,
+      "grad_norm": 0.5873286128044128,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 9157
+    },
+    {
+      "epoch": 1.063955852454255,
+      "grad_norm": 0.5283979177474976,
+      "learning_rate": 0.0001,
+      "loss": 1.3008,
+      "step": 9158
+    },
+    {
+      "epoch": 1.0640720302062154,
+      "grad_norm": 0.5319207310676575,
+      "learning_rate": 0.0001,
+      "loss": 1.3622,
+      "step": 9159
+    },
+    {
+      "epoch": 1.064188207958176,
+      "grad_norm": 0.526910662651062,
+      "learning_rate": 0.0001,
+      "loss": 1.3385,
+      "step": 9160
+    },
+    {
+      "epoch": 1.0643043857101364,
+      "grad_norm": 0.5416330099105835,
+      "learning_rate": 0.0001,
+      "loss": 1.369,
+      "step": 9161
+    },
+    {
+      "epoch": 1.0644205634620971,
+      "grad_norm": 0.5400428175926208,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 9162
+    },
+    {
+      "epoch": 1.0645367412140576,
+      "grad_norm": 0.5297672152519226,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 9163
+    },
+    {
+      "epoch": 1.064652918966018,
+      "grad_norm": 0.5449459552764893,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 9164
+    },
+    {
+      "epoch": 1.0647690967179786,
+      "grad_norm": 0.5715656876564026,
+      "learning_rate": 0.0001,
+      "loss": 1.6004,
+      "step": 9165
+    },
+    {
+      "epoch": 1.064885274469939,
+      "grad_norm": 0.5295259952545166,
+      "learning_rate": 0.0001,
+      "loss": 1.3461,
+      "step": 9166
+    },
+    {
+      "epoch": 1.0650014522218996,
+      "grad_norm": 0.5438884496688843,
+      "learning_rate": 0.0001,
+      "loss": 1.3848,
+      "step": 9167
+    },
+    {
+      "epoch": 1.06511762997386,
+      "grad_norm": 0.542167603969574,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 9168
+    },
+    {
+      "epoch": 1.0652338077258205,
+      "grad_norm": 0.5348635315895081,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 9169
+    },
+    {
+      "epoch": 1.065349985477781,
+      "grad_norm": 0.5906919836997986,
+      "learning_rate": 0.0001,
+      "loss": 1.6726,
+      "step": 9170
+    },
+    {
+      "epoch": 1.0654661632297415,
+      "grad_norm": 0.5702545642852783,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 9171
+    },
+    {
+      "epoch": 1.065582340981702,
+      "grad_norm": 0.5129088759422302,
+      "learning_rate": 0.0001,
+      "loss": 1.2603,
+      "step": 9172
+    },
+    {
+      "epoch": 1.0656985187336625,
+      "grad_norm": 0.5603011250495911,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 9173
+    },
+    {
+      "epoch": 1.065814696485623,
+      "grad_norm": 0.6020777821540833,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 9174
+    },
+    {
+      "epoch": 1.0659308742375835,
+      "grad_norm": 0.5032897591590881,
+      "learning_rate": 0.0001,
+      "loss": 1.3745,
+      "step": 9175
+    },
+    {
+      "epoch": 1.066047051989544,
+      "grad_norm": 0.5128781199455261,
+      "learning_rate": 0.0001,
+      "loss": 1.2593,
+      "step": 9176
+    },
+    {
+      "epoch": 1.0661632297415045,
+      "grad_norm": 0.5254261493682861,
+      "learning_rate": 0.0001,
+      "loss": 1.3345,
+      "step": 9177
+    },
+    {
+      "epoch": 1.066279407493465,
+      "grad_norm": 0.508584201335907,
+      "learning_rate": 0.0001,
+      "loss": 1.4429,
+      "step": 9178
+    },
+    {
+      "epoch": 1.0663955852454254,
+      "grad_norm": 0.5458223223686218,
+      "learning_rate": 0.0001,
+      "loss": 1.487,
+      "step": 9179
+    },
+    {
+      "epoch": 1.066511762997386,
+      "grad_norm": 0.5139407515525818,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 9180
+    },
+    {
+      "epoch": 1.0666279407493464,
+      "grad_norm": 0.505296528339386,
+      "learning_rate": 0.0001,
+      "loss": 1.3642,
+      "step": 9181
+    },
+    {
+      "epoch": 1.066744118501307,
+      "grad_norm": 0.5002397298812866,
+      "learning_rate": 0.0001,
+      "loss": 1.3983,
+      "step": 9182
+    },
+    {
+      "epoch": 1.0668602962532674,
+      "grad_norm": 0.551529049873352,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 9183
+    },
+    {
+      "epoch": 1.0669764740052279,
+      "grad_norm": 0.578542947769165,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 9184
+    },
+    {
+      "epoch": 1.0670926517571886,
+      "grad_norm": 0.531636655330658,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 9185
+    },
+    {
+      "epoch": 1.067208829509149,
+      "grad_norm": 0.5241780281066895,
+      "learning_rate": 0.0001,
+      "loss": 1.3676,
+      "step": 9186
+    },
+    {
+      "epoch": 1.0673250072611096,
+      "grad_norm": 0.5631434321403503,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 9187
+    },
+    {
+      "epoch": 1.06744118501307,
+      "grad_norm": 0.5912564992904663,
+      "learning_rate": 0.0001,
+      "loss": 1.5429,
+      "step": 9188
+    },
+    {
+      "epoch": 1.0675573627650305,
+      "grad_norm": 0.4988497197628021,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 9189
+    },
+    {
+      "epoch": 1.067673540516991,
+      "grad_norm": 0.5904607772827148,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 9190
+    },
+    {
+      "epoch": 1.0677897182689515,
+      "grad_norm": 0.5441171526908875,
+      "learning_rate": 0.0001,
+      "loss": 1.3963,
+      "step": 9191
+    },
+    {
+      "epoch": 1.067905896020912,
+      "grad_norm": 0.541077196598053,
+      "learning_rate": 0.0001,
+      "loss": 1.5066,
+      "step": 9192
+    },
+    {
+      "epoch": 1.0680220737728725,
+      "grad_norm": 0.53225177526474,
+      "learning_rate": 0.0001,
+      "loss": 1.4918,
+      "step": 9193
+    },
+    {
+      "epoch": 1.068138251524833,
+      "grad_norm": 0.543021559715271,
+      "learning_rate": 0.0001,
+      "loss": 1.3748,
+      "step": 9194
+    },
+    {
+      "epoch": 1.0682544292767935,
+      "grad_norm": 0.6031702756881714,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 9195
+    },
+    {
+      "epoch": 1.068370607028754,
+      "grad_norm": 0.571552574634552,
+      "learning_rate": 0.0001,
+      "loss": 1.3105,
+      "step": 9196
+    },
+    {
+      "epoch": 1.0684867847807145,
+      "grad_norm": 0.5475126504898071,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 9197
+    },
+    {
+      "epoch": 1.068602962532675,
+      "grad_norm": 0.5954582691192627,
+      "learning_rate": 0.0001,
+      "loss": 1.6204,
+      "step": 9198
+    },
+    {
+      "epoch": 1.0687191402846354,
+      "grad_norm": 0.5108276605606079,
+      "learning_rate": 0.0001,
+      "loss": 1.3587,
+      "step": 9199
+    },
+    {
+      "epoch": 1.068835318036596,
+      "grad_norm": 0.5748639106750488,
+      "learning_rate": 0.0001,
+      "loss": 1.6054,
+      "step": 9200
+    },
+    {
+      "epoch": 1.0689514957885564,
+      "grad_norm": 0.5265009999275208,
+      "learning_rate": 0.0001,
+      "loss": 1.3955,
+      "step": 9201
+    },
+    {
+      "epoch": 1.069067673540517,
+      "grad_norm": 0.5206406116485596,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 9202
+    },
+    {
+      "epoch": 1.0691838512924774,
+      "grad_norm": 0.5493237972259521,
+      "learning_rate": 0.0001,
+      "loss": 1.6656,
+      "step": 9203
+    },
+    {
+      "epoch": 1.069300029044438,
+      "grad_norm": 0.5620342493057251,
+      "learning_rate": 0.0001,
+      "loss": 1.4971,
+      "step": 9204
+    },
+    {
+      "epoch": 1.0694162067963986,
+      "grad_norm": 0.5706990361213684,
+      "learning_rate": 0.0001,
+      "loss": 1.6995,
+      "step": 9205
+    },
+    {
+      "epoch": 1.069532384548359,
+      "grad_norm": 0.5248960852622986,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 9206
+    },
+    {
+      "epoch": 1.0696485623003196,
+      "grad_norm": 0.5431904196739197,
+      "learning_rate": 0.0001,
+      "loss": 1.4966,
+      "step": 9207
+    },
+    {
+      "epoch": 1.06976474005228,
+      "grad_norm": 0.5573443174362183,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 9208
+    },
+    {
+      "epoch": 1.0698809178042405,
+      "grad_norm": 0.5166314840316772,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 9209
+    },
+    {
+      "epoch": 1.069997095556201,
+      "grad_norm": 0.5510638356208801,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 9210
+    },
+    {
+      "epoch": 1.0701132733081615,
+      "grad_norm": 0.5057435035705566,
+      "learning_rate": 0.0001,
+      "loss": 1.3959,
+      "step": 9211
+    },
+    {
+      "epoch": 1.070229451060122,
+      "grad_norm": 0.5617602467536926,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 9212
+    },
+    {
+      "epoch": 1.0703456288120825,
+      "grad_norm": 0.5629894733428955,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 9213
+    },
+    {
+      "epoch": 1.070461806564043,
+      "grad_norm": 0.5455237030982971,
+      "learning_rate": 0.0001,
+      "loss": 1.4992,
+      "step": 9214
+    },
+    {
+      "epoch": 1.0705779843160035,
+      "grad_norm": 0.5817468762397766,
+      "learning_rate": 0.0001,
+      "loss": 1.5759,
+      "step": 9215
+    },
+    {
+      "epoch": 1.070694162067964,
+      "grad_norm": 0.5639060735702515,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 9216
+    },
+    {
+      "epoch": 1.0708103398199245,
+      "grad_norm": 0.5920657515525818,
+      "learning_rate": 0.0001,
+      "loss": 1.6202,
+      "step": 9217
+    },
+    {
+      "epoch": 1.070926517571885,
+      "grad_norm": 0.513184666633606,
+      "learning_rate": 0.0001,
+      "loss": 1.394,
+      "step": 9218
+    },
+    {
+      "epoch": 1.0710426953238454,
+      "grad_norm": 0.5242840647697449,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 9219
+    },
+    {
+      "epoch": 1.071158873075806,
+      "grad_norm": 0.5487340688705444,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 9220
+    },
+    {
+      "epoch": 1.0712750508277664,
+      "grad_norm": 0.7941916584968567,
+      "learning_rate": 0.0001,
+      "loss": 1.7697,
+      "step": 9221
+    },
+    {
+      "epoch": 1.071391228579727,
+      "grad_norm": 0.5666763782501221,
+      "learning_rate": 0.0001,
+      "loss": 1.6823,
+      "step": 9222
+    },
+    {
+      "epoch": 1.0715074063316874,
+      "grad_norm": 0.5537146925926208,
+      "learning_rate": 0.0001,
+      "loss": 1.3735,
+      "step": 9223
+    },
+    {
+      "epoch": 1.0716235840836479,
+      "grad_norm": 0.5601913928985596,
+      "learning_rate": 0.0001,
+      "loss": 1.3579,
+      "step": 9224
+    },
+    {
+      "epoch": 1.0717397618356084,
+      "grad_norm": 0.5344251990318298,
+      "learning_rate": 0.0001,
+      "loss": 1.5254,
+      "step": 9225
+    },
+    {
+      "epoch": 1.071855939587569,
+      "grad_norm": 0.5359842777252197,
+      "learning_rate": 0.0001,
+      "loss": 1.3922,
+      "step": 9226
+    },
+    {
+      "epoch": 1.0719721173395296,
+      "grad_norm": 0.5405870079994202,
+      "learning_rate": 0.0001,
+      "loss": 1.4498,
+      "step": 9227
+    },
+    {
+      "epoch": 1.07208829509149,
+      "grad_norm": 0.51329505443573,
+      "learning_rate": 0.0001,
+      "loss": 1.4123,
+      "step": 9228
+    },
+    {
+      "epoch": 1.0722044728434506,
+      "grad_norm": 0.5659348368644714,
+      "learning_rate": 0.0001,
+      "loss": 1.3418,
+      "step": 9229
+    },
+    {
+      "epoch": 1.072320650595411,
+      "grad_norm": 0.5718405842781067,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 9230
+    },
+    {
+      "epoch": 1.0724368283473715,
+      "grad_norm": 0.5486005544662476,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 9231
+    },
+    {
+      "epoch": 1.072553006099332,
+      "grad_norm": 0.5066012740135193,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 9232
+    },
+    {
+      "epoch": 1.0726691838512925,
+      "grad_norm": 0.5650283098220825,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 9233
+    },
+    {
+      "epoch": 1.072785361603253,
+      "grad_norm": 0.5359314680099487,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 9234
+    },
+    {
+      "epoch": 1.0729015393552135,
+      "grad_norm": 0.5751736164093018,
+      "learning_rate": 0.0001,
+      "loss": 1.6148,
+      "step": 9235
+    },
+    {
+      "epoch": 1.073017717107174,
+      "grad_norm": 0.5163309574127197,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 9236
+    },
+    {
+      "epoch": 1.0731338948591345,
+      "grad_norm": 0.550383448600769,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 9237
+    },
+    {
+      "epoch": 1.073250072611095,
+      "grad_norm": 0.5238305330276489,
+      "learning_rate": 0.0001,
+      "loss": 1.4432,
+      "step": 9238
+    },
+    {
+      "epoch": 1.0733662503630554,
+      "grad_norm": 0.49950873851776123,
+      "learning_rate": 0.0001,
+      "loss": 1.3584,
+      "step": 9239
+    },
+    {
+      "epoch": 1.073482428115016,
+      "grad_norm": 0.508315920829773,
+      "learning_rate": 0.0001,
+      "loss": 1.3408,
+      "step": 9240
+    },
+    {
+      "epoch": 1.0735986058669764,
+      "grad_norm": 0.5786048173904419,
+      "learning_rate": 0.0001,
+      "loss": 1.27,
+      "step": 9241
+    },
+    {
+      "epoch": 1.073714783618937,
+      "grad_norm": 0.5773483514785767,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 9242
+    },
+    {
+      "epoch": 1.0738309613708974,
+      "grad_norm": 0.5315750241279602,
+      "learning_rate": 0.0001,
+      "loss": 1.469,
+      "step": 9243
+    },
+    {
+      "epoch": 1.0739471391228579,
+      "grad_norm": 0.6098319292068481,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 9244
+    },
+    {
+      "epoch": 1.0740633168748184,
+      "grad_norm": 0.5994076132774353,
+      "learning_rate": 0.0001,
+      "loss": 1.6424,
+      "step": 9245
+    },
+    {
+      "epoch": 1.074179494626779,
+      "grad_norm": 0.5081837177276611,
+      "learning_rate": 0.0001,
+      "loss": 1.3348,
+      "step": 9246
+    },
+    {
+      "epoch": 1.0742956723787396,
+      "grad_norm": 0.5301439166069031,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 9247
+    },
+    {
+      "epoch": 1.0744118501307,
+      "grad_norm": 0.5718616247177124,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 9248
+    },
+    {
+      "epoch": 1.0745280278826606,
+      "grad_norm": 0.5080563426017761,
+      "learning_rate": 0.0001,
+      "loss": 1.3807,
+      "step": 9249
+    },
+    {
+      "epoch": 1.074644205634621,
+      "grad_norm": 0.5610292553901672,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 9250
+    },
+    {
+      "epoch": 1.0747603833865815,
+      "grad_norm": 0.5332043766975403,
+      "learning_rate": 0.0001,
+      "loss": 1.4046,
+      "step": 9251
+    },
+    {
+      "epoch": 1.074876561138542,
+      "grad_norm": 0.5250877737998962,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 9252
+    },
+    {
+      "epoch": 1.0749927388905025,
+      "grad_norm": 0.5090211629867554,
+      "learning_rate": 0.0001,
+      "loss": 1.3705,
+      "step": 9253
+    },
+    {
+      "epoch": 1.075108916642463,
+      "grad_norm": 0.5132327079772949,
+      "learning_rate": 0.0001,
+      "loss": 1.4105,
+      "step": 9254
+    },
+    {
+      "epoch": 1.0752250943944235,
+      "grad_norm": 0.5518538951873779,
+      "learning_rate": 0.0001,
+      "loss": 1.5959,
+      "step": 9255
+    },
+    {
+      "epoch": 1.075341272146384,
+      "grad_norm": 0.5492732524871826,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 9256
+    },
+    {
+      "epoch": 1.0754574498983445,
+      "grad_norm": 0.5491774678230286,
+      "learning_rate": 0.0001,
+      "loss": 1.4346,
+      "step": 9257
+    },
+    {
+      "epoch": 1.075573627650305,
+      "grad_norm": 0.5792922377586365,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 9258
+    },
+    {
+      "epoch": 1.0756898054022654,
+      "grad_norm": 0.5162052512168884,
+      "learning_rate": 0.0001,
+      "loss": 1.3871,
+      "step": 9259
+    },
+    {
+      "epoch": 1.075805983154226,
+      "grad_norm": 0.5442473292350769,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 9260
+    },
+    {
+      "epoch": 1.0759221609061864,
+      "grad_norm": 0.5718018412590027,
+      "learning_rate": 0.0001,
+      "loss": 1.5434,
+      "step": 9261
+    },
+    {
+      "epoch": 1.076038338658147,
+      "grad_norm": 0.53264319896698,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 9262
+    },
+    {
+      "epoch": 1.0761545164101074,
+      "grad_norm": 0.5195859670639038,
+      "learning_rate": 0.0001,
+      "loss": 1.3004,
+      "step": 9263
+    },
+    {
+      "epoch": 1.0762706941620679,
+      "grad_norm": 0.5994012951850891,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 9264
+    },
+    {
+      "epoch": 1.0763868719140284,
+      "grad_norm": 0.5237147212028503,
+      "learning_rate": 0.0001,
+      "loss": 1.3221,
+      "step": 9265
+    },
+    {
+      "epoch": 1.0765030496659889,
+      "grad_norm": 0.5110024809837341,
+      "learning_rate": 0.0001,
+      "loss": 1.2827,
+      "step": 9266
+    },
+    {
+      "epoch": 1.0766192274179494,
+      "grad_norm": 0.5228458642959595,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 9267
+    },
+    {
+      "epoch": 1.07673540516991,
+      "grad_norm": 0.5706833004951477,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 9268
+    },
+    {
+      "epoch": 1.0768515829218706,
+      "grad_norm": 0.5496780276298523,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 9269
+    },
+    {
+      "epoch": 1.076967760673831,
+      "grad_norm": 0.5509944558143616,
+      "learning_rate": 0.0001,
+      "loss": 1.3909,
+      "step": 9270
+    },
+    {
+      "epoch": 1.0770839384257915,
+      "grad_norm": 0.5543323755264282,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 9271
+    },
+    {
+      "epoch": 1.077200116177752,
+      "grad_norm": 0.5567429661750793,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 9272
+    },
+    {
+      "epoch": 1.0773162939297125,
+      "grad_norm": 0.5358352661132812,
+      "learning_rate": 0.0001,
+      "loss": 1.3017,
+      "step": 9273
+    },
+    {
+      "epoch": 1.077432471681673,
+      "grad_norm": 0.5601693987846375,
+      "learning_rate": 0.0001,
+      "loss": 1.5659,
+      "step": 9274
+    },
+    {
+      "epoch": 1.0775486494336335,
+      "grad_norm": 0.5503692626953125,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 9275
+    },
+    {
+      "epoch": 1.077664827185594,
+      "grad_norm": 0.5987486839294434,
+      "learning_rate": 0.0001,
+      "loss": 1.459,
+      "step": 9276
+    },
+    {
+      "epoch": 1.0777810049375545,
+      "grad_norm": 0.6052455902099609,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 9277
+    },
+    {
+      "epoch": 1.077897182689515,
+      "grad_norm": 0.5339061617851257,
+      "learning_rate": 0.0001,
+      "loss": 1.4148,
+      "step": 9278
+    },
+    {
+      "epoch": 1.0780133604414754,
+      "grad_norm": 0.5292863845825195,
+      "learning_rate": 0.0001,
+      "loss": 1.5569,
+      "step": 9279
+    },
+    {
+      "epoch": 1.078129538193436,
+      "grad_norm": 0.5491698980331421,
+      "learning_rate": 0.0001,
+      "loss": 1.3836,
+      "step": 9280
+    },
+    {
+      "epoch": 1.0782457159453964,
+      "grad_norm": 0.5501977205276489,
+      "learning_rate": 0.0001,
+      "loss": 1.6123,
+      "step": 9281
+    },
+    {
+      "epoch": 1.078361893697357,
+      "grad_norm": 0.5370213389396667,
+      "learning_rate": 0.0001,
+      "loss": 1.5334,
+      "step": 9282
+    },
+    {
+      "epoch": 1.0784780714493174,
+      "grad_norm": 0.5482035279273987,
+      "learning_rate": 0.0001,
+      "loss": 1.239,
+      "step": 9283
+    },
+    {
+      "epoch": 1.0785942492012779,
+      "grad_norm": 0.5296024680137634,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 9284
+    },
+    {
+      "epoch": 1.0787104269532384,
+      "grad_norm": 0.4946173131465912,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 9285
+    },
+    {
+      "epoch": 1.0788266047051989,
+      "grad_norm": 0.5261144042015076,
+      "learning_rate": 0.0001,
+      "loss": 1.4101,
+      "step": 9286
+    },
+    {
+      "epoch": 1.0789427824571594,
+      "grad_norm": 0.5421222448348999,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 9287
+    },
+    {
+      "epoch": 1.07905896020912,
+      "grad_norm": 0.5617155432701111,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 9288
+    },
+    {
+      "epoch": 1.0791751379610806,
+      "grad_norm": 0.5525049567222595,
+      "learning_rate": 0.0001,
+      "loss": 1.4768,
+      "step": 9289
+    },
+    {
+      "epoch": 1.079291315713041,
+      "grad_norm": 0.5454285740852356,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 9290
+    },
+    {
+      "epoch": 1.0794074934650015,
+      "grad_norm": 0.5999751091003418,
+      "learning_rate": 0.0001,
+      "loss": 1.3304,
+      "step": 9291
+    },
+    {
+      "epoch": 1.079523671216962,
+      "grad_norm": 0.5454568862915039,
+      "learning_rate": 0.0001,
+      "loss": 1.375,
+      "step": 9292
+    },
+    {
+      "epoch": 1.0796398489689225,
+      "grad_norm": 0.5416858792304993,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 9293
+    },
+    {
+      "epoch": 1.079756026720883,
+      "grad_norm": 0.5695158243179321,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 9294
+    },
+    {
+      "epoch": 1.0798722044728435,
+      "grad_norm": 0.5237754583358765,
+      "learning_rate": 0.0001,
+      "loss": 1.2665,
+      "step": 9295
+    },
+    {
+      "epoch": 1.079988382224804,
+      "grad_norm": 0.5280129909515381,
+      "learning_rate": 0.0001,
+      "loss": 1.3654,
+      "step": 9296
+    },
+    {
+      "epoch": 1.0801045599767645,
+      "grad_norm": 0.5281645655632019,
+      "learning_rate": 0.0001,
+      "loss": 1.2907,
+      "step": 9297
+    },
+    {
+      "epoch": 1.080220737728725,
+      "grad_norm": 0.5463537573814392,
+      "learning_rate": 0.0001,
+      "loss": 1.4693,
+      "step": 9298
+    },
+    {
+      "epoch": 1.0803369154806854,
+      "grad_norm": 0.5508739948272705,
+      "learning_rate": 0.0001,
+      "loss": 1.304,
+      "step": 9299
+    },
+    {
+      "epoch": 1.080453093232646,
+      "grad_norm": 0.6003327369689941,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 9300
+    },
+    {
+      "epoch": 1.0805692709846064,
+      "grad_norm": 0.5155467987060547,
+      "learning_rate": 0.0001,
+      "loss": 1.3432,
+      "step": 9301
+    },
+    {
+      "epoch": 1.080685448736567,
+      "grad_norm": 0.5434906482696533,
+      "learning_rate": 0.0001,
+      "loss": 1.4054,
+      "step": 9302
+    },
+    {
+      "epoch": 1.0808016264885274,
+      "grad_norm": 0.5202093124389648,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 9303
+    },
+    {
+      "epoch": 1.080917804240488,
+      "grad_norm": 0.5816706418991089,
+      "learning_rate": 0.0001,
+      "loss": 1.4522,
+      "step": 9304
+    },
+    {
+      "epoch": 1.0810339819924484,
+      "grad_norm": 0.5725724101066589,
+      "learning_rate": 0.0001,
+      "loss": 1.5505,
+      "step": 9305
+    },
+    {
+      "epoch": 1.0811501597444089,
+      "grad_norm": 0.5437831282615662,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 9306
+    },
+    {
+      "epoch": 1.0812663374963694,
+      "grad_norm": 0.5659807920455933,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 9307
+    },
+    {
+      "epoch": 1.0813825152483298,
+      "grad_norm": 0.5198776125907898,
+      "learning_rate": 0.0001,
+      "loss": 1.4128,
+      "step": 9308
+    },
+    {
+      "epoch": 1.0814986930002903,
+      "grad_norm": 0.6664596199989319,
+      "learning_rate": 0.0001,
+      "loss": 1.7793,
+      "step": 9309
+    },
+    {
+      "epoch": 1.081614870752251,
+      "grad_norm": 0.5590860247612,
+      "learning_rate": 0.0001,
+      "loss": 1.4313,
+      "step": 9310
+    },
+    {
+      "epoch": 1.0817310485042115,
+      "grad_norm": 0.5522528290748596,
+      "learning_rate": 0.0001,
+      "loss": 1.378,
+      "step": 9311
+    },
+    {
+      "epoch": 1.081847226256172,
+      "grad_norm": 0.5644780397415161,
+      "learning_rate": 0.0001,
+      "loss": 1.3321,
+      "step": 9312
+    },
+    {
+      "epoch": 1.0819634040081325,
+      "grad_norm": 0.5319089889526367,
+      "learning_rate": 0.0001,
+      "loss": 1.2627,
+      "step": 9313
+    },
+    {
+      "epoch": 1.082079581760093,
+      "grad_norm": 0.5518312454223633,
+      "learning_rate": 0.0001,
+      "loss": 1.3252,
+      "step": 9314
+    },
+    {
+      "epoch": 1.0821957595120535,
+      "grad_norm": 0.5743319392204285,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 9315
+    },
+    {
+      "epoch": 1.082311937264014,
+      "grad_norm": 0.5341470241546631,
+      "learning_rate": 0.0001,
+      "loss": 1.348,
+      "step": 9316
+    },
+    {
+      "epoch": 1.0824281150159745,
+      "grad_norm": 0.5499274134635925,
+      "learning_rate": 0.0001,
+      "loss": 1.3848,
+      "step": 9317
+    },
+    {
+      "epoch": 1.082544292767935,
+      "grad_norm": 0.5466471910476685,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 9318
+    },
+    {
+      "epoch": 1.0826604705198954,
+      "grad_norm": 0.5436463356018066,
+      "learning_rate": 0.0001,
+      "loss": 1.3873,
+      "step": 9319
+    },
+    {
+      "epoch": 1.082776648271856,
+      "grad_norm": 0.5741965770721436,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 9320
+    },
+    {
+      "epoch": 1.0828928260238164,
+      "grad_norm": 0.5448581576347351,
+      "learning_rate": 0.0001,
+      "loss": 1.4217,
+      "step": 9321
+    },
+    {
+      "epoch": 1.083009003775777,
+      "grad_norm": 0.5822571516036987,
+      "learning_rate": 0.0001,
+      "loss": 1.5581,
+      "step": 9322
+    },
+    {
+      "epoch": 1.0831251815277374,
+      "grad_norm": 0.6465715765953064,
+      "learning_rate": 0.0001,
+      "loss": 1.5893,
+      "step": 9323
+    },
+    {
+      "epoch": 1.083241359279698,
+      "grad_norm": 0.5584543943405151,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 9324
+    },
+    {
+      "epoch": 1.0833575370316584,
+      "grad_norm": 0.5714010000228882,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 9325
+    },
+    {
+      "epoch": 1.0834737147836189,
+      "grad_norm": 0.5465094447135925,
+      "learning_rate": 0.0001,
+      "loss": 1.4766,
+      "step": 9326
+    },
+    {
+      "epoch": 1.0835898925355794,
+      "grad_norm": 0.5097503066062927,
+      "learning_rate": 0.0001,
+      "loss": 1.3755,
+      "step": 9327
+    },
+    {
+      "epoch": 1.0837060702875398,
+      "grad_norm": 0.5056722164154053,
+      "learning_rate": 0.0001,
+      "loss": 1.2346,
+      "step": 9328
+    },
+    {
+      "epoch": 1.0838222480395003,
+      "grad_norm": 0.565323531627655,
+      "learning_rate": 0.0001,
+      "loss": 1.3642,
+      "step": 9329
+    },
+    {
+      "epoch": 1.083938425791461,
+      "grad_norm": 0.5437831878662109,
+      "learning_rate": 0.0001,
+      "loss": 1.3076,
+      "step": 9330
+    },
+    {
+      "epoch": 1.0840546035434215,
+      "grad_norm": 0.5402218103408813,
+      "learning_rate": 0.0001,
+      "loss": 1.3891,
+      "step": 9331
+    },
+    {
+      "epoch": 1.084170781295382,
+      "grad_norm": 0.5713097453117371,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 9332
+    },
+    {
+      "epoch": 1.0842869590473425,
+      "grad_norm": 0.5773684978485107,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 9333
+    },
+    {
+      "epoch": 1.084403136799303,
+      "grad_norm": 0.4951843023300171,
+      "learning_rate": 0.0001,
+      "loss": 1.1599,
+      "step": 9334
+    },
+    {
+      "epoch": 1.0845193145512635,
+      "grad_norm": 0.5116522312164307,
+      "learning_rate": 0.0001,
+      "loss": 1.4187,
+      "step": 9335
+    },
+    {
+      "epoch": 1.084635492303224,
+      "grad_norm": 0.5548305511474609,
+      "learning_rate": 0.0001,
+      "loss": 1.6528,
+      "step": 9336
+    },
+    {
+      "epoch": 1.0847516700551845,
+      "grad_norm": 0.5454692840576172,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 9337
+    },
+    {
+      "epoch": 1.084867847807145,
+      "grad_norm": 0.5281533598899841,
+      "learning_rate": 0.0001,
+      "loss": 1.385,
+      "step": 9338
+    },
+    {
+      "epoch": 1.0849840255591054,
+      "grad_norm": 0.5260929465293884,
+      "learning_rate": 0.0001,
+      "loss": 1.3408,
+      "step": 9339
+    },
+    {
+      "epoch": 1.085100203311066,
+      "grad_norm": 0.5246114134788513,
+      "learning_rate": 0.0001,
+      "loss": 1.3622,
+      "step": 9340
+    },
+    {
+      "epoch": 1.0852163810630264,
+      "grad_norm": 0.5710988640785217,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 9341
+    },
+    {
+      "epoch": 1.085332558814987,
+      "grad_norm": 0.5941675901412964,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 9342
+    },
+    {
+      "epoch": 1.0854487365669474,
+      "grad_norm": 0.5492114424705505,
+      "learning_rate": 0.0001,
+      "loss": 1.4177,
+      "step": 9343
+    },
+    {
+      "epoch": 1.085564914318908,
+      "grad_norm": 0.5595629215240479,
+      "learning_rate": 0.0001,
+      "loss": 1.4012,
+      "step": 9344
+    },
+    {
+      "epoch": 1.0856810920708684,
+      "grad_norm": 0.5756628513336182,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 9345
+    },
+    {
+      "epoch": 1.0857972698228289,
+      "grad_norm": 0.5529462099075317,
+      "learning_rate": 0.0001,
+      "loss": 1.5355,
+      "step": 9346
+    },
+    {
+      "epoch": 1.0859134475747894,
+      "grad_norm": 0.5427260398864746,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 9347
+    },
+    {
+      "epoch": 1.0860296253267498,
+      "grad_norm": 0.49472326040267944,
+      "learning_rate": 0.0001,
+      "loss": 1.3941,
+      "step": 9348
+    },
+    {
+      "epoch": 1.0861458030787103,
+      "grad_norm": 0.5269715189933777,
+      "learning_rate": 0.0001,
+      "loss": 1.4096,
+      "step": 9349
+    },
+    {
+      "epoch": 1.0862619808306708,
+      "grad_norm": 0.5339666604995728,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 9350
+    },
+    {
+      "epoch": 1.0863781585826313,
+      "grad_norm": 0.5161392688751221,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 9351
+    },
+    {
+      "epoch": 1.086494336334592,
+      "grad_norm": 0.5255089998245239,
+      "learning_rate": 0.0001,
+      "loss": 1.3307,
+      "step": 9352
+    },
+    {
+      "epoch": 1.0866105140865525,
+      "grad_norm": 0.5742126107215881,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 9353
+    },
+    {
+      "epoch": 1.086726691838513,
+      "grad_norm": 0.5712875723838806,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 9354
+    },
+    {
+      "epoch": 1.0868428695904735,
+      "grad_norm": 0.5453868508338928,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 9355
+    },
+    {
+      "epoch": 1.086959047342434,
+      "grad_norm": 0.5706861019134521,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 9356
+    },
+    {
+      "epoch": 1.0870752250943945,
+      "grad_norm": 0.5392447710037231,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 9357
+    },
+    {
+      "epoch": 1.087191402846355,
+      "grad_norm": 0.5431585907936096,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 9358
+    },
+    {
+      "epoch": 1.0873075805983154,
+      "grad_norm": 0.6064168810844421,
+      "learning_rate": 0.0001,
+      "loss": 1.596,
+      "step": 9359
+    },
+    {
+      "epoch": 1.087423758350276,
+      "grad_norm": 0.6154705882072449,
+      "learning_rate": 0.0001,
+      "loss": 1.4216,
+      "step": 9360
+    },
+    {
+      "epoch": 1.0875399361022364,
+      "grad_norm": 0.5736680030822754,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 9361
+    },
+    {
+      "epoch": 1.087656113854197,
+      "grad_norm": 0.5389428734779358,
+      "learning_rate": 0.0001,
+      "loss": 1.5365,
+      "step": 9362
+    },
+    {
+      "epoch": 1.0877722916061574,
+      "grad_norm": 0.5451179146766663,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 9363
+    },
+    {
+      "epoch": 1.087888469358118,
+      "grad_norm": 0.5412275791168213,
+      "learning_rate": 0.0001,
+      "loss": 1.4005,
+      "step": 9364
+    },
+    {
+      "epoch": 1.0880046471100784,
+      "grad_norm": 0.5797693729400635,
+      "learning_rate": 0.0001,
+      "loss": 1.5728,
+      "step": 9365
+    },
+    {
+      "epoch": 1.0881208248620389,
+      "grad_norm": 0.5813042521476746,
+      "learning_rate": 0.0001,
+      "loss": 1.7605,
+      "step": 9366
+    },
+    {
+      "epoch": 1.0882370026139994,
+      "grad_norm": 0.5206052660942078,
+      "learning_rate": 0.0001,
+      "loss": 1.3434,
+      "step": 9367
+    },
+    {
+      "epoch": 1.0883531803659598,
+      "grad_norm": 0.5384072661399841,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 9368
+    },
+    {
+      "epoch": 1.0884693581179203,
+      "grad_norm": 0.49630725383758545,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 9369
+    },
+    {
+      "epoch": 1.0885855358698808,
+      "grad_norm": 0.5630286335945129,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 9370
+    },
+    {
+      "epoch": 1.0887017136218413,
+      "grad_norm": 0.5383244156837463,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 9371
+    },
+    {
+      "epoch": 1.088817891373802,
+      "grad_norm": 0.6035181879997253,
+      "learning_rate": 0.0001,
+      "loss": 1.489,
+      "step": 9372
+    },
+    {
+      "epoch": 1.0889340691257625,
+      "grad_norm": 0.5531001687049866,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 9373
+    },
+    {
+      "epoch": 1.089050246877723,
+      "grad_norm": 0.5467984676361084,
+      "learning_rate": 0.0001,
+      "loss": 1.3719,
+      "step": 9374
+    },
+    {
+      "epoch": 1.0891664246296835,
+      "grad_norm": 0.5076682567596436,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 9375
+    },
+    {
+      "epoch": 1.089282602381644,
+      "grad_norm": 0.5291927456855774,
+      "learning_rate": 0.0001,
+      "loss": 1.3413,
+      "step": 9376
+    },
+    {
+      "epoch": 1.0893987801336045,
+      "grad_norm": 0.4952625632286072,
+      "learning_rate": 0.0001,
+      "loss": 1.3823,
+      "step": 9377
+    },
+    {
+      "epoch": 1.089514957885565,
+      "grad_norm": 0.585818350315094,
+      "learning_rate": 0.0001,
+      "loss": 1.656,
+      "step": 9378
+    },
+    {
+      "epoch": 1.0896311356375254,
+      "grad_norm": 0.6007453799247742,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 9379
+    },
+    {
+      "epoch": 1.089747313389486,
+      "grad_norm": 0.5968008637428284,
+      "learning_rate": 0.0001,
+      "loss": 1.5623,
+      "step": 9380
+    },
+    {
+      "epoch": 1.0898634911414464,
+      "grad_norm": 0.5476697087287903,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 9381
+    },
+    {
+      "epoch": 1.089979668893407,
+      "grad_norm": 0.5096556544303894,
+      "learning_rate": 0.0001,
+      "loss": 1.2194,
+      "step": 9382
+    },
+    {
+      "epoch": 1.0900958466453674,
+      "grad_norm": 0.5080416798591614,
+      "learning_rate": 0.0001,
+      "loss": 1.23,
+      "step": 9383
+    },
+    {
+      "epoch": 1.090212024397328,
+      "grad_norm": 0.5420169234275818,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 9384
+    },
+    {
+      "epoch": 1.0903282021492884,
+      "grad_norm": 0.543263852596283,
+      "learning_rate": 0.0001,
+      "loss": 1.3452,
+      "step": 9385
+    },
+    {
+      "epoch": 1.0904443799012489,
+      "grad_norm": 0.500427782535553,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 9386
+    },
+    {
+      "epoch": 1.0905605576532094,
+      "grad_norm": 0.5219041109085083,
+      "learning_rate": 0.0001,
+      "loss": 1.2534,
+      "step": 9387
+    },
+    {
+      "epoch": 1.0906767354051699,
+      "grad_norm": 0.5618671774864197,
+      "learning_rate": 0.0001,
+      "loss": 1.5241,
+      "step": 9388
+    },
+    {
+      "epoch": 1.0907929131571303,
+      "grad_norm": 0.567960262298584,
+      "learning_rate": 0.0001,
+      "loss": 1.599,
+      "step": 9389
+    },
+    {
+      "epoch": 1.0909090909090908,
+      "grad_norm": 0.5592353940010071,
+      "learning_rate": 0.0001,
+      "loss": 1.4526,
+      "step": 9390
+    },
+    {
+      "epoch": 1.0910252686610513,
+      "grad_norm": 0.5943126082420349,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 9391
+    },
+    {
+      "epoch": 1.0911414464130118,
+      "grad_norm": 0.5825600028038025,
+      "learning_rate": 0.0001,
+      "loss": 1.62,
+      "step": 9392
+    },
+    {
+      "epoch": 1.0912576241649723,
+      "grad_norm": 0.5191078186035156,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 9393
+    },
+    {
+      "epoch": 1.091373801916933,
+      "grad_norm": 0.5243180394172668,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 9394
+    },
+    {
+      "epoch": 1.0914899796688935,
+      "grad_norm": 0.548667311668396,
+      "learning_rate": 0.0001,
+      "loss": 1.4179,
+      "step": 9395
+    },
+    {
+      "epoch": 1.091606157420854,
+      "grad_norm": 0.5515474081039429,
+      "learning_rate": 0.0001,
+      "loss": 1.3246,
+      "step": 9396
+    },
+    {
+      "epoch": 1.0917223351728145,
+      "grad_norm": 0.5582205057144165,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 9397
+    },
+    {
+      "epoch": 1.091838512924775,
+      "grad_norm": 0.5735296607017517,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 9398
+    },
+    {
+      "epoch": 1.0919546906767355,
+      "grad_norm": 0.5381777286529541,
+      "learning_rate": 0.0001,
+      "loss": 1.4056,
+      "step": 9399
+    },
+    {
+      "epoch": 1.092070868428696,
+      "grad_norm": 0.506504237651825,
+      "learning_rate": 0.0001,
+      "loss": 1.3763,
+      "step": 9400
+    },
+    {
+      "epoch": 1.0921870461806564,
+      "grad_norm": 0.5787509679794312,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 9401
+    },
+    {
+      "epoch": 1.092303223932617,
+      "grad_norm": 0.543312132358551,
+      "learning_rate": 0.0001,
+      "loss": 1.3438,
+      "step": 9402
+    },
+    {
+      "epoch": 1.0924194016845774,
+      "grad_norm": 0.5725640058517456,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 9403
+    },
+    {
+      "epoch": 1.092535579436538,
+      "grad_norm": 0.5353484153747559,
+      "learning_rate": 0.0001,
+      "loss": 1.2541,
+      "step": 9404
+    },
+    {
+      "epoch": 1.0926517571884984,
+      "grad_norm": 0.5454126596450806,
+      "learning_rate": 0.0001,
+      "loss": 1.3389,
+      "step": 9405
+    },
+    {
+      "epoch": 1.0927679349404589,
+      "grad_norm": 0.549069881439209,
+      "learning_rate": 0.0001,
+      "loss": 1.5135,
+      "step": 9406
+    },
+    {
+      "epoch": 1.0928841126924194,
+      "grad_norm": 0.572575032711029,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 9407
+    },
+    {
+      "epoch": 1.0930002904443799,
+      "grad_norm": 0.5477794408798218,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 9408
+    },
+    {
+      "epoch": 1.0931164681963403,
+      "grad_norm": 0.5575615167617798,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 9409
+    },
+    {
+      "epoch": 1.0932326459483008,
+      "grad_norm": 0.502451479434967,
+      "learning_rate": 0.0001,
+      "loss": 1.4112,
+      "step": 9410
+    },
+    {
+      "epoch": 1.0933488237002613,
+      "grad_norm": 0.5695579051971436,
+      "learning_rate": 0.0001,
+      "loss": 1.6426,
+      "step": 9411
+    },
+    {
+      "epoch": 1.0934650014522218,
+      "grad_norm": 0.5555123686790466,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 9412
+    },
+    {
+      "epoch": 1.0935811792041823,
+      "grad_norm": 0.5110344886779785,
+      "learning_rate": 0.0001,
+      "loss": 1.485,
+      "step": 9413
+    },
+    {
+      "epoch": 1.093697356956143,
+      "grad_norm": 0.514350414276123,
+      "learning_rate": 0.0001,
+      "loss": 1.3294,
+      "step": 9414
+    },
+    {
+      "epoch": 1.0938135347081035,
+      "grad_norm": 0.570740818977356,
+      "learning_rate": 0.0001,
+      "loss": 1.6172,
+      "step": 9415
+    },
+    {
+      "epoch": 1.093929712460064,
+      "grad_norm": 0.5143060684204102,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 9416
+    },
+    {
+      "epoch": 1.0940458902120245,
+      "grad_norm": 0.5782046914100647,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 9417
+    },
+    {
+      "epoch": 1.094162067963985,
+      "grad_norm": 0.5496023893356323,
+      "learning_rate": 0.0001,
+      "loss": 1.3474,
+      "step": 9418
+    },
+    {
+      "epoch": 1.0942782457159455,
+      "grad_norm": 0.4826601445674896,
+      "learning_rate": 0.0001,
+      "loss": 1.1978,
+      "step": 9419
+    },
+    {
+      "epoch": 1.094394423467906,
+      "grad_norm": 0.5562894940376282,
+      "learning_rate": 0.0001,
+      "loss": 1.4771,
+      "step": 9420
+    },
+    {
+      "epoch": 1.0945106012198664,
+      "grad_norm": 0.5985347032546997,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 9421
+    },
+    {
+      "epoch": 1.094626778971827,
+      "grad_norm": 0.5639268755912781,
+      "learning_rate": 0.0001,
+      "loss": 1.359,
+      "step": 9422
+    },
+    {
+      "epoch": 1.0947429567237874,
+      "grad_norm": 0.6072770357131958,
+      "learning_rate": 0.0001,
+      "loss": 1.7112,
+      "step": 9423
+    },
+    {
+      "epoch": 1.094859134475748,
+      "grad_norm": 0.5154184103012085,
+      "learning_rate": 0.0001,
+      "loss": 1.1247,
+      "step": 9424
+    },
+    {
+      "epoch": 1.0949753122277084,
+      "grad_norm": 0.6177548170089722,
+      "learning_rate": 0.0001,
+      "loss": 1.6308,
+      "step": 9425
+    },
+    {
+      "epoch": 1.0950914899796689,
+      "grad_norm": 0.5776795148849487,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 9426
+    },
+    {
+      "epoch": 1.0952076677316294,
+      "grad_norm": 0.5506322383880615,
+      "learning_rate": 0.0001,
+      "loss": 1.4099,
+      "step": 9427
+    },
+    {
+      "epoch": 1.0953238454835899,
+      "grad_norm": 0.5430233478546143,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 9428
+    },
+    {
+      "epoch": 1.0954400232355503,
+      "grad_norm": 0.5830442309379578,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 9429
+    },
+    {
+      "epoch": 1.0955562009875108,
+      "grad_norm": 0.541569173336029,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 9430
+    },
+    {
+      "epoch": 1.0956723787394713,
+      "grad_norm": 0.556121289730072,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 9431
+    },
+    {
+      "epoch": 1.0957885564914318,
+      "grad_norm": 0.5274482369422913,
+      "learning_rate": 0.0001,
+      "loss": 1.4293,
+      "step": 9432
+    },
+    {
+      "epoch": 1.0959047342433923,
+      "grad_norm": 0.5375184416770935,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 9433
+    },
+    {
+      "epoch": 1.0960209119953528,
+      "grad_norm": 0.5496751666069031,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 9434
+    },
+    {
+      "epoch": 1.0961370897473133,
+      "grad_norm": 0.5520657896995544,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 9435
+    },
+    {
+      "epoch": 1.096253267499274,
+      "grad_norm": 0.5120051503181458,
+      "learning_rate": 0.0001,
+      "loss": 1.5101,
+      "step": 9436
+    },
+    {
+      "epoch": 1.0963694452512345,
+      "grad_norm": 0.5237596035003662,
+      "learning_rate": 0.0001,
+      "loss": 1.4054,
+      "step": 9437
+    },
+    {
+      "epoch": 1.096485623003195,
+      "grad_norm": 0.5413674116134644,
+      "learning_rate": 0.0001,
+      "loss": 1.4558,
+      "step": 9438
+    },
+    {
+      "epoch": 1.0966018007551555,
+      "grad_norm": 0.5916610360145569,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 9439
+    },
+    {
+      "epoch": 1.096717978507116,
+      "grad_norm": 0.6033625602722168,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 9440
+    },
+    {
+      "epoch": 1.0968341562590764,
+      "grad_norm": 0.549626886844635,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 9441
+    },
+    {
+      "epoch": 1.096950334011037,
+      "grad_norm": 0.566194474697113,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 9442
+    },
+    {
+      "epoch": 1.0970665117629974,
+      "grad_norm": 0.5590158700942993,
+      "learning_rate": 0.0001,
+      "loss": 1.4561,
+      "step": 9443
+    },
+    {
+      "epoch": 1.097182689514958,
+      "grad_norm": 0.5413280129432678,
+      "learning_rate": 0.0001,
+      "loss": 1.4254,
+      "step": 9444
+    },
+    {
+      "epoch": 1.0972988672669184,
+      "grad_norm": 0.5623098611831665,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 9445
+    },
+    {
+      "epoch": 1.0974150450188789,
+      "grad_norm": 0.6405941843986511,
+      "learning_rate": 0.0001,
+      "loss": 1.4324,
+      "step": 9446
+    },
+    {
+      "epoch": 1.0975312227708394,
+      "grad_norm": 0.5057178735733032,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 9447
+    },
+    {
+      "epoch": 1.0976474005227999,
+      "grad_norm": 0.5987020134925842,
+      "learning_rate": 0.0001,
+      "loss": 1.308,
+      "step": 9448
+    },
+    {
+      "epoch": 1.0977635782747603,
+      "grad_norm": 0.5348968505859375,
+      "learning_rate": 0.0001,
+      "loss": 1.344,
+      "step": 9449
+    },
+    {
+      "epoch": 1.0978797560267208,
+      "grad_norm": 0.5361587405204773,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 9450
+    },
+    {
+      "epoch": 1.0979959337786813,
+      "grad_norm": 0.549802839756012,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 9451
+    },
+    {
+      "epoch": 1.0981121115306418,
+      "grad_norm": 0.5512414574623108,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 9452
+    },
+    {
+      "epoch": 1.0982282892826023,
+      "grad_norm": 0.5330852270126343,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 9453
+    },
+    {
+      "epoch": 1.0983444670345628,
+      "grad_norm": 0.5517197847366333,
+      "learning_rate": 0.0001,
+      "loss": 1.3646,
+      "step": 9454
+    },
+    {
+      "epoch": 1.0984606447865233,
+      "grad_norm": 0.5206728577613831,
+      "learning_rate": 0.0001,
+      "loss": 1.3936,
+      "step": 9455
+    },
+    {
+      "epoch": 1.098576822538484,
+      "grad_norm": 0.5265716910362244,
+      "learning_rate": 0.0001,
+      "loss": 1.3308,
+      "step": 9456
+    },
+    {
+      "epoch": 1.0986930002904445,
+      "grad_norm": 0.5415132641792297,
+      "learning_rate": 0.0001,
+      "loss": 1.385,
+      "step": 9457
+    },
+    {
+      "epoch": 1.098809178042405,
+      "grad_norm": 0.5541356801986694,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 9458
+    },
+    {
+      "epoch": 1.0989253557943655,
+      "grad_norm": 0.5190122723579407,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 9459
+    },
+    {
+      "epoch": 1.099041533546326,
+      "grad_norm": 0.554965615272522,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 9460
+    },
+    {
+      "epoch": 1.0991577112982864,
+      "grad_norm": 0.5586269497871399,
+      "learning_rate": 0.0001,
+      "loss": 1.4455,
+      "step": 9461
+    },
+    {
+      "epoch": 1.099273889050247,
+      "grad_norm": 0.5546631217002869,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 9462
+    },
+    {
+      "epoch": 1.0993900668022074,
+      "grad_norm": 0.5459680557250977,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 9463
+    },
+    {
+      "epoch": 1.099506244554168,
+      "grad_norm": 0.568166196346283,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 9464
+    },
+    {
+      "epoch": 1.0996224223061284,
+      "grad_norm": 0.5397794842720032,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 9465
+    },
+    {
+      "epoch": 1.0997386000580889,
+      "grad_norm": 0.5615697503089905,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 9466
+    },
+    {
+      "epoch": 1.0998547778100494,
+      "grad_norm": 0.6064543128013611,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 9467
+    },
+    {
+      "epoch": 1.0999709555620099,
+      "grad_norm": 0.5744174718856812,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 9468
+    },
+    {
+      "epoch": 1.1000871333139703,
+      "grad_norm": 0.5579218864440918,
+      "learning_rate": 0.0001,
+      "loss": 1.4867,
+      "step": 9469
+    },
+    {
+      "epoch": 1.1002033110659308,
+      "grad_norm": 0.5771361589431763,
+      "learning_rate": 0.0001,
+      "loss": 1.508,
+      "step": 9470
+    },
+    {
+      "epoch": 1.1003194888178913,
+      "grad_norm": 0.575631856918335,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 9471
+    },
+    {
+      "epoch": 1.1004356665698518,
+      "grad_norm": 0.5344403982162476,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 9472
+    },
+    {
+      "epoch": 1.1005518443218123,
+      "grad_norm": 0.5184001922607422,
+      "learning_rate": 0.0001,
+      "loss": 1.3486,
+      "step": 9473
+    },
+    {
+      "epoch": 1.1006680220737728,
+      "grad_norm": 0.5264585018157959,
+      "learning_rate": 0.0001,
+      "loss": 1.3588,
+      "step": 9474
+    },
+    {
+      "epoch": 1.1007841998257333,
+      "grad_norm": 0.5590055584907532,
+      "learning_rate": 0.0001,
+      "loss": 1.4734,
+      "step": 9475
+    },
+    {
+      "epoch": 1.1009003775776938,
+      "grad_norm": 0.5235846638679504,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 9476
+    },
+    {
+      "epoch": 1.1010165553296543,
+      "grad_norm": 0.5595501661300659,
+      "learning_rate": 0.0001,
+      "loss": 1.4029,
+      "step": 9477
+    },
+    {
+      "epoch": 1.101132733081615,
+      "grad_norm": 0.5435571074485779,
+      "learning_rate": 0.0001,
+      "loss": 1.4213,
+      "step": 9478
+    },
+    {
+      "epoch": 1.1012489108335755,
+      "grad_norm": 0.5264965295791626,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 9479
+    },
+    {
+      "epoch": 1.101365088585536,
+      "grad_norm": 0.529292643070221,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 9480
+    },
+    {
+      "epoch": 1.1014812663374964,
+      "grad_norm": 0.5540322661399841,
+      "learning_rate": 0.0001,
+      "loss": 1.5225,
+      "step": 9481
+    },
+    {
+      "epoch": 1.101597444089457,
+      "grad_norm": 0.5076307058334351,
+      "learning_rate": 0.0001,
+      "loss": 1.3656,
+      "step": 9482
+    },
+    {
+      "epoch": 1.1017136218414174,
+      "grad_norm": 0.5783825516700745,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 9483
+    },
+    {
+      "epoch": 1.101829799593378,
+      "grad_norm": 0.5533871650695801,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 9484
+    },
+    {
+      "epoch": 1.1019459773453384,
+      "grad_norm": 0.5388155579566956,
+      "learning_rate": 0.0001,
+      "loss": 1.4595,
+      "step": 9485
+    },
+    {
+      "epoch": 1.1020621550972989,
+      "grad_norm": 0.5467915534973145,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 9486
+    },
+    {
+      "epoch": 1.1021783328492594,
+      "grad_norm": 0.6151425242424011,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 9487
+    },
+    {
+      "epoch": 1.1022945106012199,
+      "grad_norm": 0.5877282023429871,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 9488
+    },
+    {
+      "epoch": 1.1024106883531803,
+      "grad_norm": 0.5632818937301636,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 9489
+    },
+    {
+      "epoch": 1.1025268661051408,
+      "grad_norm": 0.5703235268592834,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 9490
+    },
+    {
+      "epoch": 1.1026430438571013,
+      "grad_norm": 0.563475489616394,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 9491
+    },
+    {
+      "epoch": 1.1027592216090618,
+      "grad_norm": 0.5732678174972534,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 9492
+    },
+    {
+      "epoch": 1.1028753993610223,
+      "grad_norm": 0.5850802659988403,
+      "learning_rate": 0.0001,
+      "loss": 1.4957,
+      "step": 9493
+    },
+    {
+      "epoch": 1.1029915771129828,
+      "grad_norm": 0.5793954133987427,
+      "learning_rate": 0.0001,
+      "loss": 1.583,
+      "step": 9494
+    },
+    {
+      "epoch": 1.1031077548649433,
+      "grad_norm": 0.5984706878662109,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 9495
+    },
+    {
+      "epoch": 1.1032239326169038,
+      "grad_norm": 0.5244345664978027,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 9496
+    },
+    {
+      "epoch": 1.1033401103688645,
+      "grad_norm": 0.5405702590942383,
+      "learning_rate": 0.0001,
+      "loss": 1.359,
+      "step": 9497
+    },
+    {
+      "epoch": 1.103456288120825,
+      "grad_norm": 0.5445573329925537,
+      "learning_rate": 0.0001,
+      "loss": 1.615,
+      "step": 9498
+    },
+    {
+      "epoch": 1.1035724658727855,
+      "grad_norm": 0.5632995367050171,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 9499
+    },
+    {
+      "epoch": 1.103688643624746,
+      "grad_norm": 0.5281904935836792,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 9500
+    },
+    {
+      "epoch": 1.1038048213767064,
+      "grad_norm": 0.5330088138580322,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 9501
+    },
+    {
+      "epoch": 1.103920999128667,
+      "grad_norm": 0.514532744884491,
+      "learning_rate": 0.0001,
+      "loss": 1.4414,
+      "step": 9502
+    },
+    {
+      "epoch": 1.1040371768806274,
+      "grad_norm": 0.548172652721405,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 9503
+    },
+    {
+      "epoch": 1.104153354632588,
+      "grad_norm": 0.5549656748771667,
+      "learning_rate": 0.0001,
+      "loss": 1.3786,
+      "step": 9504
+    },
+    {
+      "epoch": 1.1042695323845484,
+      "grad_norm": 0.5625692009925842,
+      "learning_rate": 0.0001,
+      "loss": 1.5153,
+      "step": 9505
+    },
+    {
+      "epoch": 1.1043857101365089,
+      "grad_norm": 0.5380663871765137,
+      "learning_rate": 0.0001,
+      "loss": 1.3388,
+      "step": 9506
+    },
+    {
+      "epoch": 1.1045018878884694,
+      "grad_norm": 0.530598521232605,
+      "learning_rate": 0.0001,
+      "loss": 1.3424,
+      "step": 9507
+    },
+    {
+      "epoch": 1.1046180656404299,
+      "grad_norm": 0.5374849438667297,
+      "learning_rate": 0.0001,
+      "loss": 1.4022,
+      "step": 9508
+    },
+    {
+      "epoch": 1.1047342433923903,
+      "grad_norm": 0.5946294665336609,
+      "learning_rate": 0.0001,
+      "loss": 1.6452,
+      "step": 9509
+    },
+    {
+      "epoch": 1.1048504211443508,
+      "grad_norm": 0.5669423937797546,
+      "learning_rate": 0.0001,
+      "loss": 1.3185,
+      "step": 9510
+    },
+    {
+      "epoch": 1.1049665988963113,
+      "grad_norm": 0.6064764261245728,
+      "learning_rate": 0.0001,
+      "loss": 1.4928,
+      "step": 9511
+    },
+    {
+      "epoch": 1.1050827766482718,
+      "grad_norm": 0.5889172554016113,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 9512
+    },
+    {
+      "epoch": 1.1051989544002323,
+      "grad_norm": 0.5890141725540161,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 9513
+    },
+    {
+      "epoch": 1.1053151321521928,
+      "grad_norm": 0.5455361604690552,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 9514
+    },
+    {
+      "epoch": 1.1054313099041533,
+      "grad_norm": 0.5481176972389221,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 9515
+    },
+    {
+      "epoch": 1.1055474876561138,
+      "grad_norm": 0.602695882320404,
+      "learning_rate": 0.0001,
+      "loss": 1.6609,
+      "step": 9516
+    },
+    {
+      "epoch": 1.1056636654080743,
+      "grad_norm": 0.5367211103439331,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 9517
+    },
+    {
+      "epoch": 1.1057798431600347,
+      "grad_norm": 0.5153917670249939,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 9518
+    },
+    {
+      "epoch": 1.1058960209119952,
+      "grad_norm": 0.523540735244751,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 9519
+    },
+    {
+      "epoch": 1.106012198663956,
+      "grad_norm": 0.5269156098365784,
+      "learning_rate": 0.0001,
+      "loss": 1.4749,
+      "step": 9520
+    },
+    {
+      "epoch": 1.1061283764159164,
+      "grad_norm": 0.5302053689956665,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 9521
+    },
+    {
+      "epoch": 1.106244554167877,
+      "grad_norm": 0.5808258652687073,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 9522
+    },
+    {
+      "epoch": 1.1063607319198374,
+      "grad_norm": 0.6308403611183167,
+      "learning_rate": 0.0001,
+      "loss": 1.5731,
+      "step": 9523
+    },
+    {
+      "epoch": 1.106476909671798,
+      "grad_norm": 0.6031569838523865,
+      "learning_rate": 0.0001,
+      "loss": 1.5984,
+      "step": 9524
+    },
+    {
+      "epoch": 1.1065930874237584,
+      "grad_norm": 0.5299091339111328,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 9525
+    },
+    {
+      "epoch": 1.1067092651757189,
+      "grad_norm": 0.5627480745315552,
+      "learning_rate": 0.0001,
+      "loss": 1.6189,
+      "step": 9526
+    },
+    {
+      "epoch": 1.1068254429276794,
+      "grad_norm": 0.5670896172523499,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 9527
+    },
+    {
+      "epoch": 1.1069416206796399,
+      "grad_norm": 0.5727272629737854,
+      "learning_rate": 0.0001,
+      "loss": 1.3688,
+      "step": 9528
+    },
+    {
+      "epoch": 1.1070577984316003,
+      "grad_norm": 0.4948303699493408,
+      "learning_rate": 0.0001,
+      "loss": 1.3057,
+      "step": 9529
+    },
+    {
+      "epoch": 1.1071739761835608,
+      "grad_norm": 0.5203784108161926,
+      "learning_rate": 0.0001,
+      "loss": 1.4601,
+      "step": 9530
+    },
+    {
+      "epoch": 1.1072901539355213,
+      "grad_norm": 0.5920235514640808,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 9531
+    },
+    {
+      "epoch": 1.1074063316874818,
+      "grad_norm": 0.5513145923614502,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 9532
+    },
+    {
+      "epoch": 1.1075225094394423,
+      "grad_norm": 0.5774796605110168,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 9533
+    },
+    {
+      "epoch": 1.1076386871914028,
+      "grad_norm": 0.5905603170394897,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 9534
+    },
+    {
+      "epoch": 1.1077548649433633,
+      "grad_norm": 0.5975148677825928,
+      "learning_rate": 0.0001,
+      "loss": 1.5869,
+      "step": 9535
+    },
+    {
+      "epoch": 1.1078710426953238,
+      "grad_norm": 0.5394390821456909,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 9536
+    },
+    {
+      "epoch": 1.1079872204472843,
+      "grad_norm": 0.5816001296043396,
+      "learning_rate": 0.0001,
+      "loss": 1.4818,
+      "step": 9537
+    },
+    {
+      "epoch": 1.1081033981992447,
+      "grad_norm": 0.5599253177642822,
+      "learning_rate": 0.0001,
+      "loss": 1.3033,
+      "step": 9538
+    },
+    {
+      "epoch": 1.1082195759512055,
+      "grad_norm": 0.5270270109176636,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 9539
+    },
+    {
+      "epoch": 1.108335753703166,
+      "grad_norm": 0.5689064264297485,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 9540
+    },
+    {
+      "epoch": 1.1084519314551264,
+      "grad_norm": 0.5355569124221802,
+      "learning_rate": 0.0001,
+      "loss": 1.6211,
+      "step": 9541
+    },
+    {
+      "epoch": 1.108568109207087,
+      "grad_norm": 0.5691352486610413,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 9542
+    },
+    {
+      "epoch": 1.1086842869590474,
+      "grad_norm": 0.5562692284584045,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 9543
+    },
+    {
+      "epoch": 1.108800464711008,
+      "grad_norm": 0.5737292766571045,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 9544
+    },
+    {
+      "epoch": 1.1089166424629684,
+      "grad_norm": 0.5629547238349915,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 9545
+    },
+    {
+      "epoch": 1.1090328202149289,
+      "grad_norm": 0.5484119057655334,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 9546
+    },
+    {
+      "epoch": 1.1091489979668894,
+      "grad_norm": 0.5538215041160583,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 9547
+    },
+    {
+      "epoch": 1.1092651757188499,
+      "grad_norm": 0.5706968903541565,
+      "learning_rate": 0.0001,
+      "loss": 1.7345,
+      "step": 9548
+    },
+    {
+      "epoch": 1.1093813534708103,
+      "grad_norm": 0.5173110961914062,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 9549
+    },
+    {
+      "epoch": 1.1094975312227708,
+      "grad_norm": 0.5889958143234253,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 9550
+    },
+    {
+      "epoch": 1.1096137089747313,
+      "grad_norm": 0.5318642258644104,
+      "learning_rate": 0.0001,
+      "loss": 1.4558,
+      "step": 9551
+    },
+    {
+      "epoch": 1.1097298867266918,
+      "grad_norm": 0.5898750424385071,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 9552
+    },
+    {
+      "epoch": 1.1098460644786523,
+      "grad_norm": 0.5130350589752197,
+      "learning_rate": 0.0001,
+      "loss": 1.3356,
+      "step": 9553
+    },
+    {
+      "epoch": 1.1099622422306128,
+      "grad_norm": 0.5386868119239807,
+      "learning_rate": 0.0001,
+      "loss": 1.6305,
+      "step": 9554
+    },
+    {
+      "epoch": 1.1100784199825733,
+      "grad_norm": 0.5684298872947693,
+      "learning_rate": 0.0001,
+      "loss": 1.4623,
+      "step": 9555
+    },
+    {
+      "epoch": 1.1101945977345338,
+      "grad_norm": 0.5544883012771606,
+      "learning_rate": 0.0001,
+      "loss": 1.5781,
+      "step": 9556
+    },
+    {
+      "epoch": 1.1103107754864943,
+      "grad_norm": 0.5304974913597107,
+      "learning_rate": 0.0001,
+      "loss": 1.3058,
+      "step": 9557
+    },
+    {
+      "epoch": 1.1104269532384548,
+      "grad_norm": 0.5397143363952637,
+      "learning_rate": 0.0001,
+      "loss": 1.4132,
+      "step": 9558
+    },
+    {
+      "epoch": 1.1105431309904152,
+      "grad_norm": 0.5635391473770142,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 9559
+    },
+    {
+      "epoch": 1.1106593087423757,
+      "grad_norm": 0.5330328345298767,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 9560
+    },
+    {
+      "epoch": 1.1107754864943362,
+      "grad_norm": 0.5222686529159546,
+      "learning_rate": 0.0001,
+      "loss": 1.2798,
+      "step": 9561
+    },
+    {
+      "epoch": 1.110891664246297,
+      "grad_norm": 0.5562769770622253,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 9562
+    },
+    {
+      "epoch": 1.1110078419982574,
+      "grad_norm": 0.5504674911499023,
+      "learning_rate": 0.0001,
+      "loss": 1.5051,
+      "step": 9563
+    },
+    {
+      "epoch": 1.111124019750218,
+      "grad_norm": 0.5468124747276306,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 9564
+    },
+    {
+      "epoch": 1.1112401975021784,
+      "grad_norm": 0.5254167914390564,
+      "learning_rate": 0.0001,
+      "loss": 1.5379,
+      "step": 9565
+    },
+    {
+      "epoch": 1.1113563752541389,
+      "grad_norm": 0.5691385269165039,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 9566
+    },
+    {
+      "epoch": 1.1114725530060994,
+      "grad_norm": 0.570458710193634,
+      "learning_rate": 0.0001,
+      "loss": 1.4127,
+      "step": 9567
+    },
+    {
+      "epoch": 1.1115887307580599,
+      "grad_norm": 0.5755772590637207,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 9568
+    },
+    {
+      "epoch": 1.1117049085100204,
+      "grad_norm": 0.5539082288742065,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 9569
+    },
+    {
+      "epoch": 1.1118210862619808,
+      "grad_norm": 0.5218854546546936,
+      "learning_rate": 0.0001,
+      "loss": 1.2754,
+      "step": 9570
+    },
+    {
+      "epoch": 1.1119372640139413,
+      "grad_norm": 0.5608935952186584,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 9571
+    },
+    {
+      "epoch": 1.1120534417659018,
+      "grad_norm": 0.5075331330299377,
+      "learning_rate": 0.0001,
+      "loss": 1.2668,
+      "step": 9572
+    },
+    {
+      "epoch": 1.1121696195178623,
+      "grad_norm": 0.5797237753868103,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 9573
+    },
+    {
+      "epoch": 1.1122857972698228,
+      "grad_norm": 0.5733375549316406,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 9574
+    },
+    {
+      "epoch": 1.1124019750217833,
+      "grad_norm": 0.5502446293830872,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 9575
+    },
+    {
+      "epoch": 1.1125181527737438,
+      "grad_norm": 0.5844667553901672,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 9576
+    },
+    {
+      "epoch": 1.1126343305257043,
+      "grad_norm": 0.5160120129585266,
+      "learning_rate": 0.0001,
+      "loss": 1.3696,
+      "step": 9577
+    },
+    {
+      "epoch": 1.1127505082776648,
+      "grad_norm": 0.5747503042221069,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 9578
+    },
+    {
+      "epoch": 1.1128666860296252,
+      "grad_norm": 0.5804192423820496,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 9579
+    },
+    {
+      "epoch": 1.1129828637815857,
+      "grad_norm": 0.5319089293479919,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 9580
+    },
+    {
+      "epoch": 1.1130990415335464,
+      "grad_norm": 0.5397303104400635,
+      "learning_rate": 0.0001,
+      "loss": 1.3097,
+      "step": 9581
+    },
+    {
+      "epoch": 1.113215219285507,
+      "grad_norm": 0.5700457692146301,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 9582
+    },
+    {
+      "epoch": 1.1133313970374674,
+      "grad_norm": 0.5417009592056274,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 9583
+    },
+    {
+      "epoch": 1.113447574789428,
+      "grad_norm": 0.573021650314331,
+      "learning_rate": 0.0001,
+      "loss": 1.4537,
+      "step": 9584
+    },
+    {
+      "epoch": 1.1135637525413884,
+      "grad_norm": 0.5706760883331299,
+      "learning_rate": 0.0001,
+      "loss": 1.7165,
+      "step": 9585
+    },
+    {
+      "epoch": 1.1136799302933489,
+      "grad_norm": 0.5758902430534363,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 9586
+    },
+    {
+      "epoch": 1.1137961080453094,
+      "grad_norm": 0.541764497756958,
+      "learning_rate": 0.0001,
+      "loss": 1.3593,
+      "step": 9587
+    },
+    {
+      "epoch": 1.1139122857972699,
+      "grad_norm": 0.5544206500053406,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 9588
+    },
+    {
+      "epoch": 1.1140284635492304,
+      "grad_norm": 0.6012476086616516,
+      "learning_rate": 0.0001,
+      "loss": 1.6206,
+      "step": 9589
+    },
+    {
+      "epoch": 1.1141446413011908,
+      "grad_norm": 0.5391044020652771,
+      "learning_rate": 0.0001,
+      "loss": 1.2987,
+      "step": 9590
+    },
+    {
+      "epoch": 1.1142608190531513,
+      "grad_norm": 0.547203540802002,
+      "learning_rate": 0.0001,
+      "loss": 1.5008,
+      "step": 9591
+    },
+    {
+      "epoch": 1.1143769968051118,
+      "grad_norm": 0.5447909832000732,
+      "learning_rate": 0.0001,
+      "loss": 1.4603,
+      "step": 9592
+    },
+    {
+      "epoch": 1.1144931745570723,
+      "grad_norm": 0.5598756074905396,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 9593
+    },
+    {
+      "epoch": 1.1146093523090328,
+      "grad_norm": 0.5283433198928833,
+      "learning_rate": 0.0001,
+      "loss": 1.2927,
+      "step": 9594
+    },
+    {
+      "epoch": 1.1147255300609933,
+      "grad_norm": 0.5749019384384155,
+      "learning_rate": 0.0001,
+      "loss": 1.4112,
+      "step": 9595
+    },
+    {
+      "epoch": 1.1148417078129538,
+      "grad_norm": 0.5759902000427246,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 9596
+    },
+    {
+      "epoch": 1.1149578855649143,
+      "grad_norm": 0.5656595230102539,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 9597
+    },
+    {
+      "epoch": 1.1150740633168748,
+      "grad_norm": 0.5748841762542725,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 9598
+    },
+    {
+      "epoch": 1.1151902410688352,
+      "grad_norm": 0.4959535002708435,
+      "learning_rate": 0.0001,
+      "loss": 1.3178,
+      "step": 9599
+    },
+    {
+      "epoch": 1.1153064188207957,
+      "grad_norm": 0.5196616053581238,
+      "learning_rate": 0.0001,
+      "loss": 1.36,
+      "step": 9600
+    },
+    {
+      "epoch": 1.1154225965727562,
+      "grad_norm": 0.5188265442848206,
+      "learning_rate": 0.0001,
+      "loss": 1.4426,
+      "step": 9601
+    },
+    {
+      "epoch": 1.1155387743247167,
+      "grad_norm": 0.5687859058380127,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 9602
+    },
+    {
+      "epoch": 1.1156549520766774,
+      "grad_norm": 0.5094379186630249,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 9603
+    },
+    {
+      "epoch": 1.115771129828638,
+      "grad_norm": 0.5415772199630737,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 9604
+    },
+    {
+      "epoch": 1.1158873075805984,
+      "grad_norm": 0.5665364861488342,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 9605
+    },
+    {
+      "epoch": 1.1160034853325589,
+      "grad_norm": 0.5534007549285889,
+      "learning_rate": 0.0001,
+      "loss": 1.4372,
+      "step": 9606
+    },
+    {
+      "epoch": 1.1161196630845194,
+      "grad_norm": 0.5435113310813904,
+      "learning_rate": 0.0001,
+      "loss": 1.3756,
+      "step": 9607
+    },
+    {
+      "epoch": 1.1162358408364799,
+      "grad_norm": 0.5332193374633789,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 9608
+    },
+    {
+      "epoch": 1.1163520185884404,
+      "grad_norm": 0.5340940952301025,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 9609
+    },
+    {
+      "epoch": 1.1164681963404008,
+      "grad_norm": 0.5510147213935852,
+      "learning_rate": 0.0001,
+      "loss": 1.4416,
+      "step": 9610
+    },
+    {
+      "epoch": 1.1165843740923613,
+      "grad_norm": 0.5450196266174316,
+      "learning_rate": 0.0001,
+      "loss": 1.5096,
+      "step": 9611
+    },
+    {
+      "epoch": 1.1167005518443218,
+      "grad_norm": 0.5567142367362976,
+      "learning_rate": 0.0001,
+      "loss": 1.5802,
+      "step": 9612
+    },
+    {
+      "epoch": 1.1168167295962823,
+      "grad_norm": 0.5469493865966797,
+      "learning_rate": 0.0001,
+      "loss": 1.6743,
+      "step": 9613
+    },
+    {
+      "epoch": 1.1169329073482428,
+      "grad_norm": 0.5908190011978149,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 9614
+    },
+    {
+      "epoch": 1.1170490851002033,
+      "grad_norm": 0.5396865010261536,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 9615
+    },
+    {
+      "epoch": 1.1171652628521638,
+      "grad_norm": 0.5347968339920044,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 9616
+    },
+    {
+      "epoch": 1.1172814406041243,
+      "grad_norm": 0.5435789823532104,
+      "learning_rate": 0.0001,
+      "loss": 1.4568,
+      "step": 9617
+    },
+    {
+      "epoch": 1.1173976183560848,
+      "grad_norm": 0.530580461025238,
+      "learning_rate": 0.0001,
+      "loss": 1.3363,
+      "step": 9618
+    },
+    {
+      "epoch": 1.1175137961080452,
+      "grad_norm": 0.5703923106193542,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 9619
+    },
+    {
+      "epoch": 1.1176299738600057,
+      "grad_norm": 0.5795918703079224,
+      "learning_rate": 0.0001,
+      "loss": 1.6841,
+      "step": 9620
+    },
+    {
+      "epoch": 1.1177461516119662,
+      "grad_norm": 0.5287982225418091,
+      "learning_rate": 0.0001,
+      "loss": 1.4183,
+      "step": 9621
+    },
+    {
+      "epoch": 1.1178623293639267,
+      "grad_norm": 0.5755120515823364,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 9622
+    },
+    {
+      "epoch": 1.1179785071158874,
+      "grad_norm": 0.569776713848114,
+      "learning_rate": 0.0001,
+      "loss": 1.6366,
+      "step": 9623
+    },
+    {
+      "epoch": 1.118094684867848,
+      "grad_norm": 0.5526320338249207,
+      "learning_rate": 0.0001,
+      "loss": 1.4253,
+      "step": 9624
+    },
+    {
+      "epoch": 1.1182108626198084,
+      "grad_norm": 0.5705094337463379,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 9625
+    },
+    {
+      "epoch": 1.1183270403717689,
+      "grad_norm": 0.5555620789527893,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 9626
+    },
+    {
+      "epoch": 1.1184432181237294,
+      "grad_norm": 0.5216465592384338,
+      "learning_rate": 0.0001,
+      "loss": 1.3455,
+      "step": 9627
+    },
+    {
+      "epoch": 1.1185593958756899,
+      "grad_norm": 0.5284774899482727,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 9628
+    },
+    {
+      "epoch": 1.1186755736276504,
+      "grad_norm": 0.47694408893585205,
+      "learning_rate": 0.0001,
+      "loss": 1.2213,
+      "step": 9629
+    },
+    {
+      "epoch": 1.1187917513796108,
+      "grad_norm": 0.547134280204773,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 9630
+    },
+    {
+      "epoch": 1.1189079291315713,
+      "grad_norm": 0.5496186017990112,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 9631
+    },
+    {
+      "epoch": 1.1190241068835318,
+      "grad_norm": 0.6156900525093079,
+      "learning_rate": 0.0001,
+      "loss": 1.553,
+      "step": 9632
+    },
+    {
+      "epoch": 1.1191402846354923,
+      "grad_norm": 0.6084433794021606,
+      "learning_rate": 0.0001,
+      "loss": 1.5547,
+      "step": 9633
+    },
+    {
+      "epoch": 1.1192564623874528,
+      "grad_norm": 0.5293398499488831,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 9634
+    },
+    {
+      "epoch": 1.1193726401394133,
+      "grad_norm": 0.550264835357666,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 9635
+    },
+    {
+      "epoch": 1.1194888178913738,
+      "grad_norm": 0.5736042261123657,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 9636
+    },
+    {
+      "epoch": 1.1196049956433343,
+      "grad_norm": 0.5628027319908142,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 9637
+    },
+    {
+      "epoch": 1.1197211733952948,
+      "grad_norm": 0.5793375372886658,
+      "learning_rate": 0.0001,
+      "loss": 1.351,
+      "step": 9638
+    },
+    {
+      "epoch": 1.1198373511472552,
+      "grad_norm": 0.5381454229354858,
+      "learning_rate": 0.0001,
+      "loss": 1.4153,
+      "step": 9639
+    },
+    {
+      "epoch": 1.1199535288992157,
+      "grad_norm": 0.48907560110092163,
+      "learning_rate": 0.0001,
+      "loss": 1.3104,
+      "step": 9640
+    },
+    {
+      "epoch": 1.1200697066511762,
+      "grad_norm": 0.5359677672386169,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 9641
+    },
+    {
+      "epoch": 1.1201858844031367,
+      "grad_norm": 0.5884495377540588,
+      "learning_rate": 0.0001,
+      "loss": 1.4382,
+      "step": 9642
+    },
+    {
+      "epoch": 1.1203020621550972,
+      "grad_norm": 0.5695892572402954,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 9643
+    },
+    {
+      "epoch": 1.1204182399070577,
+      "grad_norm": 0.5442866683006287,
+      "learning_rate": 0.0001,
+      "loss": 1.2676,
+      "step": 9644
+    },
+    {
+      "epoch": 1.1205344176590184,
+      "grad_norm": 0.5650526881217957,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 9645
+    },
+    {
+      "epoch": 1.120650595410979,
+      "grad_norm": 0.6149001717567444,
+      "learning_rate": 0.0001,
+      "loss": 1.586,
+      "step": 9646
+    },
+    {
+      "epoch": 1.1207667731629394,
+      "grad_norm": 0.5208737254142761,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 9647
+    },
+    {
+      "epoch": 1.1208829509148999,
+      "grad_norm": 0.5633329749107361,
+      "learning_rate": 0.0001,
+      "loss": 1.4319,
+      "step": 9648
+    },
+    {
+      "epoch": 1.1209991286668604,
+      "grad_norm": 0.524625301361084,
+      "learning_rate": 0.0001,
+      "loss": 1.3997,
+      "step": 9649
+    },
+    {
+      "epoch": 1.1211153064188208,
+      "grad_norm": 0.5944722890853882,
+      "learning_rate": 0.0001,
+      "loss": 1.6914,
+      "step": 9650
+    },
+    {
+      "epoch": 1.1212314841707813,
+      "grad_norm": 0.5723394155502319,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 9651
+    },
+    {
+      "epoch": 1.1213476619227418,
+      "grad_norm": 0.5389914512634277,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 9652
+    },
+    {
+      "epoch": 1.1214638396747023,
+      "grad_norm": 0.5923259854316711,
+      "learning_rate": 0.0001,
+      "loss": 1.6293,
+      "step": 9653
+    },
+    {
+      "epoch": 1.1215800174266628,
+      "grad_norm": 0.562023401260376,
+      "learning_rate": 0.0001,
+      "loss": 1.5501,
+      "step": 9654
+    },
+    {
+      "epoch": 1.1216961951786233,
+      "grad_norm": 0.5436960458755493,
+      "learning_rate": 0.0001,
+      "loss": 1.4251,
+      "step": 9655
+    },
+    {
+      "epoch": 1.1218123729305838,
+      "grad_norm": 0.5693610310554504,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 9656
+    },
+    {
+      "epoch": 1.1219285506825443,
+      "grad_norm": 0.5483703017234802,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 9657
+    },
+    {
+      "epoch": 1.1220447284345048,
+      "grad_norm": 0.5604730844497681,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 9658
+    },
+    {
+      "epoch": 1.1221609061864652,
+      "grad_norm": 0.5464431643486023,
+      "learning_rate": 0.0001,
+      "loss": 1.3452,
+      "step": 9659
+    },
+    {
+      "epoch": 1.1222770839384257,
+      "grad_norm": 0.6087496280670166,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 9660
+    },
+    {
+      "epoch": 1.1223932616903862,
+      "grad_norm": 0.5438372492790222,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 9661
+    },
+    {
+      "epoch": 1.1225094394423467,
+      "grad_norm": 0.5792588591575623,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 9662
+    },
+    {
+      "epoch": 1.1226256171943072,
+      "grad_norm": 0.5307050347328186,
+      "learning_rate": 0.0001,
+      "loss": 1.3508,
+      "step": 9663
+    },
+    {
+      "epoch": 1.1227417949462677,
+      "grad_norm": 0.5792557001113892,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 9664
+    },
+    {
+      "epoch": 1.1228579726982284,
+      "grad_norm": 0.568598747253418,
+      "learning_rate": 0.0001,
+      "loss": 1.4901,
+      "step": 9665
+    },
+    {
+      "epoch": 1.122974150450189,
+      "grad_norm": 0.5365720987319946,
+      "learning_rate": 0.0001,
+      "loss": 1.3632,
+      "step": 9666
+    },
+    {
+      "epoch": 1.1230903282021494,
+      "grad_norm": 0.5844710469245911,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 9667
+    },
+    {
+      "epoch": 1.1232065059541099,
+      "grad_norm": 0.5613291263580322,
+      "learning_rate": 0.0001,
+      "loss": 1.3457,
+      "step": 9668
+    },
+    {
+      "epoch": 1.1233226837060704,
+      "grad_norm": 0.5521247386932373,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 9669
+    },
+    {
+      "epoch": 1.1234388614580308,
+      "grad_norm": 0.5522918701171875,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 9670
+    },
+    {
+      "epoch": 1.1235550392099913,
+      "grad_norm": 0.5793119072914124,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 9671
+    },
+    {
+      "epoch": 1.1236712169619518,
+      "grad_norm": 0.5977482199668884,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 9672
+    },
+    {
+      "epoch": 1.1237873947139123,
+      "grad_norm": 0.5508487224578857,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 9673
+    },
+    {
+      "epoch": 1.1239035724658728,
+      "grad_norm": 0.5501763820648193,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 9674
+    },
+    {
+      "epoch": 1.1240197502178333,
+      "grad_norm": 0.6087480187416077,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 9675
+    },
+    {
+      "epoch": 1.1241359279697938,
+      "grad_norm": 0.580289900302887,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 9676
+    },
+    {
+      "epoch": 1.1242521057217543,
+      "grad_norm": 0.5434087514877319,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 9677
+    },
+    {
+      "epoch": 1.1243682834737148,
+      "grad_norm": 0.5863969922065735,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 9678
+    },
+    {
+      "epoch": 1.1244844612256752,
+      "grad_norm": 0.5805991291999817,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 9679
+    },
+    {
+      "epoch": 1.1246006389776357,
+      "grad_norm": 0.562703013420105,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 9680
+    },
+    {
+      "epoch": 1.1247168167295962,
+      "grad_norm": 0.5464785695075989,
+      "learning_rate": 0.0001,
+      "loss": 1.3219,
+      "step": 9681
+    },
+    {
+      "epoch": 1.1248329944815567,
+      "grad_norm": 0.5555753707885742,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 9682
+    },
+    {
+      "epoch": 1.1249491722335172,
+      "grad_norm": 0.5941157937049866,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 9683
+    },
+    {
+      "epoch": 1.1250653499854777,
+      "grad_norm": 0.6653990745544434,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 9684
+    },
+    {
+      "epoch": 1.1251815277374382,
+      "grad_norm": 0.5832642912864685,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 9685
+    },
+    {
+      "epoch": 1.1252977054893987,
+      "grad_norm": 0.5274986028671265,
+      "learning_rate": 0.0001,
+      "loss": 1.493,
+      "step": 9686
+    },
+    {
+      "epoch": 1.1254138832413592,
+      "grad_norm": 0.5797598361968994,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 9687
+    },
+    {
+      "epoch": 1.1255300609933199,
+      "grad_norm": 0.5291623473167419,
+      "learning_rate": 0.0001,
+      "loss": 1.4156,
+      "step": 9688
+    },
+    {
+      "epoch": 1.1256462387452804,
+      "grad_norm": 0.5337206125259399,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 9689
+    },
+    {
+      "epoch": 1.1257624164972408,
+      "grad_norm": 0.5710740685462952,
+      "learning_rate": 0.0001,
+      "loss": 1.7177,
+      "step": 9690
+    },
+    {
+      "epoch": 1.1258785942492013,
+      "grad_norm": 0.5532470345497131,
+      "learning_rate": 0.0001,
+      "loss": 1.6986,
+      "step": 9691
+    },
+    {
+      "epoch": 1.1259947720011618,
+      "grad_norm": 0.5095158219337463,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 9692
+    },
+    {
+      "epoch": 1.1261109497531223,
+      "grad_norm": 0.5526217222213745,
+      "learning_rate": 0.0001,
+      "loss": 1.5832,
+      "step": 9693
+    },
+    {
+      "epoch": 1.1262271275050828,
+      "grad_norm": 0.5839323997497559,
+      "learning_rate": 0.0001,
+      "loss": 1.6062,
+      "step": 9694
+    },
+    {
+      "epoch": 1.1263433052570433,
+      "grad_norm": 0.5310125350952148,
+      "learning_rate": 0.0001,
+      "loss": 1.273,
+      "step": 9695
+    },
+    {
+      "epoch": 1.1264594830090038,
+      "grad_norm": 0.5890069007873535,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 9696
+    },
+    {
+      "epoch": 1.1265756607609643,
+      "grad_norm": 0.5359507203102112,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 9697
+    },
+    {
+      "epoch": 1.1266918385129248,
+      "grad_norm": 0.5606240630149841,
+      "learning_rate": 0.0001,
+      "loss": 1.4849,
+      "step": 9698
+    },
+    {
+      "epoch": 1.1268080162648852,
+      "grad_norm": 0.534092128276825,
+      "learning_rate": 0.0001,
+      "loss": 1.367,
+      "step": 9699
+    },
+    {
+      "epoch": 1.1269241940168457,
+      "grad_norm": 0.553125262260437,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 9700
+    },
+    {
+      "epoch": 1.1270403717688062,
+      "grad_norm": 0.5301721096038818,
+      "learning_rate": 0.0001,
+      "loss": 1.3242,
+      "step": 9701
+    },
+    {
+      "epoch": 1.1271565495207667,
+      "grad_norm": 0.6657614707946777,
+      "learning_rate": 0.0001,
+      "loss": 1.6773,
+      "step": 9702
+    },
+    {
+      "epoch": 1.1272727272727272,
+      "grad_norm": 0.541265606880188,
+      "learning_rate": 0.0001,
+      "loss": 1.3735,
+      "step": 9703
+    },
+    {
+      "epoch": 1.1273889050246877,
+      "grad_norm": 0.5840449333190918,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 9704
+    },
+    {
+      "epoch": 1.1275050827766482,
+      "grad_norm": 0.5938639044761658,
+      "learning_rate": 0.0001,
+      "loss": 1.6064,
+      "step": 9705
+    },
+    {
+      "epoch": 1.127621260528609,
+      "grad_norm": 0.5215795040130615,
+      "learning_rate": 0.0001,
+      "loss": 1.5404,
+      "step": 9706
+    },
+    {
+      "epoch": 1.1277374382805694,
+      "grad_norm": 0.537142276763916,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 9707
+    },
+    {
+      "epoch": 1.1278536160325299,
+      "grad_norm": 0.6039302349090576,
+      "learning_rate": 0.0001,
+      "loss": 1.5017,
+      "step": 9708
+    },
+    {
+      "epoch": 1.1279697937844904,
+      "grad_norm": 0.510535478591919,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 9709
+    },
+    {
+      "epoch": 1.1280859715364508,
+      "grad_norm": 0.5227552056312561,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 9710
+    },
+    {
+      "epoch": 1.1282021492884113,
+      "grad_norm": 0.5117825865745544,
+      "learning_rate": 0.0001,
+      "loss": 1.386,
+      "step": 9711
+    },
+    {
+      "epoch": 1.1283183270403718,
+      "grad_norm": 0.5713513493537903,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 9712
+    },
+    {
+      "epoch": 1.1284345047923323,
+      "grad_norm": 0.6017709374427795,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 9713
+    },
+    {
+      "epoch": 1.1285506825442928,
+      "grad_norm": 0.5840389728546143,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 9714
+    },
+    {
+      "epoch": 1.1286668602962533,
+      "grad_norm": 0.5689877867698669,
+      "learning_rate": 0.0001,
+      "loss": 1.4169,
+      "step": 9715
+    },
+    {
+      "epoch": 1.1287830380482138,
+      "grad_norm": 0.603010356426239,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 9716
+    },
+    {
+      "epoch": 1.1288992158001743,
+      "grad_norm": 0.5397545695304871,
+      "learning_rate": 0.0001,
+      "loss": 1.407,
+      "step": 9717
+    },
+    {
+      "epoch": 1.1290153935521348,
+      "grad_norm": 0.5655140280723572,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 9718
+    },
+    {
+      "epoch": 1.1291315713040952,
+      "grad_norm": 0.5146525502204895,
+      "learning_rate": 0.0001,
+      "loss": 1.4001,
+      "step": 9719
+    },
+    {
+      "epoch": 1.1292477490560557,
+      "grad_norm": 0.6142404675483704,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 9720
+    },
+    {
+      "epoch": 1.1293639268080162,
+      "grad_norm": 0.5134833455085754,
+      "learning_rate": 0.0001,
+      "loss": 1.3542,
+      "step": 9721
+    },
+    {
+      "epoch": 1.1294801045599767,
+      "grad_norm": 0.5315486788749695,
+      "learning_rate": 0.0001,
+      "loss": 1.3618,
+      "step": 9722
+    },
+    {
+      "epoch": 1.1295962823119372,
+      "grad_norm": 0.5732195973396301,
+      "learning_rate": 0.0001,
+      "loss": 1.52,
+      "step": 9723
+    },
+    {
+      "epoch": 1.1297124600638977,
+      "grad_norm": 0.5358231067657471,
+      "learning_rate": 0.0001,
+      "loss": 1.4426,
+      "step": 9724
+    },
+    {
+      "epoch": 1.1298286378158582,
+      "grad_norm": 0.5462802648544312,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 9725
+    },
+    {
+      "epoch": 1.1299448155678187,
+      "grad_norm": 0.5840174555778503,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 9726
+    },
+    {
+      "epoch": 1.1300609933197792,
+      "grad_norm": 0.536839485168457,
+      "learning_rate": 0.0001,
+      "loss": 1.2596,
+      "step": 9727
+    },
+    {
+      "epoch": 1.1301771710717397,
+      "grad_norm": 0.5742023587226868,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 9728
+    },
+    {
+      "epoch": 1.1302933488237001,
+      "grad_norm": 0.5591979026794434,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 9729
+    },
+    {
+      "epoch": 1.1304095265756609,
+      "grad_norm": 0.5407907366752625,
+      "learning_rate": 0.0001,
+      "loss": 1.368,
+      "step": 9730
+    },
+    {
+      "epoch": 1.1305257043276213,
+      "grad_norm": 0.5862224102020264,
+      "learning_rate": 0.0001,
+      "loss": 1.6457,
+      "step": 9731
+    },
+    {
+      "epoch": 1.1306418820795818,
+      "grad_norm": 0.5518715977668762,
+      "learning_rate": 0.0001,
+      "loss": 1.421,
+      "step": 9732
+    },
+    {
+      "epoch": 1.1307580598315423,
+      "grad_norm": 0.5652369260787964,
+      "learning_rate": 0.0001,
+      "loss": 1.4673,
+      "step": 9733
+    },
+    {
+      "epoch": 1.1308742375835028,
+      "grad_norm": 0.5239549279212952,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 9734
+    },
+    {
+      "epoch": 1.1309904153354633,
+      "grad_norm": 0.5143555998802185,
+      "learning_rate": 0.0001,
+      "loss": 1.3138,
+      "step": 9735
+    },
+    {
+      "epoch": 1.1311065930874238,
+      "grad_norm": 0.5065417289733887,
+      "learning_rate": 0.0001,
+      "loss": 1.4762,
+      "step": 9736
+    },
+    {
+      "epoch": 1.1312227708393843,
+      "grad_norm": 0.5721327662467957,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 9737
+    },
+    {
+      "epoch": 1.1313389485913448,
+      "grad_norm": 0.553887665271759,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 9738
+    },
+    {
+      "epoch": 1.1314551263433053,
+      "grad_norm": 0.5835604667663574,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 9739
+    },
+    {
+      "epoch": 1.1315713040952657,
+      "grad_norm": 0.6082079410552979,
+      "learning_rate": 0.0001,
+      "loss": 1.5311,
+      "step": 9740
+    },
+    {
+      "epoch": 1.1316874818472262,
+      "grad_norm": 0.5478276014328003,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 9741
+    },
+    {
+      "epoch": 1.1318036595991867,
+      "grad_norm": 0.5668119192123413,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 9742
+    },
+    {
+      "epoch": 1.1319198373511472,
+      "grad_norm": 0.5409606695175171,
+      "learning_rate": 0.0001,
+      "loss": 1.2484,
+      "step": 9743
+    },
+    {
+      "epoch": 1.1320360151031077,
+      "grad_norm": 0.5933104157447815,
+      "learning_rate": 0.0001,
+      "loss": 1.5732,
+      "step": 9744
+    },
+    {
+      "epoch": 1.1321521928550682,
+      "grad_norm": 0.5886265635490417,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 9745
+    },
+    {
+      "epoch": 1.1322683706070287,
+      "grad_norm": 0.534992516040802,
+      "learning_rate": 0.0001,
+      "loss": 1.4977,
+      "step": 9746
+    },
+    {
+      "epoch": 1.1323845483589892,
+      "grad_norm": 0.5970652103424072,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 9747
+    },
+    {
+      "epoch": 1.1325007261109499,
+      "grad_norm": 0.5775313377380371,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 9748
+    },
+    {
+      "epoch": 1.1326169038629104,
+      "grad_norm": 0.5551691055297852,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 9749
+    },
+    {
+      "epoch": 1.1327330816148709,
+      "grad_norm": 0.57322096824646,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 9750
+    },
+    {
+      "epoch": 1.1328492593668313,
+      "grad_norm": 0.5926565527915955,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 9751
+    },
+    {
+      "epoch": 1.1329654371187918,
+      "grad_norm": 0.5369096398353577,
+      "learning_rate": 0.0001,
+      "loss": 1.3547,
+      "step": 9752
+    },
+    {
+      "epoch": 1.1330816148707523,
+      "grad_norm": 0.5695880651473999,
+      "learning_rate": 0.0001,
+      "loss": 1.5324,
+      "step": 9753
+    },
+    {
+      "epoch": 1.1331977926227128,
+      "grad_norm": 0.5742299556732178,
+      "learning_rate": 0.0001,
+      "loss": 1.5644,
+      "step": 9754
+    },
+    {
+      "epoch": 1.1333139703746733,
+      "grad_norm": 0.5906009674072266,
+      "learning_rate": 0.0001,
+      "loss": 1.5804,
+      "step": 9755
+    },
+    {
+      "epoch": 1.1334301481266338,
+      "grad_norm": 0.5348914265632629,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 9756
+    },
+    {
+      "epoch": 1.1335463258785943,
+      "grad_norm": 0.5486701130867004,
+      "learning_rate": 0.0001,
+      "loss": 1.5705,
+      "step": 9757
+    },
+    {
+      "epoch": 1.1336625036305548,
+      "grad_norm": 0.5349100828170776,
+      "learning_rate": 0.0001,
+      "loss": 1.3754,
+      "step": 9758
+    },
+    {
+      "epoch": 1.1337786813825153,
+      "grad_norm": 0.6003459692001343,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 9759
+    },
+    {
+      "epoch": 1.1338948591344757,
+      "grad_norm": 0.545295774936676,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 9760
+    },
+    {
+      "epoch": 1.1340110368864362,
+      "grad_norm": 0.5949148535728455,
+      "learning_rate": 0.0001,
+      "loss": 1.4104,
+      "step": 9761
+    },
+    {
+      "epoch": 1.1341272146383967,
+      "grad_norm": 0.6306147575378418,
+      "learning_rate": 0.0001,
+      "loss": 1.6523,
+      "step": 9762
+    },
+    {
+      "epoch": 1.1342433923903572,
+      "grad_norm": 0.5710545182228088,
+      "learning_rate": 0.0001,
+      "loss": 1.3827,
+      "step": 9763
+    },
+    {
+      "epoch": 1.1343595701423177,
+      "grad_norm": 0.5881208777427673,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 9764
+    },
+    {
+      "epoch": 1.1344757478942782,
+      "grad_norm": 0.5546616315841675,
+      "learning_rate": 0.0001,
+      "loss": 1.3955,
+      "step": 9765
+    },
+    {
+      "epoch": 1.1345919256462387,
+      "grad_norm": 0.5268201231956482,
+      "learning_rate": 0.0001,
+      "loss": 1.373,
+      "step": 9766
+    },
+    {
+      "epoch": 1.1347081033981992,
+      "grad_norm": 0.5580306053161621,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 9767
+    },
+    {
+      "epoch": 1.1348242811501597,
+      "grad_norm": 0.6127236485481262,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 9768
+    },
+    {
+      "epoch": 1.1349404589021201,
+      "grad_norm": 0.5953782200813293,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 9769
+    },
+    {
+      "epoch": 1.1350566366540806,
+      "grad_norm": 0.5282092690467834,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 9770
+    },
+    {
+      "epoch": 1.1351728144060411,
+      "grad_norm": 0.5678205490112305,
+      "learning_rate": 0.0001,
+      "loss": 1.4748,
+      "step": 9771
+    },
+    {
+      "epoch": 1.1352889921580018,
+      "grad_norm": 0.522504448890686,
+      "learning_rate": 0.0001,
+      "loss": 1.4129,
+      "step": 9772
+    },
+    {
+      "epoch": 1.1354051699099623,
+      "grad_norm": 0.5682795643806458,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 9773
+    },
+    {
+      "epoch": 1.1355213476619228,
+      "grad_norm": 0.5516985654830933,
+      "learning_rate": 0.0001,
+      "loss": 1.3844,
+      "step": 9774
+    },
+    {
+      "epoch": 1.1356375254138833,
+      "grad_norm": 0.5454441905021667,
+      "learning_rate": 0.0001,
+      "loss": 1.6141,
+      "step": 9775
+    },
+    {
+      "epoch": 1.1357537031658438,
+      "grad_norm": 0.5516277551651001,
+      "learning_rate": 0.0001,
+      "loss": 1.4229,
+      "step": 9776
+    },
+    {
+      "epoch": 1.1358698809178043,
+      "grad_norm": 0.5397228002548218,
+      "learning_rate": 0.0001,
+      "loss": 1.6032,
+      "step": 9777
+    },
+    {
+      "epoch": 1.1359860586697648,
+      "grad_norm": 0.56315678358078,
+      "learning_rate": 0.0001,
+      "loss": 1.6077,
+      "step": 9778
+    },
+    {
+      "epoch": 1.1361022364217253,
+      "grad_norm": 0.5425570607185364,
+      "learning_rate": 0.0001,
+      "loss": 1.425,
+      "step": 9779
+    },
+    {
+      "epoch": 1.1362184141736857,
+      "grad_norm": 0.5313009023666382,
+      "learning_rate": 0.0001,
+      "loss": 1.4678,
+      "step": 9780
+    },
+    {
+      "epoch": 1.1363345919256462,
+      "grad_norm": 0.5546362996101379,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 9781
+    },
+    {
+      "epoch": 1.1364507696776067,
+      "grad_norm": 0.5482527017593384,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 9782
+    },
+    {
+      "epoch": 1.1365669474295672,
+      "grad_norm": 0.5520925521850586,
+      "learning_rate": 0.0001,
+      "loss": 1.358,
+      "step": 9783
+    },
+    {
+      "epoch": 1.1366831251815277,
+      "grad_norm": 0.5753809809684753,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 9784
+    },
+    {
+      "epoch": 1.1367993029334882,
+      "grad_norm": 0.5345529317855835,
+      "learning_rate": 0.0001,
+      "loss": 1.3329,
+      "step": 9785
+    },
+    {
+      "epoch": 1.1369154806854487,
+      "grad_norm": 0.5629556775093079,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 9786
+    },
+    {
+      "epoch": 1.1370316584374092,
+      "grad_norm": 0.524359405040741,
+      "learning_rate": 0.0001,
+      "loss": 1.1784,
+      "step": 9787
+    },
+    {
+      "epoch": 1.1371478361893697,
+      "grad_norm": 0.5655604004859924,
+      "learning_rate": 0.0001,
+      "loss": 1.4612,
+      "step": 9788
+    },
+    {
+      "epoch": 1.1372640139413301,
+      "grad_norm": 0.5480789542198181,
+      "learning_rate": 0.0001,
+      "loss": 1.3745,
+      "step": 9789
+    },
+    {
+      "epoch": 1.1373801916932909,
+      "grad_norm": 0.568131685256958,
+      "learning_rate": 0.0001,
+      "loss": 1.4528,
+      "step": 9790
+    },
+    {
+      "epoch": 1.1374963694452513,
+      "grad_norm": 0.5922342538833618,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 9791
+    },
+    {
+      "epoch": 1.1376125471972118,
+      "grad_norm": 0.601453423500061,
+      "learning_rate": 0.0001,
+      "loss": 1.6236,
+      "step": 9792
+    },
+    {
+      "epoch": 1.1377287249491723,
+      "grad_norm": 0.5655685663223267,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 9793
+    },
+    {
+      "epoch": 1.1378449027011328,
+      "grad_norm": 0.5532330274581909,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 9794
+    },
+    {
+      "epoch": 1.1379610804530933,
+      "grad_norm": 0.5482434034347534,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 9795
+    },
+    {
+      "epoch": 1.1380772582050538,
+      "grad_norm": 0.5250821113586426,
+      "learning_rate": 0.0001,
+      "loss": 1.3726,
+      "step": 9796
+    },
+    {
+      "epoch": 1.1381934359570143,
+      "grad_norm": 0.5487037897109985,
+      "learning_rate": 0.0001,
+      "loss": 1.5452,
+      "step": 9797
+    },
+    {
+      "epoch": 1.1383096137089748,
+      "grad_norm": 0.5330802798271179,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 9798
+    },
+    {
+      "epoch": 1.1384257914609353,
+      "grad_norm": 0.5245431065559387,
+      "learning_rate": 0.0001,
+      "loss": 1.5838,
+      "step": 9799
+    },
+    {
+      "epoch": 1.1385419692128957,
+      "grad_norm": 0.5072425007820129,
+      "learning_rate": 0.0001,
+      "loss": 1.3497,
+      "step": 9800
+    },
+    {
+      "epoch": 1.1386581469648562,
+      "grad_norm": 0.5386658310890198,
+      "learning_rate": 0.0001,
+      "loss": 1.4285,
+      "step": 9801
+    },
+    {
+      "epoch": 1.1387743247168167,
+      "grad_norm": 0.5451638698577881,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 9802
+    },
+    {
+      "epoch": 1.1388905024687772,
+      "grad_norm": 0.5327491760253906,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 9803
+    },
+    {
+      "epoch": 1.1390066802207377,
+      "grad_norm": 0.5449811220169067,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 9804
+    },
+    {
+      "epoch": 1.1391228579726982,
+      "grad_norm": 0.5945097208023071,
+      "learning_rate": 0.0001,
+      "loss": 1.5619,
+      "step": 9805
+    },
+    {
+      "epoch": 1.1392390357246587,
+      "grad_norm": 0.5785495042800903,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 9806
+    },
+    {
+      "epoch": 1.1393552134766192,
+      "grad_norm": 0.5066744089126587,
+      "learning_rate": 0.0001,
+      "loss": 1.3782,
+      "step": 9807
+    },
+    {
+      "epoch": 1.1394713912285797,
+      "grad_norm": 0.6268693208694458,
+      "learning_rate": 0.0001,
+      "loss": 1.601,
+      "step": 9808
+    },
+    {
+      "epoch": 1.1395875689805401,
+      "grad_norm": 0.5436332821846008,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 9809
+    },
+    {
+      "epoch": 1.1397037467325006,
+      "grad_norm": 0.5843521356582642,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 9810
+    },
+    {
+      "epoch": 1.1398199244844611,
+      "grad_norm": 0.5359347462654114,
+      "learning_rate": 0.0001,
+      "loss": 1.4952,
+      "step": 9811
+    },
+    {
+      "epoch": 1.1399361022364216,
+      "grad_norm": 0.5501775145530701,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 9812
+    },
+    {
+      "epoch": 1.140052279988382,
+      "grad_norm": 0.5738726258277893,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 9813
+    },
+    {
+      "epoch": 1.1401684577403428,
+      "grad_norm": 0.6035877466201782,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 9814
+    },
+    {
+      "epoch": 1.1402846354923033,
+      "grad_norm": 0.5534920692443848,
+      "learning_rate": 0.0001,
+      "loss": 1.4699,
+      "step": 9815
+    },
+    {
+      "epoch": 1.1404008132442638,
+      "grad_norm": 0.5403549075126648,
+      "learning_rate": 0.0001,
+      "loss": 1.5539,
+      "step": 9816
+    },
+    {
+      "epoch": 1.1405169909962243,
+      "grad_norm": 0.5253798961639404,
+      "learning_rate": 0.0001,
+      "loss": 1.3346,
+      "step": 9817
+    },
+    {
+      "epoch": 1.1406331687481848,
+      "grad_norm": 0.5829864144325256,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 9818
+    },
+    {
+      "epoch": 1.1407493465001453,
+      "grad_norm": 0.5337379574775696,
+      "learning_rate": 0.0001,
+      "loss": 1.447,
+      "step": 9819
+    },
+    {
+      "epoch": 1.1408655242521057,
+      "grad_norm": 0.5743491053581238,
+      "learning_rate": 0.0001,
+      "loss": 1.4244,
+      "step": 9820
+    },
+    {
+      "epoch": 1.1409817020040662,
+      "grad_norm": 0.524262011051178,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 9821
+    },
+    {
+      "epoch": 1.1410978797560267,
+      "grad_norm": 0.5501481890678406,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 9822
+    },
+    {
+      "epoch": 1.1412140575079872,
+      "grad_norm": 0.5878981351852417,
+      "learning_rate": 0.0001,
+      "loss": 1.7145,
+      "step": 9823
+    },
+    {
+      "epoch": 1.1413302352599477,
+      "grad_norm": 0.5892753601074219,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 9824
+    },
+    {
+      "epoch": 1.1414464130119082,
+      "grad_norm": 0.5729379653930664,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 9825
+    },
+    {
+      "epoch": 1.1415625907638687,
+      "grad_norm": 0.5200401544570923,
+      "learning_rate": 0.0001,
+      "loss": 1.3968,
+      "step": 9826
+    },
+    {
+      "epoch": 1.1416787685158292,
+      "grad_norm": 0.5553308725357056,
+      "learning_rate": 0.0001,
+      "loss": 1.3769,
+      "step": 9827
+    },
+    {
+      "epoch": 1.1417949462677897,
+      "grad_norm": 0.5183868408203125,
+      "learning_rate": 0.0001,
+      "loss": 1.2838,
+      "step": 9828
+    },
+    {
+      "epoch": 1.1419111240197501,
+      "grad_norm": 0.5820907950401306,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 9829
+    },
+    {
+      "epoch": 1.1420273017717106,
+      "grad_norm": 0.589948296546936,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 9830
+    },
+    {
+      "epoch": 1.1421434795236711,
+      "grad_norm": 0.5357127785682678,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 9831
+    },
+    {
+      "epoch": 1.1422596572756318,
+      "grad_norm": 0.5246970057487488,
+      "learning_rate": 0.0001,
+      "loss": 1.3825,
+      "step": 9832
+    },
+    {
+      "epoch": 1.1423758350275923,
+      "grad_norm": 0.5680531859397888,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 9833
+    },
+    {
+      "epoch": 1.1424920127795528,
+      "grad_norm": 0.5650409460067749,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 9834
+    },
+    {
+      "epoch": 1.1426081905315133,
+      "grad_norm": 0.5513635873794556,
+      "learning_rate": 0.0001,
+      "loss": 1.471,
+      "step": 9835
+    },
+    {
+      "epoch": 1.1427243682834738,
+      "grad_norm": 0.5147875547409058,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 9836
+    },
+    {
+      "epoch": 1.1428405460354343,
+      "grad_norm": 0.5659693479537964,
+      "learning_rate": 0.0001,
+      "loss": 1.4035,
+      "step": 9837
+    },
+    {
+      "epoch": 1.1429567237873948,
+      "grad_norm": 0.5677375793457031,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 9838
+    },
+    {
+      "epoch": 1.1430729015393553,
+      "grad_norm": 0.5909645557403564,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 9839
+    },
+    {
+      "epoch": 1.1431890792913157,
+      "grad_norm": 0.603756308555603,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 9840
+    },
+    {
+      "epoch": 1.1433052570432762,
+      "grad_norm": 0.6034809350967407,
+      "learning_rate": 0.0001,
+      "loss": 1.5083,
+      "step": 9841
+    },
+    {
+      "epoch": 1.1434214347952367,
+      "grad_norm": 0.5629401803016663,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 9842
+    },
+    {
+      "epoch": 1.1435376125471972,
+      "grad_norm": 0.5638735890388489,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 9843
+    },
+    {
+      "epoch": 1.1436537902991577,
+      "grad_norm": 0.572892427444458,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 9844
+    },
+    {
+      "epoch": 1.1437699680511182,
+      "grad_norm": 0.5544632077217102,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 9845
+    },
+    {
+      "epoch": 1.1438861458030787,
+      "grad_norm": 0.596784770488739,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 9846
+    },
+    {
+      "epoch": 1.1440023235550392,
+      "grad_norm": 0.5610263347625732,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 9847
+    },
+    {
+      "epoch": 1.1441185013069997,
+      "grad_norm": 0.5021182298660278,
+      "learning_rate": 0.0001,
+      "loss": 1.4044,
+      "step": 9848
+    },
+    {
+      "epoch": 1.1442346790589601,
+      "grad_norm": 0.5500348210334778,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 9849
+    },
+    {
+      "epoch": 1.1443508568109206,
+      "grad_norm": 0.5434955358505249,
+      "learning_rate": 0.0001,
+      "loss": 1.447,
+      "step": 9850
+    },
+    {
+      "epoch": 1.1444670345628811,
+      "grad_norm": 0.5625091195106506,
+      "learning_rate": 0.0001,
+      "loss": 1.4034,
+      "step": 9851
+    },
+    {
+      "epoch": 1.1445832123148416,
+      "grad_norm": 0.5443983674049377,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 9852
+    },
+    {
+      "epoch": 1.144699390066802,
+      "grad_norm": 0.5581114888191223,
+      "learning_rate": 0.0001,
+      "loss": 1.5058,
+      "step": 9853
+    },
+    {
+      "epoch": 1.1448155678187626,
+      "grad_norm": 0.5418434143066406,
+      "learning_rate": 0.0001,
+      "loss": 1.4448,
+      "step": 9854
+    },
+    {
+      "epoch": 1.1449317455707233,
+      "grad_norm": 0.5469671487808228,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 9855
+    },
+    {
+      "epoch": 1.1450479233226838,
+      "grad_norm": 0.5789541602134705,
+      "learning_rate": 0.0001,
+      "loss": 1.4371,
+      "step": 9856
+    },
+    {
+      "epoch": 1.1451641010746443,
+      "grad_norm": 0.5320298075675964,
+      "learning_rate": 0.0001,
+      "loss": 1.3405,
+      "step": 9857
+    },
+    {
+      "epoch": 1.1452802788266048,
+      "grad_norm": 0.563693642616272,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 9858
+    },
+    {
+      "epoch": 1.1453964565785653,
+      "grad_norm": 0.5295918583869934,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 9859
+    },
+    {
+      "epoch": 1.1455126343305257,
+      "grad_norm": 0.558912456035614,
+      "learning_rate": 0.0001,
+      "loss": 1.515,
+      "step": 9860
+    },
+    {
+      "epoch": 1.1456288120824862,
+      "grad_norm": 0.5343863368034363,
+      "learning_rate": 0.0001,
+      "loss": 1.5597,
+      "step": 9861
+    },
+    {
+      "epoch": 1.1457449898344467,
+      "grad_norm": 0.543212354183197,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 9862
+    },
+    {
+      "epoch": 1.1458611675864072,
+      "grad_norm": 0.5438005924224854,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 9863
+    },
+    {
+      "epoch": 1.1459773453383677,
+      "grad_norm": 0.525090217590332,
+      "learning_rate": 0.0001,
+      "loss": 1.258,
+      "step": 9864
+    },
+    {
+      "epoch": 1.1460935230903282,
+      "grad_norm": 0.5800353288650513,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 9865
+    },
+    {
+      "epoch": 1.1462097008422887,
+      "grad_norm": 0.565473198890686,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 9866
+    },
+    {
+      "epoch": 1.1463258785942492,
+      "grad_norm": 0.5632652640342712,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 9867
+    },
+    {
+      "epoch": 1.1464420563462097,
+      "grad_norm": 0.5769978165626526,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 9868
+    },
+    {
+      "epoch": 1.1465582340981701,
+      "grad_norm": 0.6273400187492371,
+      "learning_rate": 0.0001,
+      "loss": 1.6432,
+      "step": 9869
+    },
+    {
+      "epoch": 1.1466744118501306,
+      "grad_norm": 0.6044973134994507,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 9870
+    },
+    {
+      "epoch": 1.1467905896020911,
+      "grad_norm": 0.5586897730827332,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 9871
+    },
+    {
+      "epoch": 1.1469067673540516,
+      "grad_norm": 0.5836507081985474,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 9872
+    },
+    {
+      "epoch": 1.147022945106012,
+      "grad_norm": 0.5953487753868103,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 9873
+    },
+    {
+      "epoch": 1.1471391228579728,
+      "grad_norm": 0.5075315833091736,
+      "learning_rate": 0.0001,
+      "loss": 1.336,
+      "step": 9874
+    },
+    {
+      "epoch": 1.1472553006099333,
+      "grad_norm": 0.5457214713096619,
+      "learning_rate": 0.0001,
+      "loss": 1.3285,
+      "step": 9875
+    },
+    {
+      "epoch": 1.1473714783618938,
+      "grad_norm": 0.5651188492774963,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 9876
+    },
+    {
+      "epoch": 1.1474876561138543,
+      "grad_norm": 0.6411771774291992,
+      "learning_rate": 0.0001,
+      "loss": 1.6225,
+      "step": 9877
+    },
+    {
+      "epoch": 1.1476038338658148,
+      "grad_norm": 0.516982913017273,
+      "learning_rate": 0.0001,
+      "loss": 1.4613,
+      "step": 9878
+    },
+    {
+      "epoch": 1.1477200116177753,
+      "grad_norm": 0.538935661315918,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 9879
+    },
+    {
+      "epoch": 1.1478361893697357,
+      "grad_norm": 0.6248571872711182,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 9880
+    },
+    {
+      "epoch": 1.1479523671216962,
+      "grad_norm": 0.5510977506637573,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 9881
+    },
+    {
+      "epoch": 1.1480685448736567,
+      "grad_norm": 0.5983824133872986,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 9882
+    },
+    {
+      "epoch": 1.1481847226256172,
+      "grad_norm": 0.5636187195777893,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 9883
+    },
+    {
+      "epoch": 1.1483009003775777,
+      "grad_norm": 0.5564061999320984,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 9884
+    },
+    {
+      "epoch": 1.1484170781295382,
+      "grad_norm": 0.5623925924301147,
+      "learning_rate": 0.0001,
+      "loss": 1.3934,
+      "step": 9885
+    },
+    {
+      "epoch": 1.1485332558814987,
+      "grad_norm": 0.5958746075630188,
+      "learning_rate": 0.0001,
+      "loss": 1.391,
+      "step": 9886
+    },
+    {
+      "epoch": 1.1486494336334592,
+      "grad_norm": 0.6206726431846619,
+      "learning_rate": 0.0001,
+      "loss": 1.5622,
+      "step": 9887
+    },
+    {
+      "epoch": 1.1487656113854197,
+      "grad_norm": 0.5640969276428223,
+      "learning_rate": 0.0001,
+      "loss": 1.5668,
+      "step": 9888
+    },
+    {
+      "epoch": 1.1488817891373801,
+      "grad_norm": 0.5939497947692871,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 9889
+    },
+    {
+      "epoch": 1.1489979668893406,
+      "grad_norm": 0.5671570897102356,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 9890
+    },
+    {
+      "epoch": 1.1491141446413011,
+      "grad_norm": 0.5769670605659485,
+      "learning_rate": 0.0001,
+      "loss": 1.4192,
+      "step": 9891
+    },
+    {
+      "epoch": 1.1492303223932616,
+      "grad_norm": 0.5393675565719604,
+      "learning_rate": 0.0001,
+      "loss": 1.6622,
+      "step": 9892
+    },
+    {
+      "epoch": 1.149346500145222,
+      "grad_norm": 0.5987977385520935,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 9893
+    },
+    {
+      "epoch": 1.1494626778971826,
+      "grad_norm": 0.5398848652839661,
+      "learning_rate": 0.0001,
+      "loss": 1.3972,
+      "step": 9894
+    },
+    {
+      "epoch": 1.149578855649143,
+      "grad_norm": 0.5538201332092285,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 9895
+    },
+    {
+      "epoch": 1.1496950334011036,
+      "grad_norm": 0.533630907535553,
+      "learning_rate": 0.0001,
+      "loss": 1.43,
+      "step": 9896
+    },
+    {
+      "epoch": 1.1498112111530643,
+      "grad_norm": 0.5726608037948608,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 9897
+    },
+    {
+      "epoch": 1.1499273889050248,
+      "grad_norm": 0.5075186491012573,
+      "learning_rate": 0.0001,
+      "loss": 1.362,
+      "step": 9898
+    },
+    {
+      "epoch": 1.1500435666569853,
+      "grad_norm": 0.539199948310852,
+      "learning_rate": 0.0001,
+      "loss": 1.324,
+      "step": 9899
+    },
+    {
+      "epoch": 1.1501597444089458,
+      "grad_norm": 0.5326648950576782,
+      "learning_rate": 0.0001,
+      "loss": 1.5828,
+      "step": 9900
+    },
+    {
+      "epoch": 1.1502759221609062,
+      "grad_norm": 0.5526677966117859,
+      "learning_rate": 0.0001,
+      "loss": 1.4872,
+      "step": 9901
+    },
+    {
+      "epoch": 1.1503920999128667,
+      "grad_norm": 0.5454682111740112,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 9902
+    },
+    {
+      "epoch": 1.1505082776648272,
+      "grad_norm": 0.5236324071884155,
+      "learning_rate": 0.0001,
+      "loss": 1.3648,
+      "step": 9903
+    },
+    {
+      "epoch": 1.1506244554167877,
+      "grad_norm": 0.5452289581298828,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 9904
+    },
+    {
+      "epoch": 1.1507406331687482,
+      "grad_norm": 0.5598399639129639,
+      "learning_rate": 0.0001,
+      "loss": 1.5572,
+      "step": 9905
+    },
+    {
+      "epoch": 1.1508568109207087,
+      "grad_norm": 0.5843247771263123,
+      "learning_rate": 0.0001,
+      "loss": 1.6318,
+      "step": 9906
+    },
+    {
+      "epoch": 1.1509729886726692,
+      "grad_norm": 0.5665671229362488,
+      "learning_rate": 0.0001,
+      "loss": 1.5197,
+      "step": 9907
+    },
+    {
+      "epoch": 1.1510891664246297,
+      "grad_norm": 0.515862226486206,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 9908
+    },
+    {
+      "epoch": 1.1512053441765902,
+      "grad_norm": 0.542131781578064,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 9909
+    },
+    {
+      "epoch": 1.1513215219285506,
+      "grad_norm": 0.5564214587211609,
+      "learning_rate": 0.0001,
+      "loss": 1.3456,
+      "step": 9910
+    },
+    {
+      "epoch": 1.1514376996805111,
+      "grad_norm": 0.5730974674224854,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 9911
+    },
+    {
+      "epoch": 1.1515538774324716,
+      "grad_norm": 0.5575627088546753,
+      "learning_rate": 0.0001,
+      "loss": 1.4728,
+      "step": 9912
+    },
+    {
+      "epoch": 1.151670055184432,
+      "grad_norm": 0.5514092445373535,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 9913
+    },
+    {
+      "epoch": 1.1517862329363926,
+      "grad_norm": 0.5962027311325073,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 9914
+    },
+    {
+      "epoch": 1.151902410688353,
+      "grad_norm": 0.5560566186904907,
+      "learning_rate": 0.0001,
+      "loss": 1.4282,
+      "step": 9915
+    },
+    {
+      "epoch": 1.1520185884403138,
+      "grad_norm": 0.5740779042243958,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 9916
+    },
+    {
+      "epoch": 1.1521347661922743,
+      "grad_norm": 0.5669369101524353,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 9917
+    },
+    {
+      "epoch": 1.1522509439442348,
+      "grad_norm": 0.5397534370422363,
+      "learning_rate": 0.0001,
+      "loss": 1.3715,
+      "step": 9918
+    },
+    {
+      "epoch": 1.1523671216961953,
+      "grad_norm": 0.5681161284446716,
+      "learning_rate": 0.0001,
+      "loss": 1.3601,
+      "step": 9919
+    },
+    {
+      "epoch": 1.1524832994481558,
+      "grad_norm": 0.5549933910369873,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 9920
+    },
+    {
+      "epoch": 1.1525994772001162,
+      "grad_norm": 0.5869593024253845,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 9921
+    },
+    {
+      "epoch": 1.1527156549520767,
+      "grad_norm": 0.5578819513320923,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 9922
+    },
+    {
+      "epoch": 1.1528318327040372,
+      "grad_norm": 0.5640743970870972,
+      "learning_rate": 0.0001,
+      "loss": 1.4265,
+      "step": 9923
+    },
+    {
+      "epoch": 1.1529480104559977,
+      "grad_norm": 0.5549669861793518,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 9924
+    },
+    {
+      "epoch": 1.1530641882079582,
+      "grad_norm": 0.5610499978065491,
+      "learning_rate": 0.0001,
+      "loss": 1.4156,
+      "step": 9925
+    },
+    {
+      "epoch": 1.1531803659599187,
+      "grad_norm": 0.5248560905456543,
+      "learning_rate": 0.0001,
+      "loss": 1.4177,
+      "step": 9926
+    },
+    {
+      "epoch": 1.1532965437118792,
+      "grad_norm": 0.5581945180892944,
+      "learning_rate": 0.0001,
+      "loss": 1.4434,
+      "step": 9927
+    },
+    {
+      "epoch": 1.1534127214638397,
+      "grad_norm": 0.5940243601799011,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 9928
+    },
+    {
+      "epoch": 1.1535288992158002,
+      "grad_norm": 0.5478717088699341,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 9929
+    },
+    {
+      "epoch": 1.1536450769677606,
+      "grad_norm": 0.6184346079826355,
+      "learning_rate": 0.0001,
+      "loss": 1.4189,
+      "step": 9930
+    },
+    {
+      "epoch": 1.1537612547197211,
+      "grad_norm": 0.5686115622520447,
+      "learning_rate": 0.0001,
+      "loss": 1.4612,
+      "step": 9931
+    },
+    {
+      "epoch": 1.1538774324716816,
+      "grad_norm": 0.5752643942832947,
+      "learning_rate": 0.0001,
+      "loss": 1.7198,
+      "step": 9932
+    },
+    {
+      "epoch": 1.153993610223642,
+      "grad_norm": 0.5206838250160217,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 9933
+    },
+    {
+      "epoch": 1.1541097879756026,
+      "grad_norm": 0.5340604186058044,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 9934
+    },
+    {
+      "epoch": 1.154225965727563,
+      "grad_norm": 0.5354512929916382,
+      "learning_rate": 0.0001,
+      "loss": 1.4083,
+      "step": 9935
+    },
+    {
+      "epoch": 1.1543421434795236,
+      "grad_norm": 0.5514809489250183,
+      "learning_rate": 0.0001,
+      "loss": 1.5335,
+      "step": 9936
+    },
+    {
+      "epoch": 1.154458321231484,
+      "grad_norm": 0.547510027885437,
+      "learning_rate": 0.0001,
+      "loss": 1.4518,
+      "step": 9937
+    },
+    {
+      "epoch": 1.1545744989834446,
+      "grad_norm": 0.5750606060028076,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 9938
+    },
+    {
+      "epoch": 1.1546906767354053,
+      "grad_norm": 0.5391731858253479,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 9939
+    },
+    {
+      "epoch": 1.1548068544873658,
+      "grad_norm": 0.5582312345504761,
+      "learning_rate": 0.0001,
+      "loss": 1.4282,
+      "step": 9940
+    },
+    {
+      "epoch": 1.1549230322393262,
+      "grad_norm": 0.5333447456359863,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 9941
+    },
+    {
+      "epoch": 1.1550392099912867,
+      "grad_norm": 0.5500908493995667,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 9942
+    },
+    {
+      "epoch": 1.1551553877432472,
+      "grad_norm": 0.5635502934455872,
+      "learning_rate": 0.0001,
+      "loss": 1.5813,
+      "step": 9943
+    },
+    {
+      "epoch": 1.1552715654952077,
+      "grad_norm": 0.5841420292854309,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 9944
+    },
+    {
+      "epoch": 1.1553877432471682,
+      "grad_norm": 0.5658759474754333,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 9945
+    },
+    {
+      "epoch": 1.1555039209991287,
+      "grad_norm": 0.5532236099243164,
+      "learning_rate": 0.0001,
+      "loss": 1.4737,
+      "step": 9946
+    },
+    {
+      "epoch": 1.1556200987510892,
+      "grad_norm": 0.5682868361473083,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 9947
+    },
+    {
+      "epoch": 1.1557362765030497,
+      "grad_norm": 0.5833743810653687,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 9948
+    },
+    {
+      "epoch": 1.1558524542550102,
+      "grad_norm": 0.5212302803993225,
+      "learning_rate": 0.0001,
+      "loss": 1.2786,
+      "step": 9949
+    },
+    {
+      "epoch": 1.1559686320069706,
+      "grad_norm": 0.6027308702468872,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 9950
+    },
+    {
+      "epoch": 1.1560848097589311,
+      "grad_norm": 0.5450575351715088,
+      "learning_rate": 0.0001,
+      "loss": 1.4678,
+      "step": 9951
+    },
+    {
+      "epoch": 1.1562009875108916,
+      "grad_norm": 0.5215844511985779,
+      "learning_rate": 0.0001,
+      "loss": 1.4646,
+      "step": 9952
+    },
+    {
+      "epoch": 1.156317165262852,
+      "grad_norm": 0.5741782188415527,
+      "learning_rate": 0.0001,
+      "loss": 1.4504,
+      "step": 9953
+    },
+    {
+      "epoch": 1.1564333430148126,
+      "grad_norm": 0.5755845904350281,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 9954
+    },
+    {
+      "epoch": 1.156549520766773,
+      "grad_norm": 0.593981921672821,
+      "learning_rate": 0.0001,
+      "loss": 1.6518,
+      "step": 9955
+    },
+    {
+      "epoch": 1.1566656985187336,
+      "grad_norm": 0.5386795401573181,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 9956
+    },
+    {
+      "epoch": 1.156781876270694,
+      "grad_norm": 0.5577713847160339,
+      "learning_rate": 0.0001,
+      "loss": 1.3742,
+      "step": 9957
+    },
+    {
+      "epoch": 1.1568980540226548,
+      "grad_norm": 0.5421655178070068,
+      "learning_rate": 0.0001,
+      "loss": 1.3991,
+      "step": 9958
+    },
+    {
+      "epoch": 1.1570142317746153,
+      "grad_norm": 0.5427201986312866,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 9959
+    },
+    {
+      "epoch": 1.1571304095265758,
+      "grad_norm": 0.5640281438827515,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 9960
+    },
+    {
+      "epoch": 1.1572465872785362,
+      "grad_norm": 0.5602498054504395,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 9961
+    },
+    {
+      "epoch": 1.1573627650304967,
+      "grad_norm": 0.5524298548698425,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 9962
+    },
+    {
+      "epoch": 1.1574789427824572,
+      "grad_norm": 0.5706125497817993,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 9963
+    },
+    {
+      "epoch": 1.1575951205344177,
+      "grad_norm": 0.5506826639175415,
+      "learning_rate": 0.0001,
+      "loss": 1.3571,
+      "step": 9964
+    },
+    {
+      "epoch": 1.1577112982863782,
+      "grad_norm": 0.5594316720962524,
+      "learning_rate": 0.0001,
+      "loss": 1.482,
+      "step": 9965
+    },
+    {
+      "epoch": 1.1578274760383387,
+      "grad_norm": 0.5849363207817078,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 9966
+    },
+    {
+      "epoch": 1.1579436537902992,
+      "grad_norm": 0.5251225233078003,
+      "learning_rate": 0.0001,
+      "loss": 1.2302,
+      "step": 9967
+    },
+    {
+      "epoch": 1.1580598315422597,
+      "grad_norm": 0.6071436405181885,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 9968
+    },
+    {
+      "epoch": 1.1581760092942202,
+      "grad_norm": 0.5688941478729248,
+      "learning_rate": 0.0001,
+      "loss": 1.6126,
+      "step": 9969
+    },
+    {
+      "epoch": 1.1582921870461806,
+      "grad_norm": 0.5586535930633545,
+      "learning_rate": 0.0001,
+      "loss": 1.4697,
+      "step": 9970
+    },
+    {
+      "epoch": 1.1584083647981411,
+      "grad_norm": 0.6454538702964783,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 9971
+    },
+    {
+      "epoch": 1.1585245425501016,
+      "grad_norm": 0.5616946816444397,
+      "learning_rate": 0.0001,
+      "loss": 1.389,
+      "step": 9972
+    },
+    {
+      "epoch": 1.158640720302062,
+      "grad_norm": 0.6290552616119385,
+      "learning_rate": 0.0001,
+      "loss": 1.6434,
+      "step": 9973
+    },
+    {
+      "epoch": 1.1587568980540226,
+      "grad_norm": 0.5599773526191711,
+      "learning_rate": 0.0001,
+      "loss": 1.3618,
+      "step": 9974
+    },
+    {
+      "epoch": 1.158873075805983,
+      "grad_norm": 0.5668715834617615,
+      "learning_rate": 0.0001,
+      "loss": 1.403,
+      "step": 9975
+    },
+    {
+      "epoch": 1.1589892535579436,
+      "grad_norm": 0.576244056224823,
+      "learning_rate": 0.0001,
+      "loss": 1.5943,
+      "step": 9976
+    },
+    {
+      "epoch": 1.159105431309904,
+      "grad_norm": 0.5568404793739319,
+      "learning_rate": 0.0001,
+      "loss": 1.3909,
+      "step": 9977
+    },
+    {
+      "epoch": 1.1592216090618646,
+      "grad_norm": 0.5190873146057129,
+      "learning_rate": 0.0001,
+      "loss": 1.3981,
+      "step": 9978
+    },
+    {
+      "epoch": 1.159337786813825,
+      "grad_norm": 0.580040693283081,
+      "learning_rate": 0.0001,
+      "loss": 1.3702,
+      "step": 9979
+    },
+    {
+      "epoch": 1.1594539645657855,
+      "grad_norm": 0.5603786110877991,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 9980
+    },
+    {
+      "epoch": 1.1595701423177462,
+      "grad_norm": 0.5624469518661499,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 9981
+    },
+    {
+      "epoch": 1.1596863200697067,
+      "grad_norm": 0.5951324701309204,
+      "learning_rate": 0.0001,
+      "loss": 1.6157,
+      "step": 9982
+    },
+    {
+      "epoch": 1.1598024978216672,
+      "grad_norm": 0.547521710395813,
+      "learning_rate": 0.0001,
+      "loss": 1.474,
+      "step": 9983
+    },
+    {
+      "epoch": 1.1599186755736277,
+      "grad_norm": 0.5596818923950195,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 9984
+    },
+    {
+      "epoch": 1.1600348533255882,
+      "grad_norm": 0.5726192593574524,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 9985
+    },
+    {
+      "epoch": 1.1601510310775487,
+      "grad_norm": 0.5832332968711853,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 9986
+    },
+    {
+      "epoch": 1.1602672088295092,
+      "grad_norm": 0.5925440192222595,
+      "learning_rate": 0.0001,
+      "loss": 1.5852,
+      "step": 9987
+    },
+    {
+      "epoch": 1.1603833865814697,
+      "grad_norm": 0.6479099988937378,
+      "learning_rate": 0.0001,
+      "loss": 1.6276,
+      "step": 9988
+    },
+    {
+      "epoch": 1.1604995643334302,
+      "grad_norm": 0.5257347226142883,
+      "learning_rate": 0.0001,
+      "loss": 1.4354,
+      "step": 9989
+    },
+    {
+      "epoch": 1.1606157420853906,
+      "grad_norm": 0.5414207577705383,
+      "learning_rate": 0.0001,
+      "loss": 1.4025,
+      "step": 9990
+    },
+    {
+      "epoch": 1.1607319198373511,
+      "grad_norm": 0.5247325897216797,
+      "learning_rate": 0.0001,
+      "loss": 1.511,
+      "step": 9991
+    },
+    {
+      "epoch": 1.1608480975893116,
+      "grad_norm": 0.5498883724212646,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 9992
+    },
+    {
+      "epoch": 1.1609642753412721,
+      "grad_norm": 0.5376471877098083,
+      "learning_rate": 0.0001,
+      "loss": 1.4189,
+      "step": 9993
+    },
+    {
+      "epoch": 1.1610804530932326,
+      "grad_norm": 0.5790148377418518,
+      "learning_rate": 0.0001,
+      "loss": 1.6122,
+      "step": 9994
+    },
+    {
+      "epoch": 1.161196630845193,
+      "grad_norm": 0.5751370191574097,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 9995
+    },
+    {
+      "epoch": 1.1613128085971536,
+      "grad_norm": 0.5805469751358032,
+      "learning_rate": 0.0001,
+      "loss": 1.4589,
+      "step": 9996
+    },
+    {
+      "epoch": 1.161428986349114,
+      "grad_norm": 0.5531786680221558,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 9997
+    },
+    {
+      "epoch": 1.1615451641010746,
+      "grad_norm": 0.5185563564300537,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 9998
+    },
+    {
+      "epoch": 1.1616613418530353,
+      "grad_norm": 0.5276340246200562,
+      "learning_rate": 0.0001,
+      "loss": 1.3336,
+      "step": 9999
+    },
+    {
+      "epoch": 1.1617775196049958,
+      "grad_norm": 0.5624109506607056,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 10000
+    },
+    {
+      "epoch": 1.1618936973569562,
+      "grad_norm": 0.5590581297874451,
+      "learning_rate": 0.0001,
+      "loss": 1.4464,
+      "step": 10001
+    },
+    {
+      "epoch": 1.1620098751089167,
+      "grad_norm": 0.60872483253479,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 10002
+    },
+    {
+      "epoch": 1.1621260528608772,
+      "grad_norm": 0.5246327519416809,
+      "learning_rate": 0.0001,
+      "loss": 1.3447,
+      "step": 10003
+    },
+    {
+      "epoch": 1.1622422306128377,
+      "grad_norm": 0.5634382963180542,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 10004
+    },
+    {
+      "epoch": 1.1623584083647982,
+      "grad_norm": 0.55689537525177,
+      "learning_rate": 0.0001,
+      "loss": 1.3776,
+      "step": 10005
+    },
+    {
+      "epoch": 1.1624745861167587,
+      "grad_norm": 0.5736417174339294,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 10006
+    },
+    {
+      "epoch": 1.1625907638687192,
+      "grad_norm": 0.5619802474975586,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 10007
+    },
+    {
+      "epoch": 1.1627069416206797,
+      "grad_norm": 0.5841330885887146,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 10008
+    },
+    {
+      "epoch": 1.1628231193726402,
+      "grad_norm": 0.5336185693740845,
+      "learning_rate": 0.0001,
+      "loss": 1.2244,
+      "step": 10009
+    },
+    {
+      "epoch": 1.1629392971246006,
+      "grad_norm": 0.5390673875808716,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 10010
+    },
+    {
+      "epoch": 1.1630554748765611,
+      "grad_norm": 0.5778940320014954,
+      "learning_rate": 0.0001,
+      "loss": 1.2823,
+      "step": 10011
+    },
+    {
+      "epoch": 1.1631716526285216,
+      "grad_norm": 0.5761231184005737,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 10012
+    },
+    {
+      "epoch": 1.1632878303804821,
+      "grad_norm": 0.5847422480583191,
+      "learning_rate": 0.0001,
+      "loss": 1.5658,
+      "step": 10013
+    },
+    {
+      "epoch": 1.1634040081324426,
+      "grad_norm": 0.6390685439109802,
+      "learning_rate": 0.0001,
+      "loss": 1.5276,
+      "step": 10014
+    },
+    {
+      "epoch": 1.163520185884403,
+      "grad_norm": 0.5886343121528625,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 10015
+    },
+    {
+      "epoch": 1.1636363636363636,
+      "grad_norm": 0.581598699092865,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 10016
+    },
+    {
+      "epoch": 1.163752541388324,
+      "grad_norm": 0.6227822303771973,
+      "learning_rate": 0.0001,
+      "loss": 1.4497,
+      "step": 10017
+    },
+    {
+      "epoch": 1.1638687191402846,
+      "grad_norm": 0.5321307182312012,
+      "learning_rate": 0.0001,
+      "loss": 1.4031,
+      "step": 10018
+    },
+    {
+      "epoch": 1.163984896892245,
+      "grad_norm": 0.6062250137329102,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 10019
+    },
+    {
+      "epoch": 1.1641010746442055,
+      "grad_norm": 0.5998744964599609,
+      "learning_rate": 0.0001,
+      "loss": 1.4752,
+      "step": 10020
+    },
+    {
+      "epoch": 1.164217252396166,
+      "grad_norm": 0.6060172319412231,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 10021
+    },
+    {
+      "epoch": 1.1643334301481265,
+      "grad_norm": 0.5388416051864624,
+      "learning_rate": 0.0001,
+      "loss": 1.5048,
+      "step": 10022
+    },
+    {
+      "epoch": 1.1644496079000872,
+      "grad_norm": 0.5970954895019531,
+      "learning_rate": 0.0001,
+      "loss": 1.545,
+      "step": 10023
+    },
+    {
+      "epoch": 1.1645657856520477,
+      "grad_norm": 0.5507951974868774,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 10024
+    },
+    {
+      "epoch": 1.1646819634040082,
+      "grad_norm": 0.5912706255912781,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 10025
+    },
+    {
+      "epoch": 1.1647981411559687,
+      "grad_norm": 0.5806775093078613,
+      "learning_rate": 0.0001,
+      "loss": 1.4093,
+      "step": 10026
+    },
+    {
+      "epoch": 1.1649143189079292,
+      "grad_norm": 0.6133882403373718,
+      "learning_rate": 0.0001,
+      "loss": 1.5617,
+      "step": 10027
+    },
+    {
+      "epoch": 1.1650304966598897,
+      "grad_norm": 0.5739568471908569,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 10028
+    },
+    {
+      "epoch": 1.1651466744118502,
+      "grad_norm": 0.5220316052436829,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 10029
+    },
+    {
+      "epoch": 1.1652628521638106,
+      "grad_norm": 0.5443993806838989,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 10030
+    },
+    {
+      "epoch": 1.1653790299157711,
+      "grad_norm": 0.5931938290596008,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 10031
+    },
+    {
+      "epoch": 1.1654952076677316,
+      "grad_norm": 0.5669753551483154,
+      "learning_rate": 0.0001,
+      "loss": 1.3503,
+      "step": 10032
+    },
+    {
+      "epoch": 1.1656113854196921,
+      "grad_norm": 0.5309962034225464,
+      "learning_rate": 0.0001,
+      "loss": 1.3674,
+      "step": 10033
+    },
+    {
+      "epoch": 1.1657275631716526,
+      "grad_norm": 0.5499364137649536,
+      "learning_rate": 0.0001,
+      "loss": 1.3246,
+      "step": 10034
+    },
+    {
+      "epoch": 1.165843740923613,
+      "grad_norm": 0.5379160046577454,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 10035
+    },
+    {
+      "epoch": 1.1659599186755736,
+      "grad_norm": 0.605657696723938,
+      "learning_rate": 0.0001,
+      "loss": 1.5954,
+      "step": 10036
+    },
+    {
+      "epoch": 1.166076096427534,
+      "grad_norm": 0.5854047536849976,
+      "learning_rate": 0.0001,
+      "loss": 1.4527,
+      "step": 10037
+    },
+    {
+      "epoch": 1.1661922741794946,
+      "grad_norm": 0.6003736853599548,
+      "learning_rate": 0.0001,
+      "loss": 1.4352,
+      "step": 10038
+    },
+    {
+      "epoch": 1.166308451931455,
+      "grad_norm": 0.5782053470611572,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 10039
+    },
+    {
+      "epoch": 1.1664246296834155,
+      "grad_norm": 0.5899875164031982,
+      "learning_rate": 0.0001,
+      "loss": 1.689,
+      "step": 10040
+    },
+    {
+      "epoch": 1.1665408074353762,
+      "grad_norm": 0.580557107925415,
+      "learning_rate": 0.0001,
+      "loss": 1.5173,
+      "step": 10041
+    },
+    {
+      "epoch": 1.1666569851873367,
+      "grad_norm": 0.5655171275138855,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 10042
+    },
+    {
+      "epoch": 1.1667731629392972,
+      "grad_norm": 0.5810794830322266,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 10043
+    },
+    {
+      "epoch": 1.1668893406912577,
+      "grad_norm": 0.5902233123779297,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 10044
+    },
+    {
+      "epoch": 1.1670055184432182,
+      "grad_norm": 0.5441036820411682,
+      "learning_rate": 0.0001,
+      "loss": 1.4305,
+      "step": 10045
+    },
+    {
+      "epoch": 1.1671216961951787,
+      "grad_norm": 0.5872811675071716,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 10046
+    },
+    {
+      "epoch": 1.1672378739471392,
+      "grad_norm": 0.5635554194450378,
+      "learning_rate": 0.0001,
+      "loss": 1.3447,
+      "step": 10047
+    },
+    {
+      "epoch": 1.1673540516990997,
+      "grad_norm": 0.5794597864151001,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 10048
+    },
+    {
+      "epoch": 1.1674702294510602,
+      "grad_norm": 0.5093443393707275,
+      "learning_rate": 0.0001,
+      "loss": 1.3911,
+      "step": 10049
+    },
+    {
+      "epoch": 1.1675864072030206,
+      "grad_norm": 0.5254181623458862,
+      "learning_rate": 0.0001,
+      "loss": 1.415,
+      "step": 10050
+    },
+    {
+      "epoch": 1.1677025849549811,
+      "grad_norm": 0.5460498929023743,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 10051
+    },
+    {
+      "epoch": 1.1678187627069416,
+      "grad_norm": 0.5802798867225647,
+      "learning_rate": 0.0001,
+      "loss": 1.4191,
+      "step": 10052
+    },
+    {
+      "epoch": 1.1679349404589021,
+      "grad_norm": 0.5423253178596497,
+      "learning_rate": 0.0001,
+      "loss": 1.4365,
+      "step": 10053
+    },
+    {
+      "epoch": 1.1680511182108626,
+      "grad_norm": 0.5617722272872925,
+      "learning_rate": 0.0001,
+      "loss": 1.2869,
+      "step": 10054
+    },
+    {
+      "epoch": 1.168167295962823,
+      "grad_norm": 0.542448103427887,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 10055
+    },
+    {
+      "epoch": 1.1682834737147836,
+      "grad_norm": 0.5292364954948425,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 10056
+    },
+    {
+      "epoch": 1.168399651466744,
+      "grad_norm": 0.6243206858634949,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 10057
+    },
+    {
+      "epoch": 1.1685158292187046,
+      "grad_norm": 0.5304319858551025,
+      "learning_rate": 0.0001,
+      "loss": 1.36,
+      "step": 10058
+    },
+    {
+      "epoch": 1.168632006970665,
+      "grad_norm": 0.5566378831863403,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 10059
+    },
+    {
+      "epoch": 1.1687481847226255,
+      "grad_norm": 0.6295537352561951,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 10060
+    },
+    {
+      "epoch": 1.168864362474586,
+      "grad_norm": 0.5838227868080139,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 10061
+    },
+    {
+      "epoch": 1.1689805402265465,
+      "grad_norm": 0.5756407380104065,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 10062
+    },
+    {
+      "epoch": 1.169096717978507,
+      "grad_norm": 0.5596350431442261,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 10063
+    },
+    {
+      "epoch": 1.1692128957304675,
+      "grad_norm": 0.5502869486808777,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 10064
+    },
+    {
+      "epoch": 1.1693290734824282,
+      "grad_norm": 0.5689765214920044,
+      "learning_rate": 0.0001,
+      "loss": 1.3957,
+      "step": 10065
+    },
+    {
+      "epoch": 1.1694452512343887,
+      "grad_norm": 0.5550212264060974,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 10066
+    },
+    {
+      "epoch": 1.1695614289863492,
+      "grad_norm": 0.5285496711730957,
+      "learning_rate": 0.0001,
+      "loss": 1.3828,
+      "step": 10067
+    },
+    {
+      "epoch": 1.1696776067383097,
+      "grad_norm": 0.5546186566352844,
+      "learning_rate": 0.0001,
+      "loss": 1.3966,
+      "step": 10068
+    },
+    {
+      "epoch": 1.1697937844902702,
+      "grad_norm": 0.5387147068977356,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 10069
+    },
+    {
+      "epoch": 1.1699099622422306,
+      "grad_norm": 0.6305792927742004,
+      "learning_rate": 0.0001,
+      "loss": 1.6158,
+      "step": 10070
+    },
+    {
+      "epoch": 1.1700261399941911,
+      "grad_norm": 0.5712059736251831,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 10071
+    },
+    {
+      "epoch": 1.1701423177461516,
+      "grad_norm": 0.5652614831924438,
+      "learning_rate": 0.0001,
+      "loss": 1.5069,
+      "step": 10072
+    },
+    {
+      "epoch": 1.1702584954981121,
+      "grad_norm": 0.6206264495849609,
+      "learning_rate": 0.0001,
+      "loss": 1.667,
+      "step": 10073
+    },
+    {
+      "epoch": 1.1703746732500726,
+      "grad_norm": 0.5684471726417542,
+      "learning_rate": 0.0001,
+      "loss": 1.3606,
+      "step": 10074
+    },
+    {
+      "epoch": 1.170490851002033,
+      "grad_norm": 0.5581884980201721,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 10075
+    },
+    {
+      "epoch": 1.1706070287539936,
+      "grad_norm": 0.568840742111206,
+      "learning_rate": 0.0001,
+      "loss": 1.4518,
+      "step": 10076
+    },
+    {
+      "epoch": 1.170723206505954,
+      "grad_norm": 0.5445674061775208,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 10077
+    },
+    {
+      "epoch": 1.1708393842579146,
+      "grad_norm": 0.52695631980896,
+      "learning_rate": 0.0001,
+      "loss": 1.332,
+      "step": 10078
+    },
+    {
+      "epoch": 1.170955562009875,
+      "grad_norm": 0.5562939643859863,
+      "learning_rate": 0.0001,
+      "loss": 1.5677,
+      "step": 10079
+    },
+    {
+      "epoch": 1.1710717397618355,
+      "grad_norm": 0.5368567705154419,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 10080
+    },
+    {
+      "epoch": 1.171187917513796,
+      "grad_norm": 0.569368839263916,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 10081
+    },
+    {
+      "epoch": 1.1713040952657565,
+      "grad_norm": 0.5831530690193176,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 10082
+    },
+    {
+      "epoch": 1.1714202730177172,
+      "grad_norm": 0.5623541474342346,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 10083
+    },
+    {
+      "epoch": 1.1715364507696777,
+      "grad_norm": 0.5392382144927979,
+      "learning_rate": 0.0001,
+      "loss": 1.3762,
+      "step": 10084
+    },
+    {
+      "epoch": 1.1716526285216382,
+      "grad_norm": 0.5557898283004761,
+      "learning_rate": 0.0001,
+      "loss": 1.2674,
+      "step": 10085
+    },
+    {
+      "epoch": 1.1717688062735987,
+      "grad_norm": 0.5454236268997192,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 10086
+    },
+    {
+      "epoch": 1.1718849840255592,
+      "grad_norm": 0.5669856071472168,
+      "learning_rate": 0.0001,
+      "loss": 1.365,
+      "step": 10087
+    },
+    {
+      "epoch": 1.1720011617775197,
+      "grad_norm": 0.5392054319381714,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 10088
+    },
+    {
+      "epoch": 1.1721173395294802,
+      "grad_norm": 0.544360876083374,
+      "learning_rate": 0.0001,
+      "loss": 1.4126,
+      "step": 10089
+    },
+    {
+      "epoch": 1.1722335172814407,
+      "grad_norm": 0.5895025730133057,
+      "learning_rate": 0.0001,
+      "loss": 1.5437,
+      "step": 10090
+    },
+    {
+      "epoch": 1.1723496950334011,
+      "grad_norm": 0.504751443862915,
+      "learning_rate": 0.0001,
+      "loss": 1.308,
+      "step": 10091
+    },
+    {
+      "epoch": 1.1724658727853616,
+      "grad_norm": 0.5912155508995056,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 10092
+    },
+    {
+      "epoch": 1.1725820505373221,
+      "grad_norm": 0.5704768300056458,
+      "learning_rate": 0.0001,
+      "loss": 1.397,
+      "step": 10093
+    },
+    {
+      "epoch": 1.1726982282892826,
+      "grad_norm": 0.6013458371162415,
+      "learning_rate": 0.0001,
+      "loss": 1.4999,
+      "step": 10094
+    },
+    {
+      "epoch": 1.172814406041243,
+      "grad_norm": 0.5588251352310181,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 10095
+    },
+    {
+      "epoch": 1.1729305837932036,
+      "grad_norm": 0.5580037832260132,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 10096
+    },
+    {
+      "epoch": 1.173046761545164,
+      "grad_norm": 0.5618078708648682,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 10097
+    },
+    {
+      "epoch": 1.1731629392971246,
+      "grad_norm": 0.5638139247894287,
+      "learning_rate": 0.0001,
+      "loss": 1.392,
+      "step": 10098
+    },
+    {
+      "epoch": 1.173279117049085,
+      "grad_norm": 0.49057668447494507,
+      "learning_rate": 0.0001,
+      "loss": 1.3168,
+      "step": 10099
+    },
+    {
+      "epoch": 1.1733952948010455,
+      "grad_norm": 0.5525259971618652,
+      "learning_rate": 0.0001,
+      "loss": 1.5545,
+      "step": 10100
+    },
+    {
+      "epoch": 1.173511472553006,
+      "grad_norm": 0.5700230598449707,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 10101
+    },
+    {
+      "epoch": 1.1736276503049665,
+      "grad_norm": 0.517629861831665,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 10102
+    },
+    {
+      "epoch": 1.173743828056927,
+      "grad_norm": 0.5485448241233826,
+      "learning_rate": 0.0001,
+      "loss": 1.4147,
+      "step": 10103
+    },
+    {
+      "epoch": 1.1738600058088875,
+      "grad_norm": 0.5651414394378662,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 10104
+    },
+    {
+      "epoch": 1.173976183560848,
+      "grad_norm": 0.5873313546180725,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 10105
+    },
+    {
+      "epoch": 1.1740923613128085,
+      "grad_norm": 0.5807081460952759,
+      "learning_rate": 0.0001,
+      "loss": 1.4068,
+      "step": 10106
+    },
+    {
+      "epoch": 1.1742085390647692,
+      "grad_norm": 0.5531933903694153,
+      "learning_rate": 0.0001,
+      "loss": 1.4888,
+      "step": 10107
+    },
+    {
+      "epoch": 1.1743247168167297,
+      "grad_norm": 0.6018572449684143,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 10108
+    },
+    {
+      "epoch": 1.1744408945686902,
+      "grad_norm": 0.5930827260017395,
+      "learning_rate": 0.0001,
+      "loss": 1.5595,
+      "step": 10109
+    },
+    {
+      "epoch": 1.1745570723206507,
+      "grad_norm": 0.5503697991371155,
+      "learning_rate": 0.0001,
+      "loss": 1.3225,
+      "step": 10110
+    },
+    {
+      "epoch": 1.1746732500726111,
+      "grad_norm": 0.5934546589851379,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 10111
+    },
+    {
+      "epoch": 1.1747894278245716,
+      "grad_norm": 0.6462099552154541,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 10112
+    },
+    {
+      "epoch": 1.1749056055765321,
+      "grad_norm": 0.5270904302597046,
+      "learning_rate": 0.0001,
+      "loss": 1.3637,
+      "step": 10113
+    },
+    {
+      "epoch": 1.1750217833284926,
+      "grad_norm": 0.6062031984329224,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 10114
+    },
+    {
+      "epoch": 1.175137961080453,
+      "grad_norm": 0.6364587545394897,
+      "learning_rate": 0.0001,
+      "loss": 1.502,
+      "step": 10115
+    },
+    {
+      "epoch": 1.1752541388324136,
+      "grad_norm": 0.6208913326263428,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 10116
+    },
+    {
+      "epoch": 1.175370316584374,
+      "grad_norm": 0.5763484835624695,
+      "learning_rate": 0.0001,
+      "loss": 1.4576,
+      "step": 10117
+    },
+    {
+      "epoch": 1.1754864943363346,
+      "grad_norm": 0.6294147968292236,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 10118
+    },
+    {
+      "epoch": 1.175602672088295,
+      "grad_norm": 0.5596899390220642,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 10119
+    },
+    {
+      "epoch": 1.1757188498402555,
+      "grad_norm": 0.6140156984329224,
+      "learning_rate": 0.0001,
+      "loss": 1.6399,
+      "step": 10120
+    },
+    {
+      "epoch": 1.175835027592216,
+      "grad_norm": 0.5578920841217041,
+      "learning_rate": 0.0001,
+      "loss": 1.4892,
+      "step": 10121
+    },
+    {
+      "epoch": 1.1759512053441765,
+      "grad_norm": 0.5545052886009216,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 10122
+    },
+    {
+      "epoch": 1.176067383096137,
+      "grad_norm": 0.5487355589866638,
+      "learning_rate": 0.0001,
+      "loss": 1.3392,
+      "step": 10123
+    },
+    {
+      "epoch": 1.1761835608480975,
+      "grad_norm": 0.5947970747947693,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 10124
+    },
+    {
+      "epoch": 1.1762997386000582,
+      "grad_norm": 0.5524084568023682,
+      "learning_rate": 0.0001,
+      "loss": 1.4826,
+      "step": 10125
+    },
+    {
+      "epoch": 1.1764159163520187,
+      "grad_norm": 0.5855798125267029,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 10126
+    },
+    {
+      "epoch": 1.1765320941039792,
+      "grad_norm": 0.5738104581832886,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 10127
+    },
+    {
+      "epoch": 1.1766482718559397,
+      "grad_norm": 0.6277796030044556,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 10128
+    },
+    {
+      "epoch": 1.1767644496079002,
+      "grad_norm": 0.558943510055542,
+      "learning_rate": 0.0001,
+      "loss": 1.4912,
+      "step": 10129
+    },
+    {
+      "epoch": 1.1768806273598607,
+      "grad_norm": 0.5805367231369019,
+      "learning_rate": 0.0001,
+      "loss": 1.6493,
+      "step": 10130
+    },
+    {
+      "epoch": 1.1769968051118211,
+      "grad_norm": 0.5456610918045044,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 10131
+    },
+    {
+      "epoch": 1.1771129828637816,
+      "grad_norm": 0.5442765355110168,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 10132
+    },
+    {
+      "epoch": 1.1772291606157421,
+      "grad_norm": 0.5121324062347412,
+      "learning_rate": 0.0001,
+      "loss": 1.3845,
+      "step": 10133
+    },
+    {
+      "epoch": 1.1773453383677026,
+      "grad_norm": 0.5197672843933105,
+      "learning_rate": 0.0001,
+      "loss": 1.3748,
+      "step": 10134
+    },
+    {
+      "epoch": 1.177461516119663,
+      "grad_norm": 0.5549630522727966,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 10135
+    },
+    {
+      "epoch": 1.1775776938716236,
+      "grad_norm": 0.6009778380393982,
+      "learning_rate": 0.0001,
+      "loss": 1.6594,
+      "step": 10136
+    },
+    {
+      "epoch": 1.177693871623584,
+      "grad_norm": 0.6478327512741089,
+      "learning_rate": 0.0001,
+      "loss": 1.3931,
+      "step": 10137
+    },
+    {
+      "epoch": 1.1778100493755446,
+      "grad_norm": 0.5748621821403503,
+      "learning_rate": 0.0001,
+      "loss": 1.5719,
+      "step": 10138
+    },
+    {
+      "epoch": 1.177926227127505,
+      "grad_norm": 0.5634380578994751,
+      "learning_rate": 0.0001,
+      "loss": 1.4188,
+      "step": 10139
+    },
+    {
+      "epoch": 1.1780424048794655,
+      "grad_norm": 0.5881836414337158,
+      "learning_rate": 0.0001,
+      "loss": 1.5994,
+      "step": 10140
+    },
+    {
+      "epoch": 1.178158582631426,
+      "grad_norm": 0.5679641962051392,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 10141
+    },
+    {
+      "epoch": 1.1782747603833865,
+      "grad_norm": 0.5754004120826721,
+      "learning_rate": 0.0001,
+      "loss": 1.6292,
+      "step": 10142
+    },
+    {
+      "epoch": 1.178390938135347,
+      "grad_norm": 0.6078845858573914,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 10143
+    },
+    {
+      "epoch": 1.1785071158873075,
+      "grad_norm": 0.5382782220840454,
+      "learning_rate": 0.0001,
+      "loss": 1.3716,
+      "step": 10144
+    },
+    {
+      "epoch": 1.178623293639268,
+      "grad_norm": 0.565054178237915,
+      "learning_rate": 0.0001,
+      "loss": 1.4706,
+      "step": 10145
+    },
+    {
+      "epoch": 1.1787394713912285,
+      "grad_norm": 0.6017858982086182,
+      "learning_rate": 0.0001,
+      "loss": 1.5564,
+      "step": 10146
+    },
+    {
+      "epoch": 1.178855649143189,
+      "grad_norm": 0.533458948135376,
+      "learning_rate": 0.0001,
+      "loss": 1.2963,
+      "step": 10147
+    },
+    {
+      "epoch": 1.1789718268951495,
+      "grad_norm": 0.6092103719711304,
+      "learning_rate": 0.0001,
+      "loss": 1.5096,
+      "step": 10148
+    },
+    {
+      "epoch": 1.1790880046471102,
+      "grad_norm": 0.5930465459823608,
+      "learning_rate": 0.0001,
+      "loss": 1.476,
+      "step": 10149
+    },
+    {
+      "epoch": 1.1792041823990707,
+      "grad_norm": 0.5392512083053589,
+      "learning_rate": 0.0001,
+      "loss": 1.4036,
+      "step": 10150
+    },
+    {
+      "epoch": 1.1793203601510311,
+      "grad_norm": 0.5717263221740723,
+      "learning_rate": 0.0001,
+      "loss": 1.5386,
+      "step": 10151
+    },
+    {
+      "epoch": 1.1794365379029916,
+      "grad_norm": 0.5188798308372498,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 10152
+    },
+    {
+      "epoch": 1.1795527156549521,
+      "grad_norm": 0.5673980712890625,
+      "learning_rate": 0.0001,
+      "loss": 1.446,
+      "step": 10153
+    },
+    {
+      "epoch": 1.1796688934069126,
+      "grad_norm": 0.5817063450813293,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 10154
+    },
+    {
+      "epoch": 1.179785071158873,
+      "grad_norm": 0.5734279751777649,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 10155
+    },
+    {
+      "epoch": 1.1799012489108336,
+      "grad_norm": 0.540127158164978,
+      "learning_rate": 0.0001,
+      "loss": 1.5124,
+      "step": 10156
+    },
+    {
+      "epoch": 1.180017426662794,
+      "grad_norm": 0.6068228483200073,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 10157
+    },
+    {
+      "epoch": 1.1801336044147546,
+      "grad_norm": 0.5281007289886475,
+      "learning_rate": 0.0001,
+      "loss": 1.3727,
+      "step": 10158
+    },
+    {
+      "epoch": 1.180249782166715,
+      "grad_norm": 0.57302325963974,
+      "learning_rate": 0.0001,
+      "loss": 1.4104,
+      "step": 10159
+    },
+    {
+      "epoch": 1.1803659599186755,
+      "grad_norm": 0.5550638437271118,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 10160
+    },
+    {
+      "epoch": 1.180482137670636,
+      "grad_norm": 0.559135377407074,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 10161
+    },
+    {
+      "epoch": 1.1805983154225965,
+      "grad_norm": 0.5715136528015137,
+      "learning_rate": 0.0001,
+      "loss": 1.481,
+      "step": 10162
+    },
+    {
+      "epoch": 1.180714493174557,
+      "grad_norm": 0.5593242645263672,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 10163
+    },
+    {
+      "epoch": 1.1808306709265175,
+      "grad_norm": 0.5487082004547119,
+      "learning_rate": 0.0001,
+      "loss": 1.4297,
+      "step": 10164
+    },
+    {
+      "epoch": 1.180946848678478,
+      "grad_norm": 0.5988155603408813,
+      "learning_rate": 0.0001,
+      "loss": 1.6336,
+      "step": 10165
+    },
+    {
+      "epoch": 1.1810630264304385,
+      "grad_norm": 0.5767937302589417,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 10166
+    },
+    {
+      "epoch": 1.1811792041823992,
+      "grad_norm": 0.6116029620170593,
+      "learning_rate": 0.0001,
+      "loss": 1.5717,
+      "step": 10167
+    },
+    {
+      "epoch": 1.1812953819343597,
+      "grad_norm": 0.5582594275474548,
+      "learning_rate": 0.0001,
+      "loss": 1.361,
+      "step": 10168
+    },
+    {
+      "epoch": 1.1814115596863202,
+      "grad_norm": 0.5252476930618286,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 10169
+    },
+    {
+      "epoch": 1.1815277374382807,
+      "grad_norm": 0.5445083379745483,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 10170
+    },
+    {
+      "epoch": 1.1816439151902411,
+      "grad_norm": 0.5257118344306946,
+      "learning_rate": 0.0001,
+      "loss": 1.3504,
+      "step": 10171
+    },
+    {
+      "epoch": 1.1817600929422016,
+      "grad_norm": 0.5568935871124268,
+      "learning_rate": 0.0001,
+      "loss": 1.4852,
+      "step": 10172
+    },
+    {
+      "epoch": 1.1818762706941621,
+      "grad_norm": 0.5395621061325073,
+      "learning_rate": 0.0001,
+      "loss": 1.5846,
+      "step": 10173
+    },
+    {
+      "epoch": 1.1819924484461226,
+      "grad_norm": 0.5763841867446899,
+      "learning_rate": 0.0001,
+      "loss": 1.57,
+      "step": 10174
+    },
+    {
+      "epoch": 1.182108626198083,
+      "grad_norm": 0.5146093964576721,
+      "learning_rate": 0.0001,
+      "loss": 1.3141,
+      "step": 10175
+    },
+    {
+      "epoch": 1.1822248039500436,
+      "grad_norm": 0.5548669099807739,
+      "learning_rate": 0.0001,
+      "loss": 1.4279,
+      "step": 10176
+    },
+    {
+      "epoch": 1.182340981702004,
+      "grad_norm": 0.5692338347434998,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 10177
+    },
+    {
+      "epoch": 1.1824571594539646,
+      "grad_norm": 0.5822345614433289,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 10178
+    },
+    {
+      "epoch": 1.182573337205925,
+      "grad_norm": 0.5448129177093506,
+      "learning_rate": 0.0001,
+      "loss": 1.3834,
+      "step": 10179
+    },
+    {
+      "epoch": 1.1826895149578855,
+      "grad_norm": 0.6466269493103027,
+      "learning_rate": 0.0001,
+      "loss": 1.4509,
+      "step": 10180
+    },
+    {
+      "epoch": 1.182805692709846,
+      "grad_norm": 0.6030110120773315,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 10181
+    },
+    {
+      "epoch": 1.1829218704618065,
+      "grad_norm": 0.5867081880569458,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 10182
+    },
+    {
+      "epoch": 1.183038048213767,
+      "grad_norm": 0.543514609336853,
+      "learning_rate": 0.0001,
+      "loss": 1.4714,
+      "step": 10183
+    },
+    {
+      "epoch": 1.1831542259657275,
+      "grad_norm": 0.585824728012085,
+      "learning_rate": 0.0001,
+      "loss": 1.6173,
+      "step": 10184
+    },
+    {
+      "epoch": 1.183270403717688,
+      "grad_norm": 0.5779645442962646,
+      "learning_rate": 0.0001,
+      "loss": 1.5448,
+      "step": 10185
+    },
+    {
+      "epoch": 1.1833865814696485,
+      "grad_norm": 0.5890458226203918,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 10186
+    },
+    {
+      "epoch": 1.183502759221609,
+      "grad_norm": 0.5767694711685181,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 10187
+    },
+    {
+      "epoch": 1.1836189369735695,
+      "grad_norm": 0.5827814936637878,
+      "learning_rate": 0.0001,
+      "loss": 1.4696,
+      "step": 10188
+    },
+    {
+      "epoch": 1.18373511472553,
+      "grad_norm": 0.5493532419204712,
+      "learning_rate": 0.0001,
+      "loss": 1.4297,
+      "step": 10189
+    },
+    {
+      "epoch": 1.1838512924774907,
+      "grad_norm": 0.5599081516265869,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 10190
+    },
+    {
+      "epoch": 1.1839674702294511,
+      "grad_norm": 0.5541647672653198,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 10191
+    },
+    {
+      "epoch": 1.1840836479814116,
+      "grad_norm": 0.5816530585289001,
+      "learning_rate": 0.0001,
+      "loss": 1.611,
+      "step": 10192
+    },
+    {
+      "epoch": 1.1841998257333721,
+      "grad_norm": 0.6219608187675476,
+      "learning_rate": 0.0001,
+      "loss": 1.5939,
+      "step": 10193
+    },
+    {
+      "epoch": 1.1843160034853326,
+      "grad_norm": 0.5906521081924438,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 10194
+    },
+    {
+      "epoch": 1.184432181237293,
+      "grad_norm": 0.5859937071800232,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 10195
+    },
+    {
+      "epoch": 1.1845483589892536,
+      "grad_norm": 0.5565850138664246,
+      "learning_rate": 0.0001,
+      "loss": 1.4176,
+      "step": 10196
+    },
+    {
+      "epoch": 1.184664536741214,
+      "grad_norm": 0.5286614298820496,
+      "learning_rate": 0.0001,
+      "loss": 1.2829,
+      "step": 10197
+    },
+    {
+      "epoch": 1.1847807144931746,
+      "grad_norm": 0.5622779726982117,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 10198
+    },
+    {
+      "epoch": 1.184896892245135,
+      "grad_norm": 0.5423121452331543,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 10199
+    },
+    {
+      "epoch": 1.1850130699970955,
+      "grad_norm": 0.564444899559021,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 10200
+    },
+    {
+      "epoch": 1.185129247749056,
+      "grad_norm": 0.5337464213371277,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 10201
+    },
+    {
+      "epoch": 1.1852454255010165,
+      "grad_norm": 0.5334920287132263,
+      "learning_rate": 0.0001,
+      "loss": 1.369,
+      "step": 10202
+    },
+    {
+      "epoch": 1.185361603252977,
+      "grad_norm": 0.5582081079483032,
+      "learning_rate": 0.0001,
+      "loss": 1.5346,
+      "step": 10203
+    },
+    {
+      "epoch": 1.1854777810049375,
+      "grad_norm": 0.5935722589492798,
+      "learning_rate": 0.0001,
+      "loss": 1.5975,
+      "step": 10204
+    },
+    {
+      "epoch": 1.185593958756898,
+      "grad_norm": 0.606367290019989,
+      "learning_rate": 0.0001,
+      "loss": 1.6052,
+      "step": 10205
+    },
+    {
+      "epoch": 1.1857101365088585,
+      "grad_norm": 0.5262031555175781,
+      "learning_rate": 0.0001,
+      "loss": 1.3391,
+      "step": 10206
+    },
+    {
+      "epoch": 1.185826314260819,
+      "grad_norm": 0.5637447237968445,
+      "learning_rate": 0.0001,
+      "loss": 1.2968,
+      "step": 10207
+    },
+    {
+      "epoch": 1.1859424920127795,
+      "grad_norm": 0.5710691213607788,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 10208
+    },
+    {
+      "epoch": 1.1860586697647402,
+      "grad_norm": 0.5453031063079834,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 10209
+    },
+    {
+      "epoch": 1.1861748475167007,
+      "grad_norm": 0.530640721321106,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 10210
+    },
+    {
+      "epoch": 1.1862910252686611,
+      "grad_norm": 0.5612127184867859,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 10211
+    },
+    {
+      "epoch": 1.1864072030206216,
+      "grad_norm": 0.5393279194831848,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 10212
+    },
+    {
+      "epoch": 1.1865233807725821,
+      "grad_norm": 0.5792052149772644,
+      "learning_rate": 0.0001,
+      "loss": 1.6498,
+      "step": 10213
+    },
+    {
+      "epoch": 1.1866395585245426,
+      "grad_norm": 0.5454745292663574,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 10214
+    },
+    {
+      "epoch": 1.186755736276503,
+      "grad_norm": 0.574486255645752,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 10215
+    },
+    {
+      "epoch": 1.1868719140284636,
+      "grad_norm": 0.5667810440063477,
+      "learning_rate": 0.0001,
+      "loss": 1.4863,
+      "step": 10216
+    },
+    {
+      "epoch": 1.186988091780424,
+      "grad_norm": 0.540420651435852,
+      "learning_rate": 0.0001,
+      "loss": 1.3731,
+      "step": 10217
+    },
+    {
+      "epoch": 1.1871042695323846,
+      "grad_norm": 0.5797957181930542,
+      "learning_rate": 0.0001,
+      "loss": 1.5712,
+      "step": 10218
+    },
+    {
+      "epoch": 1.187220447284345,
+      "grad_norm": 0.5845776200294495,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 10219
+    },
+    {
+      "epoch": 1.1873366250363055,
+      "grad_norm": 0.5893095135688782,
+      "learning_rate": 0.0001,
+      "loss": 1.5909,
+      "step": 10220
+    },
+    {
+      "epoch": 1.187452802788266,
+      "grad_norm": 0.5685760378837585,
+      "learning_rate": 0.0001,
+      "loss": 1.4096,
+      "step": 10221
+    },
+    {
+      "epoch": 1.1875689805402265,
+      "grad_norm": 0.5455953478813171,
+      "learning_rate": 0.0001,
+      "loss": 1.3178,
+      "step": 10222
+    },
+    {
+      "epoch": 1.187685158292187,
+      "grad_norm": 0.5952999591827393,
+      "learning_rate": 0.0001,
+      "loss": 1.4628,
+      "step": 10223
+    },
+    {
+      "epoch": 1.1878013360441475,
+      "grad_norm": 0.5808957815170288,
+      "learning_rate": 0.0001,
+      "loss": 1.529,
+      "step": 10224
+    },
+    {
+      "epoch": 1.187917513796108,
+      "grad_norm": 0.5412543416023254,
+      "learning_rate": 0.0001,
+      "loss": 1.444,
+      "step": 10225
+    },
+    {
+      "epoch": 1.1880336915480685,
+      "grad_norm": 0.5611621141433716,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 10226
+    },
+    {
+      "epoch": 1.188149869300029,
+      "grad_norm": 0.5732444524765015,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 10227
+    },
+    {
+      "epoch": 1.1882660470519895,
+      "grad_norm": 0.6303038597106934,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 10228
+    },
+    {
+      "epoch": 1.18838222480395,
+      "grad_norm": 0.5532310009002686,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 10229
+    },
+    {
+      "epoch": 1.1884984025559104,
+      "grad_norm": 0.5555628538131714,
+      "learning_rate": 0.0001,
+      "loss": 1.5655,
+      "step": 10230
+    },
+    {
+      "epoch": 1.188614580307871,
+      "grad_norm": 0.5487105846405029,
+      "learning_rate": 0.0001,
+      "loss": 1.3955,
+      "step": 10231
+    },
+    {
+      "epoch": 1.1887307580598316,
+      "grad_norm": 0.5543938279151917,
+      "learning_rate": 0.0001,
+      "loss": 1.3153,
+      "step": 10232
+    },
+    {
+      "epoch": 1.1888469358117921,
+      "grad_norm": 0.5894563794136047,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 10233
+    },
+    {
+      "epoch": 1.1889631135637526,
+      "grad_norm": 0.5444560647010803,
+      "learning_rate": 0.0001,
+      "loss": 1.5692,
+      "step": 10234
+    },
+    {
+      "epoch": 1.189079291315713,
+      "grad_norm": 0.578478991985321,
+      "learning_rate": 0.0001,
+      "loss": 1.4049,
+      "step": 10235
+    },
+    {
+      "epoch": 1.1891954690676736,
+      "grad_norm": 0.5442149639129639,
+      "learning_rate": 0.0001,
+      "loss": 1.5456,
+      "step": 10236
+    },
+    {
+      "epoch": 1.189311646819634,
+      "grad_norm": 0.5239068269729614,
+      "learning_rate": 0.0001,
+      "loss": 1.2212,
+      "step": 10237
+    },
+    {
+      "epoch": 1.1894278245715946,
+      "grad_norm": 0.59259033203125,
+      "learning_rate": 0.0001,
+      "loss": 1.2868,
+      "step": 10238
+    },
+    {
+      "epoch": 1.189544002323555,
+      "grad_norm": 0.536356508731842,
+      "learning_rate": 0.0001,
+      "loss": 1.3534,
+      "step": 10239
+    },
+    {
+      "epoch": 1.1896601800755155,
+      "grad_norm": 0.6062889099121094,
+      "learning_rate": 0.0001,
+      "loss": 1.5112,
+      "step": 10240
+    },
+    {
+      "epoch": 1.189776357827476,
+      "grad_norm": 0.5797139406204224,
+      "learning_rate": 0.0001,
+      "loss": 1.618,
+      "step": 10241
+    },
+    {
+      "epoch": 1.1898925355794365,
+      "grad_norm": 0.6242355704307556,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 10242
+    },
+    {
+      "epoch": 1.190008713331397,
+      "grad_norm": 0.5886520147323608,
+      "learning_rate": 0.0001,
+      "loss": 1.4482,
+      "step": 10243
+    },
+    {
+      "epoch": 1.1901248910833575,
+      "grad_norm": 0.6011394262313843,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 10244
+    },
+    {
+      "epoch": 1.190241068835318,
+      "grad_norm": 0.5653864145278931,
+      "learning_rate": 0.0001,
+      "loss": 1.5659,
+      "step": 10245
+    },
+    {
+      "epoch": 1.1903572465872785,
+      "grad_norm": 0.5808043479919434,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 10246
+    },
+    {
+      "epoch": 1.190473424339239,
+      "grad_norm": 0.5999670028686523,
+      "learning_rate": 0.0001,
+      "loss": 1.5587,
+      "step": 10247
+    },
+    {
+      "epoch": 1.1905896020911995,
+      "grad_norm": 0.6177708506584167,
+      "learning_rate": 0.0001,
+      "loss": 1.7287,
+      "step": 10248
+    },
+    {
+      "epoch": 1.19070577984316,
+      "grad_norm": 0.6043134331703186,
+      "learning_rate": 0.0001,
+      "loss": 1.6253,
+      "step": 10249
+    },
+    {
+      "epoch": 1.1908219575951204,
+      "grad_norm": 0.5593743324279785,
+      "learning_rate": 0.0001,
+      "loss": 1.4044,
+      "step": 10250
+    },
+    {
+      "epoch": 1.1909381353470812,
+      "grad_norm": 0.5556023716926575,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 10251
+    },
+    {
+      "epoch": 1.1910543130990416,
+      "grad_norm": 0.5386412143707275,
+      "learning_rate": 0.0001,
+      "loss": 1.521,
+      "step": 10252
+    },
+    {
+      "epoch": 1.1911704908510021,
+      "grad_norm": 0.5457038879394531,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 10253
+    },
+    {
+      "epoch": 1.1912866686029626,
+      "grad_norm": 0.5371884107589722,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 10254
+    },
+    {
+      "epoch": 1.191402846354923,
+      "grad_norm": 0.5815069675445557,
+      "learning_rate": 0.0001,
+      "loss": 1.6517,
+      "step": 10255
+    },
+    {
+      "epoch": 1.1915190241068836,
+      "grad_norm": 0.5567863583564758,
+      "learning_rate": 0.0001,
+      "loss": 1.6194,
+      "step": 10256
+    },
+    {
+      "epoch": 1.191635201858844,
+      "grad_norm": 0.5386295318603516,
+      "learning_rate": 0.0001,
+      "loss": 1.5486,
+      "step": 10257
+    },
+    {
+      "epoch": 1.1917513796108046,
+      "grad_norm": 0.5735299587249756,
+      "learning_rate": 0.0001,
+      "loss": 1.4414,
+      "step": 10258
+    },
+    {
+      "epoch": 1.191867557362765,
+      "grad_norm": 0.5478100776672363,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 10259
+    },
+    {
+      "epoch": 1.1919837351147256,
+      "grad_norm": 0.5614447593688965,
+      "learning_rate": 0.0001,
+      "loss": 1.4083,
+      "step": 10260
+    },
+    {
+      "epoch": 1.192099912866686,
+      "grad_norm": 0.5918525457382202,
+      "learning_rate": 0.0001,
+      "loss": 1.3954,
+      "step": 10261
+    },
+    {
+      "epoch": 1.1922160906186465,
+      "grad_norm": 0.5843601822853088,
+      "learning_rate": 0.0001,
+      "loss": 1.4277,
+      "step": 10262
+    },
+    {
+      "epoch": 1.192332268370607,
+      "grad_norm": 0.5581550598144531,
+      "learning_rate": 0.0001,
+      "loss": 1.4389,
+      "step": 10263
+    },
+    {
+      "epoch": 1.1924484461225675,
+      "grad_norm": 0.6755667328834534,
+      "learning_rate": 0.0001,
+      "loss": 1.6176,
+      "step": 10264
+    },
+    {
+      "epoch": 1.192564623874528,
+      "grad_norm": 0.6215943694114685,
+      "learning_rate": 0.0001,
+      "loss": 1.6203,
+      "step": 10265
+    },
+    {
+      "epoch": 1.1926808016264885,
+      "grad_norm": 0.6235942840576172,
+      "learning_rate": 0.0001,
+      "loss": 1.5156,
+      "step": 10266
+    },
+    {
+      "epoch": 1.192796979378449,
+      "grad_norm": 0.5801181197166443,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 10267
+    },
+    {
+      "epoch": 1.1929131571304095,
+      "grad_norm": 0.5831395983695984,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 10268
+    },
+    {
+      "epoch": 1.19302933488237,
+      "grad_norm": 0.5884434580802917,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 10269
+    },
+    {
+      "epoch": 1.1931455126343304,
+      "grad_norm": 0.6119523048400879,
+      "learning_rate": 0.0001,
+      "loss": 1.6301,
+      "step": 10270
+    },
+    {
+      "epoch": 1.193261690386291,
+      "grad_norm": 0.5671009421348572,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 10271
+    },
+    {
+      "epoch": 1.1933778681382514,
+      "grad_norm": 0.5740747451782227,
+      "learning_rate": 0.0001,
+      "loss": 1.2502,
+      "step": 10272
+    },
+    {
+      "epoch": 1.193494045890212,
+      "grad_norm": 0.5850660800933838,
+      "learning_rate": 0.0001,
+      "loss": 1.3729,
+      "step": 10273
+    },
+    {
+      "epoch": 1.1936102236421726,
+      "grad_norm": 0.5821871757507324,
+      "learning_rate": 0.0001,
+      "loss": 1.3578,
+      "step": 10274
+    },
+    {
+      "epoch": 1.193726401394133,
+      "grad_norm": 0.5659012198448181,
+      "learning_rate": 0.0001,
+      "loss": 1.418,
+      "step": 10275
+    },
+    {
+      "epoch": 1.1938425791460936,
+      "grad_norm": 0.6060782074928284,
+      "learning_rate": 0.0001,
+      "loss": 1.6198,
+      "step": 10276
+    },
+    {
+      "epoch": 1.193958756898054,
+      "grad_norm": 0.6207935214042664,
+      "learning_rate": 0.0001,
+      "loss": 1.4444,
+      "step": 10277
+    },
+    {
+      "epoch": 1.1940749346500146,
+      "grad_norm": 0.549321711063385,
+      "learning_rate": 0.0001,
+      "loss": 1.3837,
+      "step": 10278
+    },
+    {
+      "epoch": 1.194191112401975,
+      "grad_norm": 0.5631564259529114,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 10279
+    },
+    {
+      "epoch": 1.1943072901539356,
+      "grad_norm": 0.5477903485298157,
+      "learning_rate": 0.0001,
+      "loss": 1.4314,
+      "step": 10280
+    },
+    {
+      "epoch": 1.194423467905896,
+      "grad_norm": 0.6051657199859619,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 10281
+    },
+    {
+      "epoch": 1.1945396456578565,
+      "grad_norm": 0.6054665446281433,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 10282
+    },
+    {
+      "epoch": 1.194655823409817,
+      "grad_norm": 0.57652747631073,
+      "learning_rate": 0.0001,
+      "loss": 1.5908,
+      "step": 10283
+    },
+    {
+      "epoch": 1.1947720011617775,
+      "grad_norm": 0.5573504567146301,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 10284
+    },
+    {
+      "epoch": 1.194888178913738,
+      "grad_norm": 0.5924124121665955,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 10285
+    },
+    {
+      "epoch": 1.1950043566656985,
+      "grad_norm": 0.5285975337028503,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 10286
+    },
+    {
+      "epoch": 1.195120534417659,
+      "grad_norm": 0.5311710834503174,
+      "learning_rate": 0.0001,
+      "loss": 1.4012,
+      "step": 10287
+    },
+    {
+      "epoch": 1.1952367121696195,
+      "grad_norm": 0.5347359776496887,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 10288
+    },
+    {
+      "epoch": 1.19535288992158,
+      "grad_norm": 0.563667356967926,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 10289
+    },
+    {
+      "epoch": 1.1954690676735404,
+      "grad_norm": 0.568260669708252,
+      "learning_rate": 0.0001,
+      "loss": 1.3454,
+      "step": 10290
+    },
+    {
+      "epoch": 1.195585245425501,
+      "grad_norm": 0.638574481010437,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 10291
+    },
+    {
+      "epoch": 1.1957014231774614,
+      "grad_norm": 0.6492049098014832,
+      "learning_rate": 0.0001,
+      "loss": 1.6565,
+      "step": 10292
+    },
+    {
+      "epoch": 1.1958176009294221,
+      "grad_norm": 0.5210595726966858,
+      "learning_rate": 0.0001,
+      "loss": 1.4179,
+      "step": 10293
+    },
+    {
+      "epoch": 1.1959337786813826,
+      "grad_norm": 0.5696560144424438,
+      "learning_rate": 0.0001,
+      "loss": 1.5132,
+      "step": 10294
+    },
+    {
+      "epoch": 1.196049956433343,
+      "grad_norm": 0.5875056982040405,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 10295
+    },
+    {
+      "epoch": 1.1961661341853036,
+      "grad_norm": 0.5168067216873169,
+      "learning_rate": 0.0001,
+      "loss": 1.436,
+      "step": 10296
+    },
+    {
+      "epoch": 1.196282311937264,
+      "grad_norm": 0.5388482213020325,
+      "learning_rate": 0.0001,
+      "loss": 1.3848,
+      "step": 10297
+    },
+    {
+      "epoch": 1.1963984896892246,
+      "grad_norm": 0.595585823059082,
+      "learning_rate": 0.0001,
+      "loss": 1.3237,
+      "step": 10298
+    },
+    {
+      "epoch": 1.196514667441185,
+      "grad_norm": 0.5622440576553345,
+      "learning_rate": 0.0001,
+      "loss": 1.5511,
+      "step": 10299
+    },
+    {
+      "epoch": 1.1966308451931456,
+      "grad_norm": 0.6017595529556274,
+      "learning_rate": 0.0001,
+      "loss": 1.6546,
+      "step": 10300
+    },
+    {
+      "epoch": 1.196747022945106,
+      "grad_norm": 0.5587907433509827,
+      "learning_rate": 0.0001,
+      "loss": 1.419,
+      "step": 10301
+    },
+    {
+      "epoch": 1.1968632006970665,
+      "grad_norm": 0.5632277727127075,
+      "learning_rate": 0.0001,
+      "loss": 1.3717,
+      "step": 10302
+    },
+    {
+      "epoch": 1.196979378449027,
+      "grad_norm": 0.5504304766654968,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 10303
+    },
+    {
+      "epoch": 1.1970955562009875,
+      "grad_norm": 0.5563431978225708,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 10304
+    },
+    {
+      "epoch": 1.197211733952948,
+      "grad_norm": 0.5547022223472595,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 10305
+    },
+    {
+      "epoch": 1.1973279117049085,
+      "grad_norm": 0.5395157337188721,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 10306
+    },
+    {
+      "epoch": 1.197444089456869,
+      "grad_norm": 0.6372838020324707,
+      "learning_rate": 0.0001,
+      "loss": 1.757,
+      "step": 10307
+    },
+    {
+      "epoch": 1.1975602672088295,
+      "grad_norm": 0.5683894157409668,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 10308
+    },
+    {
+      "epoch": 1.19767644496079,
+      "grad_norm": 0.5659453868865967,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 10309
+    },
+    {
+      "epoch": 1.1977926227127504,
+      "grad_norm": 0.5578935146331787,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 10310
+    },
+    {
+      "epoch": 1.197908800464711,
+      "grad_norm": 0.5382188558578491,
+      "learning_rate": 0.0001,
+      "loss": 1.543,
+      "step": 10311
+    },
+    {
+      "epoch": 1.1980249782166714,
+      "grad_norm": 0.5252891182899475,
+      "learning_rate": 0.0001,
+      "loss": 1.3211,
+      "step": 10312
+    },
+    {
+      "epoch": 1.198141155968632,
+      "grad_norm": 0.545369029045105,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 10313
+    },
+    {
+      "epoch": 1.1982573337205924,
+      "grad_norm": 0.5860430598258972,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 10314
+    },
+    {
+      "epoch": 1.1983735114725529,
+      "grad_norm": 0.5563281774520874,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 10315
+    },
+    {
+      "epoch": 1.1984896892245136,
+      "grad_norm": 0.5620051622390747,
+      "learning_rate": 0.0001,
+      "loss": 1.3899,
+      "step": 10316
+    },
+    {
+      "epoch": 1.198605866976474,
+      "grad_norm": 0.5942341685295105,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 10317
+    },
+    {
+      "epoch": 1.1987220447284346,
+      "grad_norm": 0.6117276549339294,
+      "learning_rate": 0.0001,
+      "loss": 1.6346,
+      "step": 10318
+    },
+    {
+      "epoch": 1.198838222480395,
+      "grad_norm": 0.5562477707862854,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 10319
+    },
+    {
+      "epoch": 1.1989544002323556,
+      "grad_norm": 0.5761712193489075,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 10320
+    },
+    {
+      "epoch": 1.199070577984316,
+      "grad_norm": 0.552450954914093,
+      "learning_rate": 0.0001,
+      "loss": 1.3354,
+      "step": 10321
+    },
+    {
+      "epoch": 1.1991867557362765,
+      "grad_norm": 0.5395271182060242,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 10322
+    },
+    {
+      "epoch": 1.199302933488237,
+      "grad_norm": 0.555846631526947,
+      "learning_rate": 0.0001,
+      "loss": 1.4032,
+      "step": 10323
+    },
+    {
+      "epoch": 1.1994191112401975,
+      "grad_norm": 0.5601608157157898,
+      "learning_rate": 0.0001,
+      "loss": 1.591,
+      "step": 10324
+    },
+    {
+      "epoch": 1.199535288992158,
+      "grad_norm": 0.5598324537277222,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 10325
+    },
+    {
+      "epoch": 1.1996514667441185,
+      "grad_norm": 0.5127769112586975,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 10326
+    },
+    {
+      "epoch": 1.199767644496079,
+      "grad_norm": 0.5928462147712708,
+      "learning_rate": 0.0001,
+      "loss": 1.6712,
+      "step": 10327
+    },
+    {
+      "epoch": 1.1998838222480395,
+      "grad_norm": 0.5600936412811279,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 10328
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.5994905233383179,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 10329
+    },
+    {
+      "epoch": 1.2001161777519604,
+      "grad_norm": 0.5742284059524536,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 10330
+    },
+    {
+      "epoch": 1.200232355503921,
+      "grad_norm": 0.5460923314094543,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 10331
+    },
+    {
+      "epoch": 1.2003485332558814,
+      "grad_norm": 0.5468488931655884,
+      "learning_rate": 0.0001,
+      "loss": 1.4799,
+      "step": 10332
+    },
+    {
+      "epoch": 1.200464711007842,
+      "grad_norm": 0.5500593781471252,
+      "learning_rate": 0.0001,
+      "loss": 1.2928,
+      "step": 10333
+    },
+    {
+      "epoch": 1.2005808887598026,
+      "grad_norm": 0.5604000091552734,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 10334
+    },
+    {
+      "epoch": 1.2006970665117631,
+      "grad_norm": 0.5383157730102539,
+      "learning_rate": 0.0001,
+      "loss": 1.414,
+      "step": 10335
+    },
+    {
+      "epoch": 1.2008132442637236,
+      "grad_norm": 0.6212254166603088,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 10336
+    },
+    {
+      "epoch": 1.200929422015684,
+      "grad_norm": 0.5845767259597778,
+      "learning_rate": 0.0001,
+      "loss": 1.4383,
+      "step": 10337
+    },
+    {
+      "epoch": 1.2010455997676446,
+      "grad_norm": 0.517145574092865,
+      "learning_rate": 0.0001,
+      "loss": 1.2527,
+      "step": 10338
+    },
+    {
+      "epoch": 1.201161777519605,
+      "grad_norm": 0.617876410484314,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 10339
+    },
+    {
+      "epoch": 1.2012779552715656,
+      "grad_norm": 0.5485627055168152,
+      "learning_rate": 0.0001,
+      "loss": 1.3546,
+      "step": 10340
+    },
+    {
+      "epoch": 1.201394133023526,
+      "grad_norm": 0.5494995713233948,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 10341
+    },
+    {
+      "epoch": 1.2015103107754865,
+      "grad_norm": 0.5931583642959595,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 10342
+    },
+    {
+      "epoch": 1.201626488527447,
+      "grad_norm": 0.57831871509552,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 10343
+    },
+    {
+      "epoch": 1.2017426662794075,
+      "grad_norm": 0.5546940565109253,
+      "learning_rate": 0.0001,
+      "loss": 1.2925,
+      "step": 10344
+    },
+    {
+      "epoch": 1.201858844031368,
+      "grad_norm": 0.5821717381477356,
+      "learning_rate": 0.0001,
+      "loss": 1.4944,
+      "step": 10345
+    },
+    {
+      "epoch": 1.2019750217833285,
+      "grad_norm": 0.5914159417152405,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 10346
+    },
+    {
+      "epoch": 1.202091199535289,
+      "grad_norm": 0.5691181421279907,
+      "learning_rate": 0.0001,
+      "loss": 1.3892,
+      "step": 10347
+    },
+    {
+      "epoch": 1.2022073772872495,
+      "grad_norm": 0.5852946639060974,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 10348
+    },
+    {
+      "epoch": 1.20232355503921,
+      "grad_norm": 0.5942548513412476,
+      "learning_rate": 0.0001,
+      "loss": 1.3972,
+      "step": 10349
+    },
+    {
+      "epoch": 1.2024397327911704,
+      "grad_norm": 0.595831036567688,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 10350
+    },
+    {
+      "epoch": 1.202555910543131,
+      "grad_norm": 0.5792760848999023,
+      "learning_rate": 0.0001,
+      "loss": 1.4927,
+      "step": 10351
+    },
+    {
+      "epoch": 1.2026720882950914,
+      "grad_norm": 0.5801286101341248,
+      "learning_rate": 0.0001,
+      "loss": 1.3312,
+      "step": 10352
+    },
+    {
+      "epoch": 1.202788266047052,
+      "grad_norm": 0.6192400455474854,
+      "learning_rate": 0.0001,
+      "loss": 1.746,
+      "step": 10353
+    },
+    {
+      "epoch": 1.2029044437990124,
+      "grad_norm": 0.6081662774085999,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 10354
+    },
+    {
+      "epoch": 1.203020621550973,
+      "grad_norm": 0.6102173924446106,
+      "learning_rate": 0.0001,
+      "loss": 1.6073,
+      "step": 10355
+    },
+    {
+      "epoch": 1.2031367993029334,
+      "grad_norm": 0.5657384395599365,
+      "learning_rate": 0.0001,
+      "loss": 1.5184,
+      "step": 10356
+    },
+    {
+      "epoch": 1.2032529770548939,
+      "grad_norm": 0.5585178732872009,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 10357
+    },
+    {
+      "epoch": 1.2033691548068546,
+      "grad_norm": 0.5775259137153625,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 10358
+    },
+    {
+      "epoch": 1.203485332558815,
+      "grad_norm": 0.557350754737854,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 10359
+    },
+    {
+      "epoch": 1.2036015103107756,
+      "grad_norm": 0.5576322078704834,
+      "learning_rate": 0.0001,
+      "loss": 1.7029,
+      "step": 10360
+    },
+    {
+      "epoch": 1.203717688062736,
+      "grad_norm": 0.5494878888130188,
+      "learning_rate": 0.0001,
+      "loss": 1.4551,
+      "step": 10361
+    },
+    {
+      "epoch": 1.2038338658146965,
+      "grad_norm": 0.580441951751709,
+      "learning_rate": 0.0001,
+      "loss": 1.5666,
+      "step": 10362
+    },
+    {
+      "epoch": 1.203950043566657,
+      "grad_norm": 0.5665770173072815,
+      "learning_rate": 0.0001,
+      "loss": 1.5302,
+      "step": 10363
+    },
+    {
+      "epoch": 1.2040662213186175,
+      "grad_norm": 0.5588644742965698,
+      "learning_rate": 0.0001,
+      "loss": 1.3927,
+      "step": 10364
+    },
+    {
+      "epoch": 1.204182399070578,
+      "grad_norm": 0.6142786741256714,
+      "learning_rate": 0.0001,
+      "loss": 1.5917,
+      "step": 10365
+    },
+    {
+      "epoch": 1.2042985768225385,
+      "grad_norm": 0.6669145822525024,
+      "learning_rate": 0.0001,
+      "loss": 1.5222,
+      "step": 10366
+    },
+    {
+      "epoch": 1.204414754574499,
+      "grad_norm": 0.5337054133415222,
+      "learning_rate": 0.0001,
+      "loss": 1.3741,
+      "step": 10367
+    },
+    {
+      "epoch": 1.2045309323264595,
+      "grad_norm": 0.5432060956954956,
+      "learning_rate": 0.0001,
+      "loss": 1.3621,
+      "step": 10368
+    },
+    {
+      "epoch": 1.20464711007842,
+      "grad_norm": 0.6041423678398132,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 10369
+    },
+    {
+      "epoch": 1.2047632878303804,
+      "grad_norm": 0.5839071869850159,
+      "learning_rate": 0.0001,
+      "loss": 1.4869,
+      "step": 10370
+    },
+    {
+      "epoch": 1.204879465582341,
+      "grad_norm": 0.5747912526130676,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 10371
+    },
+    {
+      "epoch": 1.2049956433343014,
+      "grad_norm": 0.5971471667289734,
+      "learning_rate": 0.0001,
+      "loss": 1.5796,
+      "step": 10372
+    },
+    {
+      "epoch": 1.205111821086262,
+      "grad_norm": 0.5386415123939514,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 10373
+    },
+    {
+      "epoch": 1.2052279988382224,
+      "grad_norm": 0.5704327821731567,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 10374
+    },
+    {
+      "epoch": 1.205344176590183,
+      "grad_norm": 0.5206299424171448,
+      "learning_rate": 0.0001,
+      "loss": 1.4509,
+      "step": 10375
+    },
+    {
+      "epoch": 1.2054603543421436,
+      "grad_norm": 0.5865551829338074,
+      "learning_rate": 0.0001,
+      "loss": 1.342,
+      "step": 10376
+    },
+    {
+      "epoch": 1.205576532094104,
+      "grad_norm": 0.5369005799293518,
+      "learning_rate": 0.0001,
+      "loss": 1.487,
+      "step": 10377
+    },
+    {
+      "epoch": 1.2056927098460646,
+      "grad_norm": 0.5041610598564148,
+      "learning_rate": 0.0001,
+      "loss": 1.1966,
+      "step": 10378
+    },
+    {
+      "epoch": 1.205808887598025,
+      "grad_norm": 0.5628383755683899,
+      "learning_rate": 0.0001,
+      "loss": 1.6165,
+      "step": 10379
+    },
+    {
+      "epoch": 1.2059250653499856,
+      "grad_norm": 0.5555375218391418,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 10380
+    },
+    {
+      "epoch": 1.206041243101946,
+      "grad_norm": 0.563068687915802,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 10381
+    },
+    {
+      "epoch": 1.2061574208539065,
+      "grad_norm": 0.5873087644577026,
+      "learning_rate": 0.0001,
+      "loss": 1.604,
+      "step": 10382
+    },
+    {
+      "epoch": 1.206273598605867,
+      "grad_norm": 0.5575566291809082,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 10383
+    },
+    {
+      "epoch": 1.2063897763578275,
+      "grad_norm": 0.58903968334198,
+      "learning_rate": 0.0001,
+      "loss": 1.3605,
+      "step": 10384
+    },
+    {
+      "epoch": 1.206505954109788,
+      "grad_norm": 0.5939505696296692,
+      "learning_rate": 0.0001,
+      "loss": 1.7172,
+      "step": 10385
+    },
+    {
+      "epoch": 1.2066221318617485,
+      "grad_norm": 0.5417733192443848,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 10386
+    },
+    {
+      "epoch": 1.206738309613709,
+      "grad_norm": 0.52766352891922,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 10387
+    },
+    {
+      "epoch": 1.2068544873656695,
+      "grad_norm": 0.5897431373596191,
+      "learning_rate": 0.0001,
+      "loss": 1.4175,
+      "step": 10388
+    },
+    {
+      "epoch": 1.20697066511763,
+      "grad_norm": 0.6240753531455994,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 10389
+    },
+    {
+      "epoch": 1.2070868428695904,
+      "grad_norm": 0.6030203104019165,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 10390
+    },
+    {
+      "epoch": 1.207203020621551,
+      "grad_norm": 0.6183759570121765,
+      "learning_rate": 0.0001,
+      "loss": 1.4371,
+      "step": 10391
+    },
+    {
+      "epoch": 1.2073191983735114,
+      "grad_norm": 0.5684889554977417,
+      "learning_rate": 0.0001,
+      "loss": 1.4891,
+      "step": 10392
+    },
+    {
+      "epoch": 1.207435376125472,
+      "grad_norm": 0.5397784113883972,
+      "learning_rate": 0.0001,
+      "loss": 1.3721,
+      "step": 10393
+    },
+    {
+      "epoch": 1.2075515538774324,
+      "grad_norm": 0.5187193751335144,
+      "learning_rate": 0.0001,
+      "loss": 1.3977,
+      "step": 10394
+    },
+    {
+      "epoch": 1.207667731629393,
+      "grad_norm": 0.5413998961448669,
+      "learning_rate": 0.0001,
+      "loss": 1.3461,
+      "step": 10395
+    },
+    {
+      "epoch": 1.2077839093813534,
+      "grad_norm": 0.545616626739502,
+      "learning_rate": 0.0001,
+      "loss": 1.3606,
+      "step": 10396
+    },
+    {
+      "epoch": 1.2079000871333139,
+      "grad_norm": 0.5437708497047424,
+      "learning_rate": 0.0001,
+      "loss": 1.4803,
+      "step": 10397
+    },
+    {
+      "epoch": 1.2080162648852744,
+      "grad_norm": 0.56759113073349,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 10398
+    },
+    {
+      "epoch": 1.2081324426372348,
+      "grad_norm": 0.5678495764732361,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 10399
+    },
+    {
+      "epoch": 1.2082486203891956,
+      "grad_norm": 0.5646939277648926,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 10400
+    },
+    {
+      "epoch": 1.208364798141156,
+      "grad_norm": 0.554522693157196,
+      "learning_rate": 0.0001,
+      "loss": 1.4276,
+      "step": 10401
+    },
+    {
+      "epoch": 1.2084809758931165,
+      "grad_norm": 0.5794949531555176,
+      "learning_rate": 0.0001,
+      "loss": 1.438,
+      "step": 10402
+    },
+    {
+      "epoch": 1.208597153645077,
+      "grad_norm": 0.5517016053199768,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 10403
+    },
+    {
+      "epoch": 1.2087133313970375,
+      "grad_norm": 0.5888990759849548,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 10404
+    },
+    {
+      "epoch": 1.208829509148998,
+      "grad_norm": 0.5803018808364868,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 10405
+    },
+    {
+      "epoch": 1.2089456869009585,
+      "grad_norm": 0.5619463324546814,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 10406
+    },
+    {
+      "epoch": 1.209061864652919,
+      "grad_norm": 0.5671004056930542,
+      "learning_rate": 0.0001,
+      "loss": 1.7036,
+      "step": 10407
+    },
+    {
+      "epoch": 1.2091780424048795,
+      "grad_norm": 0.5472593307495117,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 10408
+    },
+    {
+      "epoch": 1.20929422015684,
+      "grad_norm": 0.5378280878067017,
+      "learning_rate": 0.0001,
+      "loss": 1.4031,
+      "step": 10409
+    },
+    {
+      "epoch": 1.2094103979088004,
+      "grad_norm": 0.5641921758651733,
+      "learning_rate": 0.0001,
+      "loss": 1.3906,
+      "step": 10410
+    },
+    {
+      "epoch": 1.209526575660761,
+      "grad_norm": 0.5625640749931335,
+      "learning_rate": 0.0001,
+      "loss": 1.4926,
+      "step": 10411
+    },
+    {
+      "epoch": 1.2096427534127214,
+      "grad_norm": 0.5496149063110352,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 10412
+    },
+    {
+      "epoch": 1.209758931164682,
+      "grad_norm": 0.5828267931938171,
+      "learning_rate": 0.0001,
+      "loss": 1.6947,
+      "step": 10413
+    },
+    {
+      "epoch": 1.2098751089166424,
+      "grad_norm": 0.5525361895561218,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 10414
+    },
+    {
+      "epoch": 1.209991286668603,
+      "grad_norm": 0.524874746799469,
+      "learning_rate": 0.0001,
+      "loss": 1.3639,
+      "step": 10415
+    },
+    {
+      "epoch": 1.2101074644205634,
+      "grad_norm": 0.6062628626823425,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 10416
+    },
+    {
+      "epoch": 1.2102236421725239,
+      "grad_norm": 0.5272751450538635,
+      "learning_rate": 0.0001,
+      "loss": 1.4494,
+      "step": 10417
+    },
+    {
+      "epoch": 1.2103398199244846,
+      "grad_norm": 0.5667693614959717,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 10418
+    },
+    {
+      "epoch": 1.210455997676445,
+      "grad_norm": 0.5765136480331421,
+      "learning_rate": 0.0001,
+      "loss": 1.3549,
+      "step": 10419
+    },
+    {
+      "epoch": 1.2105721754284056,
+      "grad_norm": 0.5195456743240356,
+      "learning_rate": 0.0001,
+      "loss": 1.286,
+      "step": 10420
+    },
+    {
+      "epoch": 1.210688353180366,
+      "grad_norm": 0.6368436813354492,
+      "learning_rate": 0.0001,
+      "loss": 1.6614,
+      "step": 10421
+    },
+    {
+      "epoch": 1.2108045309323265,
+      "grad_norm": 0.5609405040740967,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 10422
+    },
+    {
+      "epoch": 1.210920708684287,
+      "grad_norm": 0.6603508591651917,
+      "learning_rate": 0.0001,
+      "loss": 1.7794,
+      "step": 10423
+    },
+    {
+      "epoch": 1.2110368864362475,
+      "grad_norm": 0.5625258684158325,
+      "learning_rate": 0.0001,
+      "loss": 1.4226,
+      "step": 10424
+    },
+    {
+      "epoch": 1.211153064188208,
+      "grad_norm": 0.5961827039718628,
+      "learning_rate": 0.0001,
+      "loss": 1.4507,
+      "step": 10425
+    },
+    {
+      "epoch": 1.2112692419401685,
+      "grad_norm": 0.5945589542388916,
+      "learning_rate": 0.0001,
+      "loss": 1.7156,
+      "step": 10426
+    },
+    {
+      "epoch": 1.211385419692129,
+      "grad_norm": 0.547545850276947,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 10427
+    },
+    {
+      "epoch": 1.2115015974440895,
+      "grad_norm": 0.5815989971160889,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 10428
+    },
+    {
+      "epoch": 1.21161777519605,
+      "grad_norm": 0.5638428926467896,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 10429
+    },
+    {
+      "epoch": 1.2117339529480105,
+      "grad_norm": 0.5702305436134338,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 10430
+    },
+    {
+      "epoch": 1.211850130699971,
+      "grad_norm": 0.5992822051048279,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 10431
+    },
+    {
+      "epoch": 1.2119663084519314,
+      "grad_norm": 0.5825338363647461,
+      "learning_rate": 0.0001,
+      "loss": 1.5207,
+      "step": 10432
+    },
+    {
+      "epoch": 1.212082486203892,
+      "grad_norm": 0.5452916622161865,
+      "learning_rate": 0.0001,
+      "loss": 1.3638,
+      "step": 10433
+    },
+    {
+      "epoch": 1.2121986639558524,
+      "grad_norm": 0.6069163084030151,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 10434
+    },
+    {
+      "epoch": 1.212314841707813,
+      "grad_norm": 0.6051025390625,
+      "learning_rate": 0.0001,
+      "loss": 1.608,
+      "step": 10435
+    },
+    {
+      "epoch": 1.2124310194597734,
+      "grad_norm": 0.5363568663597107,
+      "learning_rate": 0.0001,
+      "loss": 1.274,
+      "step": 10436
+    },
+    {
+      "epoch": 1.2125471972117339,
+      "grad_norm": 0.5586710572242737,
+      "learning_rate": 0.0001,
+      "loss": 1.3582,
+      "step": 10437
+    },
+    {
+      "epoch": 1.2126633749636944,
+      "grad_norm": 0.5546153783798218,
+      "learning_rate": 0.0001,
+      "loss": 1.4534,
+      "step": 10438
+    },
+    {
+      "epoch": 1.2127795527156549,
+      "grad_norm": 0.5305795073509216,
+      "learning_rate": 0.0001,
+      "loss": 1.4127,
+      "step": 10439
+    },
+    {
+      "epoch": 1.2128957304676153,
+      "grad_norm": 0.5443617701530457,
+      "learning_rate": 0.0001,
+      "loss": 1.453,
+      "step": 10440
+    },
+    {
+      "epoch": 1.2130119082195758,
+      "grad_norm": 0.6291701197624207,
+      "learning_rate": 0.0001,
+      "loss": 1.7236,
+      "step": 10441
+    },
+    {
+      "epoch": 1.2131280859715365,
+      "grad_norm": 0.6024119853973389,
+      "learning_rate": 0.0001,
+      "loss": 1.5023,
+      "step": 10442
+    },
+    {
+      "epoch": 1.213244263723497,
+      "grad_norm": 0.566421389579773,
+      "learning_rate": 0.0001,
+      "loss": 1.3462,
+      "step": 10443
+    },
+    {
+      "epoch": 1.2133604414754575,
+      "grad_norm": 0.6556349992752075,
+      "learning_rate": 0.0001,
+      "loss": 1.4306,
+      "step": 10444
+    },
+    {
+      "epoch": 1.213476619227418,
+      "grad_norm": 0.5445832014083862,
+      "learning_rate": 0.0001,
+      "loss": 1.3331,
+      "step": 10445
+    },
+    {
+      "epoch": 1.2135927969793785,
+      "grad_norm": 0.5676811933517456,
+      "learning_rate": 0.0001,
+      "loss": 1.585,
+      "step": 10446
+    },
+    {
+      "epoch": 1.213708974731339,
+      "grad_norm": 0.5820676684379578,
+      "learning_rate": 0.0001,
+      "loss": 1.4923,
+      "step": 10447
+    },
+    {
+      "epoch": 1.2138251524832995,
+      "grad_norm": 0.5882712602615356,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 10448
+    },
+    {
+      "epoch": 1.21394133023526,
+      "grad_norm": 0.5337011218070984,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 10449
+    },
+    {
+      "epoch": 1.2140575079872205,
+      "grad_norm": 0.6516126394271851,
+      "learning_rate": 0.0001,
+      "loss": 1.4206,
+      "step": 10450
+    },
+    {
+      "epoch": 1.214173685739181,
+      "grad_norm": 0.537398099899292,
+      "learning_rate": 0.0001,
+      "loss": 1.3852,
+      "step": 10451
+    },
+    {
+      "epoch": 1.2142898634911414,
+      "grad_norm": 0.5600173473358154,
+      "learning_rate": 0.0001,
+      "loss": 1.2863,
+      "step": 10452
+    },
+    {
+      "epoch": 1.214406041243102,
+      "grad_norm": 0.5451828241348267,
+      "learning_rate": 0.0001,
+      "loss": 1.3837,
+      "step": 10453
+    },
+    {
+      "epoch": 1.2145222189950624,
+      "grad_norm": 0.6120473742485046,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 10454
+    },
+    {
+      "epoch": 1.214638396747023,
+      "grad_norm": 0.5406019687652588,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 10455
+    },
+    {
+      "epoch": 1.2147545744989834,
+      "grad_norm": 0.49340879917144775,
+      "learning_rate": 0.0001,
+      "loss": 1.2085,
+      "step": 10456
+    },
+    {
+      "epoch": 1.2148707522509439,
+      "grad_norm": 0.5855630040168762,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 10457
+    },
+    {
+      "epoch": 1.2149869300029044,
+      "grad_norm": 0.5757850408554077,
+      "learning_rate": 0.0001,
+      "loss": 1.4777,
+      "step": 10458
+    },
+    {
+      "epoch": 1.2151031077548649,
+      "grad_norm": 0.6049354076385498,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 10459
+    },
+    {
+      "epoch": 1.2152192855068256,
+      "grad_norm": 0.5614561438560486,
+      "learning_rate": 0.0001,
+      "loss": 1.1913,
+      "step": 10460
+    },
+    {
+      "epoch": 1.215335463258786,
+      "grad_norm": 0.58977210521698,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 10461
+    },
+    {
+      "epoch": 1.2154516410107465,
+      "grad_norm": 0.5945138335227966,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 10462
+    },
+    {
+      "epoch": 1.215567818762707,
+      "grad_norm": 0.5707786083221436,
+      "learning_rate": 0.0001,
+      "loss": 1.4646,
+      "step": 10463
+    },
+    {
+      "epoch": 1.2156839965146675,
+      "grad_norm": 0.6250419020652771,
+      "learning_rate": 0.0001,
+      "loss": 1.5176,
+      "step": 10464
+    },
+    {
+      "epoch": 1.215800174266628,
+      "grad_norm": 0.5555652379989624,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 10465
+    },
+    {
+      "epoch": 1.2159163520185885,
+      "grad_norm": 0.5932562947273254,
+      "learning_rate": 0.0001,
+      "loss": 1.3049,
+      "step": 10466
+    },
+    {
+      "epoch": 1.216032529770549,
+      "grad_norm": 0.5786982178688049,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 10467
+    },
+    {
+      "epoch": 1.2161487075225095,
+      "grad_norm": 0.5523450970649719,
+      "learning_rate": 0.0001,
+      "loss": 1.5297,
+      "step": 10468
+    },
+    {
+      "epoch": 1.21626488527447,
+      "grad_norm": 0.5957953333854675,
+      "learning_rate": 0.0001,
+      "loss": 1.4569,
+      "step": 10469
+    },
+    {
+      "epoch": 1.2163810630264305,
+      "grad_norm": 0.6025353074073792,
+      "learning_rate": 0.0001,
+      "loss": 1.6472,
+      "step": 10470
+    },
+    {
+      "epoch": 1.216497240778391,
+      "grad_norm": 0.6158897280693054,
+      "learning_rate": 0.0001,
+      "loss": 1.6028,
+      "step": 10471
+    },
+    {
+      "epoch": 1.2166134185303514,
+      "grad_norm": 0.5769653916358948,
+      "learning_rate": 0.0001,
+      "loss": 1.4907,
+      "step": 10472
+    },
+    {
+      "epoch": 1.216729596282312,
+      "grad_norm": 0.5845011472702026,
+      "learning_rate": 0.0001,
+      "loss": 1.6525,
+      "step": 10473
+    },
+    {
+      "epoch": 1.2168457740342724,
+      "grad_norm": 0.5649383068084717,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 10474
+    },
+    {
+      "epoch": 1.216961951786233,
+      "grad_norm": 0.5390584468841553,
+      "learning_rate": 0.0001,
+      "loss": 1.4094,
+      "step": 10475
+    },
+    {
+      "epoch": 1.2170781295381934,
+      "grad_norm": 0.5726324319839478,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 10476
+    },
+    {
+      "epoch": 1.2171943072901539,
+      "grad_norm": 0.5737690329551697,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 10477
+    },
+    {
+      "epoch": 1.2173104850421144,
+      "grad_norm": 0.5977112650871277,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 10478
+    },
+    {
+      "epoch": 1.2174266627940749,
+      "grad_norm": 0.5659692287445068,
+      "learning_rate": 0.0001,
+      "loss": 1.6041,
+      "step": 10479
+    },
+    {
+      "epoch": 1.2175428405460353,
+      "grad_norm": 0.6409531831741333,
+      "learning_rate": 0.0001,
+      "loss": 1.5836,
+      "step": 10480
+    },
+    {
+      "epoch": 1.2176590182979958,
+      "grad_norm": 0.5651158094406128,
+      "learning_rate": 0.0001,
+      "loss": 1.6229,
+      "step": 10481
+    },
+    {
+      "epoch": 1.2177751960499563,
+      "grad_norm": 0.5669698119163513,
+      "learning_rate": 0.0001,
+      "loss": 1.4251,
+      "step": 10482
+    },
+    {
+      "epoch": 1.2178913738019168,
+      "grad_norm": 0.5658094882965088,
+      "learning_rate": 0.0001,
+      "loss": 1.5442,
+      "step": 10483
+    },
+    {
+      "epoch": 1.2180075515538775,
+      "grad_norm": 0.5667807459831238,
+      "learning_rate": 0.0001,
+      "loss": 1.4133,
+      "step": 10484
+    },
+    {
+      "epoch": 1.218123729305838,
+      "grad_norm": 0.5489072203636169,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 10485
+    },
+    {
+      "epoch": 1.2182399070577985,
+      "grad_norm": 0.6018260717391968,
+      "learning_rate": 0.0001,
+      "loss": 1.7007,
+      "step": 10486
+    },
+    {
+      "epoch": 1.218356084809759,
+      "grad_norm": 0.5693395137786865,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 10487
+    },
+    {
+      "epoch": 1.2184722625617195,
+      "grad_norm": 0.5397615432739258,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 10488
+    },
+    {
+      "epoch": 1.21858844031368,
+      "grad_norm": 0.5639338493347168,
+      "learning_rate": 0.0001,
+      "loss": 1.3537,
+      "step": 10489
+    },
+    {
+      "epoch": 1.2187046180656405,
+      "grad_norm": 0.5436851382255554,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 10490
+    },
+    {
+      "epoch": 1.218820795817601,
+      "grad_norm": 0.5400302410125732,
+      "learning_rate": 0.0001,
+      "loss": 1.401,
+      "step": 10491
+    },
+    {
+      "epoch": 1.2189369735695614,
+      "grad_norm": 0.5486143231391907,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 10492
+    },
+    {
+      "epoch": 1.219053151321522,
+      "grad_norm": 0.6082191467285156,
+      "learning_rate": 0.0001,
+      "loss": 1.4852,
+      "step": 10493
+    },
+    {
+      "epoch": 1.2191693290734824,
+      "grad_norm": 0.5857893228530884,
+      "learning_rate": 0.0001,
+      "loss": 1.3528,
+      "step": 10494
+    },
+    {
+      "epoch": 1.219285506825443,
+      "grad_norm": 0.592745304107666,
+      "learning_rate": 0.0001,
+      "loss": 1.4238,
+      "step": 10495
+    },
+    {
+      "epoch": 1.2194016845774034,
+      "grad_norm": 0.5825994610786438,
+      "learning_rate": 0.0001,
+      "loss": 1.5337,
+      "step": 10496
+    },
+    {
+      "epoch": 1.2195178623293639,
+      "grad_norm": 0.5833573341369629,
+      "learning_rate": 0.0001,
+      "loss": 1.437,
+      "step": 10497
+    },
+    {
+      "epoch": 1.2196340400813244,
+      "grad_norm": 0.5565202236175537,
+      "learning_rate": 0.0001,
+      "loss": 1.3335,
+      "step": 10498
+    },
+    {
+      "epoch": 1.2197502178332849,
+      "grad_norm": 0.5850576758384705,
+      "learning_rate": 0.0001,
+      "loss": 1.5889,
+      "step": 10499
+    },
+    {
+      "epoch": 1.2198663955852453,
+      "grad_norm": 0.5984856486320496,
+      "learning_rate": 0.0001,
+      "loss": 1.3888,
+      "step": 10500
+    },
+    {
+      "epoch": 1.2199825733372058,
+      "grad_norm": 0.5830519795417786,
+      "learning_rate": 0.0001,
+      "loss": 1.6643,
+      "step": 10501
+    },
+    {
+      "epoch": 1.2200987510891665,
+      "grad_norm": 0.5867604613304138,
+      "learning_rate": 0.0001,
+      "loss": 1.2365,
+      "step": 10502
+    },
+    {
+      "epoch": 1.220214928841127,
+      "grad_norm": 0.5673847198486328,
+      "learning_rate": 0.0001,
+      "loss": 1.3835,
+      "step": 10503
+    },
+    {
+      "epoch": 1.2203311065930875,
+      "grad_norm": 0.5929017066955566,
+      "learning_rate": 0.0001,
+      "loss": 1.3362,
+      "step": 10504
+    },
+    {
+      "epoch": 1.220447284345048,
+      "grad_norm": 0.5458045601844788,
+      "learning_rate": 0.0001,
+      "loss": 1.356,
+      "step": 10505
+    },
+    {
+      "epoch": 1.2205634620970085,
+      "grad_norm": 0.6023081541061401,
+      "learning_rate": 0.0001,
+      "loss": 1.4702,
+      "step": 10506
+    },
+    {
+      "epoch": 1.220679639848969,
+      "grad_norm": 0.6090680956840515,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 10507
+    },
+    {
+      "epoch": 1.2207958176009295,
+      "grad_norm": 0.5901305079460144,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 10508
+    },
+    {
+      "epoch": 1.22091199535289,
+      "grad_norm": 0.6401813626289368,
+      "learning_rate": 0.0001,
+      "loss": 1.6139,
+      "step": 10509
+    },
+    {
+      "epoch": 1.2210281731048505,
+      "grad_norm": 0.6004911661148071,
+      "learning_rate": 0.0001,
+      "loss": 1.4363,
+      "step": 10510
+    },
+    {
+      "epoch": 1.221144350856811,
+      "grad_norm": 0.5764592289924622,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 10511
+    },
+    {
+      "epoch": 1.2212605286087714,
+      "grad_norm": 0.5585955381393433,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 10512
+    },
+    {
+      "epoch": 1.221376706360732,
+      "grad_norm": 0.5347375273704529,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 10513
+    },
+    {
+      "epoch": 1.2214928841126924,
+      "grad_norm": 0.5225241780281067,
+      "learning_rate": 0.0001,
+      "loss": 1.35,
+      "step": 10514
+    },
+    {
+      "epoch": 1.221609061864653,
+      "grad_norm": 0.5799035429954529,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 10515
+    },
+    {
+      "epoch": 1.2217252396166134,
+      "grad_norm": 0.5327029824256897,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 10516
+    },
+    {
+      "epoch": 1.2218414173685739,
+      "grad_norm": 0.5418916344642639,
+      "learning_rate": 0.0001,
+      "loss": 1.4701,
+      "step": 10517
+    },
+    {
+      "epoch": 1.2219575951205344,
+      "grad_norm": 0.5544294714927673,
+      "learning_rate": 0.0001,
+      "loss": 1.3398,
+      "step": 10518
+    },
+    {
+      "epoch": 1.2220737728724949,
+      "grad_norm": 0.6825329065322876,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 10519
+    },
+    {
+      "epoch": 1.2221899506244553,
+      "grad_norm": 0.5987062454223633,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 10520
+    },
+    {
+      "epoch": 1.2223061283764158,
+      "grad_norm": 0.6384679675102234,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 10521
+    },
+    {
+      "epoch": 1.2224223061283763,
+      "grad_norm": 0.5685727000236511,
+      "learning_rate": 0.0001,
+      "loss": 1.3943,
+      "step": 10522
+    },
+    {
+      "epoch": 1.2225384838803368,
+      "grad_norm": 0.5603922009468079,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 10523
+    },
+    {
+      "epoch": 1.2226546616322973,
+      "grad_norm": 0.5447299480438232,
+      "learning_rate": 0.0001,
+      "loss": 1.2081,
+      "step": 10524
+    },
+    {
+      "epoch": 1.2227708393842578,
+      "grad_norm": 0.6237144470214844,
+      "learning_rate": 0.0001,
+      "loss": 1.7241,
+      "step": 10525
+    },
+    {
+      "epoch": 1.2228870171362185,
+      "grad_norm": 0.567017674446106,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 10526
+    },
+    {
+      "epoch": 1.223003194888179,
+      "grad_norm": 0.5752435326576233,
+      "learning_rate": 0.0001,
+      "loss": 1.3276,
+      "step": 10527
+    },
+    {
+      "epoch": 1.2231193726401395,
+      "grad_norm": 0.5610300898551941,
+      "learning_rate": 0.0001,
+      "loss": 1.4231,
+      "step": 10528
+    },
+    {
+      "epoch": 1.2232355503921,
+      "grad_norm": 0.5971804857254028,
+      "learning_rate": 0.0001,
+      "loss": 1.3758,
+      "step": 10529
+    },
+    {
+      "epoch": 1.2233517281440605,
+      "grad_norm": 0.5715692639350891,
+      "learning_rate": 0.0001,
+      "loss": 1.4178,
+      "step": 10530
+    },
+    {
+      "epoch": 1.223467905896021,
+      "grad_norm": 0.5297221541404724,
+      "learning_rate": 0.0001,
+      "loss": 1.3276,
+      "step": 10531
+    },
+    {
+      "epoch": 1.2235840836479814,
+      "grad_norm": 0.5569010376930237,
+      "learning_rate": 0.0001,
+      "loss": 1.3424,
+      "step": 10532
+    },
+    {
+      "epoch": 1.223700261399942,
+      "grad_norm": 0.5953126549720764,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 10533
+    },
+    {
+      "epoch": 1.2238164391519024,
+      "grad_norm": 0.6054908037185669,
+      "learning_rate": 0.0001,
+      "loss": 1.4562,
+      "step": 10534
+    },
+    {
+      "epoch": 1.223932616903863,
+      "grad_norm": 0.5700466632843018,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 10535
+    },
+    {
+      "epoch": 1.2240487946558234,
+      "grad_norm": 0.5606374144554138,
+      "learning_rate": 0.0001,
+      "loss": 1.4028,
+      "step": 10536
+    },
+    {
+      "epoch": 1.2241649724077839,
+      "grad_norm": 0.5463014245033264,
+      "learning_rate": 0.0001,
+      "loss": 1.5375,
+      "step": 10537
+    },
+    {
+      "epoch": 1.2242811501597444,
+      "grad_norm": 0.5794979333877563,
+      "learning_rate": 0.0001,
+      "loss": 1.4477,
+      "step": 10538
+    },
+    {
+      "epoch": 1.2243973279117049,
+      "grad_norm": 0.5675046443939209,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 10539
+    },
+    {
+      "epoch": 1.2245135056636653,
+      "grad_norm": 0.5605546236038208,
+      "learning_rate": 0.0001,
+      "loss": 1.4852,
+      "step": 10540
+    },
+    {
+      "epoch": 1.2246296834156258,
+      "grad_norm": 0.5566608905792236,
+      "learning_rate": 0.0001,
+      "loss": 1.3657,
+      "step": 10541
+    },
+    {
+      "epoch": 1.2247458611675863,
+      "grad_norm": 0.60628741979599,
+      "learning_rate": 0.0001,
+      "loss": 1.6062,
+      "step": 10542
+    },
+    {
+      "epoch": 1.2248620389195468,
+      "grad_norm": 0.5989949107170105,
+      "learning_rate": 0.0001,
+      "loss": 1.4082,
+      "step": 10543
+    },
+    {
+      "epoch": 1.2249782166715075,
+      "grad_norm": 0.601296067237854,
+      "learning_rate": 0.0001,
+      "loss": 1.5123,
+      "step": 10544
+    },
+    {
+      "epoch": 1.225094394423468,
+      "grad_norm": 0.6421616673469543,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 10545
+    },
+    {
+      "epoch": 1.2252105721754285,
+      "grad_norm": 0.5911616086959839,
+      "learning_rate": 0.0001,
+      "loss": 1.7276,
+      "step": 10546
+    },
+    {
+      "epoch": 1.225326749927389,
+      "grad_norm": 0.6583091020584106,
+      "learning_rate": 0.0001,
+      "loss": 1.6858,
+      "step": 10547
+    },
+    {
+      "epoch": 1.2254429276793495,
+      "grad_norm": 0.575006365776062,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 10548
+    },
+    {
+      "epoch": 1.22555910543131,
+      "grad_norm": 0.5844583511352539,
+      "learning_rate": 0.0001,
+      "loss": 1.5218,
+      "step": 10549
+    },
+    {
+      "epoch": 1.2256752831832705,
+      "grad_norm": 0.5701775550842285,
+      "learning_rate": 0.0001,
+      "loss": 1.5935,
+      "step": 10550
+    },
+    {
+      "epoch": 1.225791460935231,
+      "grad_norm": 0.5439273118972778,
+      "learning_rate": 0.0001,
+      "loss": 1.3318,
+      "step": 10551
+    },
+    {
+      "epoch": 1.2259076386871914,
+      "grad_norm": 0.5806834697723389,
+      "learning_rate": 0.0001,
+      "loss": 1.3744,
+      "step": 10552
+    },
+    {
+      "epoch": 1.226023816439152,
+      "grad_norm": 0.5335421562194824,
+      "learning_rate": 0.0001,
+      "loss": 1.1911,
+      "step": 10553
+    },
+    {
+      "epoch": 1.2261399941911124,
+      "grad_norm": 0.5800783038139343,
+      "learning_rate": 0.0001,
+      "loss": 1.4885,
+      "step": 10554
+    },
+    {
+      "epoch": 1.226256171943073,
+      "grad_norm": 0.5973320007324219,
+      "learning_rate": 0.0001,
+      "loss": 1.3504,
+      "step": 10555
+    },
+    {
+      "epoch": 1.2263723496950334,
+      "grad_norm": 0.5879946351051331,
+      "learning_rate": 0.0001,
+      "loss": 1.4086,
+      "step": 10556
+    },
+    {
+      "epoch": 1.2264885274469939,
+      "grad_norm": 0.621277928352356,
+      "learning_rate": 0.0001,
+      "loss": 1.5551,
+      "step": 10557
+    },
+    {
+      "epoch": 1.2266047051989544,
+      "grad_norm": 0.5970613360404968,
+      "learning_rate": 0.0001,
+      "loss": 1.2918,
+      "step": 10558
+    },
+    {
+      "epoch": 1.2267208829509149,
+      "grad_norm": 0.593627393245697,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 10559
+    },
+    {
+      "epoch": 1.2268370607028753,
+      "grad_norm": 0.6126559972763062,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 10560
+    },
+    {
+      "epoch": 1.2269532384548358,
+      "grad_norm": 0.5869115591049194,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 10561
+    },
+    {
+      "epoch": 1.2270694162067963,
+      "grad_norm": 0.587081789970398,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 10562
+    },
+    {
+      "epoch": 1.2271855939587568,
+      "grad_norm": 0.5886223316192627,
+      "learning_rate": 0.0001,
+      "loss": 1.5962,
+      "step": 10563
+    },
+    {
+      "epoch": 1.2273017717107173,
+      "grad_norm": 0.5749689936637878,
+      "learning_rate": 0.0001,
+      "loss": 1.4966,
+      "step": 10564
+    },
+    {
+      "epoch": 1.2274179494626778,
+      "grad_norm": 0.5669975876808167,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 10565
+    },
+    {
+      "epoch": 1.2275341272146383,
+      "grad_norm": 0.5260981917381287,
+      "learning_rate": 0.0001,
+      "loss": 1.4386,
+      "step": 10566
+    },
+    {
+      "epoch": 1.227650304966599,
+      "grad_norm": 0.5275533199310303,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 10567
+    },
+    {
+      "epoch": 1.2277664827185595,
+      "grad_norm": 0.5393403768539429,
+      "learning_rate": 0.0001,
+      "loss": 1.37,
+      "step": 10568
+    },
+    {
+      "epoch": 1.22788266047052,
+      "grad_norm": 0.5790442824363708,
+      "learning_rate": 0.0001,
+      "loss": 1.4346,
+      "step": 10569
+    },
+    {
+      "epoch": 1.2279988382224805,
+      "grad_norm": 0.579831063747406,
+      "learning_rate": 0.0001,
+      "loss": 1.3978,
+      "step": 10570
+    },
+    {
+      "epoch": 1.228115015974441,
+      "grad_norm": 0.5463832020759583,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 10571
+    },
+    {
+      "epoch": 1.2282311937264014,
+      "grad_norm": 0.551559567451477,
+      "learning_rate": 0.0001,
+      "loss": 1.4143,
+      "step": 10572
+    },
+    {
+      "epoch": 1.228347371478362,
+      "grad_norm": 0.544151782989502,
+      "learning_rate": 0.0001,
+      "loss": 1.3611,
+      "step": 10573
+    },
+    {
+      "epoch": 1.2284635492303224,
+      "grad_norm": 0.5910764932632446,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 10574
+    },
+    {
+      "epoch": 1.228579726982283,
+      "grad_norm": 0.5585051774978638,
+      "learning_rate": 0.0001,
+      "loss": 1.376,
+      "step": 10575
+    },
+    {
+      "epoch": 1.2286959047342434,
+      "grad_norm": 0.5787076354026794,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 10576
+    },
+    {
+      "epoch": 1.2288120824862039,
+      "grad_norm": 0.5803587436676025,
+      "learning_rate": 0.0001,
+      "loss": 1.5013,
+      "step": 10577
+    },
+    {
+      "epoch": 1.2289282602381644,
+      "grad_norm": 0.5648260116577148,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 10578
+    },
+    {
+      "epoch": 1.2290444379901249,
+      "grad_norm": 0.5887954235076904,
+      "learning_rate": 0.0001,
+      "loss": 1.6515,
+      "step": 10579
+    },
+    {
+      "epoch": 1.2291606157420853,
+      "grad_norm": 0.6301035284996033,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 10580
+    },
+    {
+      "epoch": 1.2292767934940458,
+      "grad_norm": 0.5514262914657593,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 10581
+    },
+    {
+      "epoch": 1.2293929712460063,
+      "grad_norm": 0.5427731275558472,
+      "learning_rate": 0.0001,
+      "loss": 1.5574,
+      "step": 10582
+    },
+    {
+      "epoch": 1.2295091489979668,
+      "grad_norm": 0.5385409593582153,
+      "learning_rate": 0.0001,
+      "loss": 1.3763,
+      "step": 10583
+    },
+    {
+      "epoch": 1.2296253267499273,
+      "grad_norm": 0.5767699480056763,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 10584
+    },
+    {
+      "epoch": 1.2297415045018878,
+      "grad_norm": 0.5255317687988281,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 10585
+    },
+    {
+      "epoch": 1.2298576822538485,
+      "grad_norm": 0.5520825982093811,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 10586
+    },
+    {
+      "epoch": 1.229973860005809,
+      "grad_norm": 0.5681281089782715,
+      "learning_rate": 0.0001,
+      "loss": 1.6923,
+      "step": 10587
+    },
+    {
+      "epoch": 1.2300900377577695,
+      "grad_norm": 0.5493902564048767,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 10588
+    },
+    {
+      "epoch": 1.23020621550973,
+      "grad_norm": 0.5706425309181213,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 10589
+    },
+    {
+      "epoch": 1.2303223932616905,
+      "grad_norm": 0.5630264282226562,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 10590
+    },
+    {
+      "epoch": 1.230438571013651,
+      "grad_norm": 0.5919108986854553,
+      "learning_rate": 0.0001,
+      "loss": 1.4042,
+      "step": 10591
+    },
+    {
+      "epoch": 1.2305547487656114,
+      "grad_norm": 0.5534734129905701,
+      "learning_rate": 0.0001,
+      "loss": 1.4322,
+      "step": 10592
+    },
+    {
+      "epoch": 1.230670926517572,
+      "grad_norm": 0.5363836884498596,
+      "learning_rate": 0.0001,
+      "loss": 1.336,
+      "step": 10593
+    },
+    {
+      "epoch": 1.2307871042695324,
+      "grad_norm": 0.5639822483062744,
+      "learning_rate": 0.0001,
+      "loss": 1.3161,
+      "step": 10594
+    },
+    {
+      "epoch": 1.230903282021493,
+      "grad_norm": 0.5711953639984131,
+      "learning_rate": 0.0001,
+      "loss": 1.5166,
+      "step": 10595
+    },
+    {
+      "epoch": 1.2310194597734534,
+      "grad_norm": 0.5398626327514648,
+      "learning_rate": 0.0001,
+      "loss": 1.3888,
+      "step": 10596
+    },
+    {
+      "epoch": 1.2311356375254139,
+      "grad_norm": 0.5533974766731262,
+      "learning_rate": 0.0001,
+      "loss": 1.3355,
+      "step": 10597
+    },
+    {
+      "epoch": 1.2312518152773744,
+      "grad_norm": 0.5976875424385071,
+      "learning_rate": 0.0001,
+      "loss": 1.4595,
+      "step": 10598
+    },
+    {
+      "epoch": 1.2313679930293349,
+      "grad_norm": 0.5598598122596741,
+      "learning_rate": 0.0001,
+      "loss": 1.3329,
+      "step": 10599
+    },
+    {
+      "epoch": 1.2314841707812954,
+      "grad_norm": 0.5962227582931519,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 10600
+    },
+    {
+      "epoch": 1.2316003485332558,
+      "grad_norm": 0.523517370223999,
+      "learning_rate": 0.0001,
+      "loss": 1.3384,
+      "step": 10601
+    },
+    {
+      "epoch": 1.2317165262852163,
+      "grad_norm": 0.5865228772163391,
+      "learning_rate": 0.0001,
+      "loss": 1.4522,
+      "step": 10602
+    },
+    {
+      "epoch": 1.2318327040371768,
+      "grad_norm": 0.5959430932998657,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 10603
+    },
+    {
+      "epoch": 1.2319488817891373,
+      "grad_norm": 0.5698263645172119,
+      "learning_rate": 0.0001,
+      "loss": 1.3024,
+      "step": 10604
+    },
+    {
+      "epoch": 1.2320650595410978,
+      "grad_norm": 0.5440550446510315,
+      "learning_rate": 0.0001,
+      "loss": 1.3271,
+      "step": 10605
+    },
+    {
+      "epoch": 1.2321812372930583,
+      "grad_norm": 0.5678781270980835,
+      "learning_rate": 0.0001,
+      "loss": 1.468,
+      "step": 10606
+    },
+    {
+      "epoch": 1.2322974150450188,
+      "grad_norm": 0.6232162714004517,
+      "learning_rate": 0.0001,
+      "loss": 1.5875,
+      "step": 10607
+    },
+    {
+      "epoch": 1.2324135927969793,
+      "grad_norm": 0.5675930380821228,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 10608
+    },
+    {
+      "epoch": 1.23252977054894,
+      "grad_norm": 0.5585689544677734,
+      "learning_rate": 0.0001,
+      "loss": 1.379,
+      "step": 10609
+    },
+    {
+      "epoch": 1.2326459483009005,
+      "grad_norm": 0.5632004141807556,
+      "learning_rate": 0.0001,
+      "loss": 1.3636,
+      "step": 10610
+    },
+    {
+      "epoch": 1.232762126052861,
+      "grad_norm": 0.5808938145637512,
+      "learning_rate": 0.0001,
+      "loss": 1.6049,
+      "step": 10611
+    },
+    {
+      "epoch": 1.2328783038048214,
+      "grad_norm": 0.5879512429237366,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 10612
+    },
+    {
+      "epoch": 1.232994481556782,
+      "grad_norm": 0.5673100352287292,
+      "learning_rate": 0.0001,
+      "loss": 1.5315,
+      "step": 10613
+    },
+    {
+      "epoch": 1.2331106593087424,
+      "grad_norm": 0.6437278389930725,
+      "learning_rate": 0.0001,
+      "loss": 1.5295,
+      "step": 10614
+    },
+    {
+      "epoch": 1.233226837060703,
+      "grad_norm": 0.554803192615509,
+      "learning_rate": 0.0001,
+      "loss": 1.5641,
+      "step": 10615
+    },
+    {
+      "epoch": 1.2333430148126634,
+      "grad_norm": 0.5712867379188538,
+      "learning_rate": 0.0001,
+      "loss": 1.3592,
+      "step": 10616
+    },
+    {
+      "epoch": 1.2334591925646239,
+      "grad_norm": 0.5592646598815918,
+      "learning_rate": 0.0001,
+      "loss": 1.4562,
+      "step": 10617
+    },
+    {
+      "epoch": 1.2335753703165844,
+      "grad_norm": 0.5389765501022339,
+      "learning_rate": 0.0001,
+      "loss": 1.3461,
+      "step": 10618
+    },
+    {
+      "epoch": 1.2336915480685449,
+      "grad_norm": 0.622361958026886,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 10619
+    },
+    {
+      "epoch": 1.2338077258205054,
+      "grad_norm": 0.5598737597465515,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 10620
+    },
+    {
+      "epoch": 1.2339239035724658,
+      "grad_norm": 0.5794302821159363,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 10621
+    },
+    {
+      "epoch": 1.2340400813244263,
+      "grad_norm": 0.5385465025901794,
+      "learning_rate": 0.0001,
+      "loss": 1.3243,
+      "step": 10622
+    },
+    {
+      "epoch": 1.2341562590763868,
+      "grad_norm": 0.5621716380119324,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 10623
+    },
+    {
+      "epoch": 1.2342724368283473,
+      "grad_norm": 0.5806053876876831,
+      "learning_rate": 0.0001,
+      "loss": 1.6138,
+      "step": 10624
+    },
+    {
+      "epoch": 1.2343886145803078,
+      "grad_norm": 0.5949336290359497,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 10625
+    },
+    {
+      "epoch": 1.2345047923322683,
+      "grad_norm": 0.6420254707336426,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 10626
+    },
+    {
+      "epoch": 1.2346209700842288,
+      "grad_norm": 0.5900982618331909,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 10627
+    },
+    {
+      "epoch": 1.2347371478361895,
+      "grad_norm": 0.5828454494476318,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 10628
+    },
+    {
+      "epoch": 1.23485332558815,
+      "grad_norm": 0.553706705570221,
+      "learning_rate": 0.0001,
+      "loss": 1.3842,
+      "step": 10629
+    },
+    {
+      "epoch": 1.2349695033401105,
+      "grad_norm": 0.5648255944252014,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 10630
+    },
+    {
+      "epoch": 1.235085681092071,
+      "grad_norm": 0.5571692585945129,
+      "learning_rate": 0.0001,
+      "loss": 1.4299,
+      "step": 10631
+    },
+    {
+      "epoch": 1.2352018588440314,
+      "grad_norm": 0.5914157032966614,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 10632
+    },
+    {
+      "epoch": 1.235318036595992,
+      "grad_norm": 0.5305378437042236,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 10633
+    },
+    {
+      "epoch": 1.2354342143479524,
+      "grad_norm": 0.5221390724182129,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 10634
+    },
+    {
+      "epoch": 1.235550392099913,
+      "grad_norm": 0.5526520013809204,
+      "learning_rate": 0.0001,
+      "loss": 1.5835,
+      "step": 10635
+    },
+    {
+      "epoch": 1.2356665698518734,
+      "grad_norm": 0.552923321723938,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 10636
+    },
+    {
+      "epoch": 1.2357827476038339,
+      "grad_norm": 0.5690024495124817,
+      "learning_rate": 0.0001,
+      "loss": 1.5021,
+      "step": 10637
+    },
+    {
+      "epoch": 1.2358989253557944,
+      "grad_norm": 0.5636914372444153,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 10638
+    },
+    {
+      "epoch": 1.2360151031077549,
+      "grad_norm": 0.6583706736564636,
+      "learning_rate": 0.0001,
+      "loss": 1.7168,
+      "step": 10639
+    },
+    {
+      "epoch": 1.2361312808597154,
+      "grad_norm": 0.600277304649353,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 10640
+    },
+    {
+      "epoch": 1.2362474586116758,
+      "grad_norm": 0.6001637578010559,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 10641
+    },
+    {
+      "epoch": 1.2363636363636363,
+      "grad_norm": 0.5948476195335388,
+      "learning_rate": 0.0001,
+      "loss": 1.5354,
+      "step": 10642
+    },
+    {
+      "epoch": 1.2364798141155968,
+      "grad_norm": 0.6298750638961792,
+      "learning_rate": 0.0001,
+      "loss": 1.4664,
+      "step": 10643
+    },
+    {
+      "epoch": 1.2365959918675573,
+      "grad_norm": 0.578194797039032,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 10644
+    },
+    {
+      "epoch": 1.2367121696195178,
+      "grad_norm": 0.5624285340309143,
+      "learning_rate": 0.0001,
+      "loss": 1.5575,
+      "step": 10645
+    },
+    {
+      "epoch": 1.2368283473714783,
+      "grad_norm": 0.5956478714942932,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 10646
+    },
+    {
+      "epoch": 1.2369445251234388,
+      "grad_norm": 0.5751087665557861,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 10647
+    },
+    {
+      "epoch": 1.2370607028753993,
+      "grad_norm": 0.5814287066459656,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 10648
+    },
+    {
+      "epoch": 1.2371768806273598,
+      "grad_norm": 0.5077279210090637,
+      "learning_rate": 0.0001,
+      "loss": 1.2642,
+      "step": 10649
+    },
+    {
+      "epoch": 1.2372930583793202,
+      "grad_norm": 0.579255223274231,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 10650
+    },
+    {
+      "epoch": 1.237409236131281,
+      "grad_norm": 0.6207264065742493,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 10651
+    },
+    {
+      "epoch": 1.2375254138832414,
+      "grad_norm": 0.6084131598472595,
+      "learning_rate": 0.0001,
+      "loss": 1.4875,
+      "step": 10652
+    },
+    {
+      "epoch": 1.237641591635202,
+      "grad_norm": 0.5942187905311584,
+      "learning_rate": 0.0001,
+      "loss": 1.6195,
+      "step": 10653
+    },
+    {
+      "epoch": 1.2377577693871624,
+      "grad_norm": 0.5793184638023376,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 10654
+    },
+    {
+      "epoch": 1.237873947139123,
+      "grad_norm": 0.5800839066505432,
+      "learning_rate": 0.0001,
+      "loss": 1.4165,
+      "step": 10655
+    },
+    {
+      "epoch": 1.2379901248910834,
+      "grad_norm": 0.5806957483291626,
+      "learning_rate": 0.0001,
+      "loss": 1.4971,
+      "step": 10656
+    },
+    {
+      "epoch": 1.2381063026430439,
+      "grad_norm": 0.5428292751312256,
+      "learning_rate": 0.0001,
+      "loss": 1.2147,
+      "step": 10657
+    },
+    {
+      "epoch": 1.2382224803950044,
+      "grad_norm": 0.5477049350738525,
+      "learning_rate": 0.0001,
+      "loss": 1.3976,
+      "step": 10658
+    },
+    {
+      "epoch": 1.2383386581469649,
+      "grad_norm": 0.5499922037124634,
+      "learning_rate": 0.0001,
+      "loss": 1.3248,
+      "step": 10659
+    },
+    {
+      "epoch": 1.2384548358989254,
+      "grad_norm": 0.5535095930099487,
+      "learning_rate": 0.0001,
+      "loss": 1.312,
+      "step": 10660
+    },
+    {
+      "epoch": 1.2385710136508858,
+      "grad_norm": 0.5951287150382996,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 10661
+    },
+    {
+      "epoch": 1.2386871914028463,
+      "grad_norm": 0.5869619846343994,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 10662
+    },
+    {
+      "epoch": 1.2388033691548068,
+      "grad_norm": 0.5614014267921448,
+      "learning_rate": 0.0001,
+      "loss": 1.4689,
+      "step": 10663
+    },
+    {
+      "epoch": 1.2389195469067673,
+      "grad_norm": 0.5796566605567932,
+      "learning_rate": 0.0001,
+      "loss": 1.4835,
+      "step": 10664
+    },
+    {
+      "epoch": 1.2390357246587278,
+      "grad_norm": 0.5910658240318298,
+      "learning_rate": 0.0001,
+      "loss": 1.5578,
+      "step": 10665
+    },
+    {
+      "epoch": 1.2391519024106883,
+      "grad_norm": 0.564470112323761,
+      "learning_rate": 0.0001,
+      "loss": 1.4322,
+      "step": 10666
+    },
+    {
+      "epoch": 1.2392680801626488,
+      "grad_norm": 0.5840427279472351,
+      "learning_rate": 0.0001,
+      "loss": 1.3944,
+      "step": 10667
+    },
+    {
+      "epoch": 1.2393842579146093,
+      "grad_norm": 0.5382651090621948,
+      "learning_rate": 0.0001,
+      "loss": 1.2162,
+      "step": 10668
+    },
+    {
+      "epoch": 1.2395004356665698,
+      "grad_norm": 0.5443084836006165,
+      "learning_rate": 0.0001,
+      "loss": 1.3367,
+      "step": 10669
+    },
+    {
+      "epoch": 1.2396166134185305,
+      "grad_norm": 0.5765810012817383,
+      "learning_rate": 0.0001,
+      "loss": 1.3894,
+      "step": 10670
+    },
+    {
+      "epoch": 1.239732791170491,
+      "grad_norm": 0.5465205907821655,
+      "learning_rate": 0.0001,
+      "loss": 1.3926,
+      "step": 10671
+    },
+    {
+      "epoch": 1.2398489689224514,
+      "grad_norm": 0.5790244936943054,
+      "learning_rate": 0.0001,
+      "loss": 1.3868,
+      "step": 10672
+    },
+    {
+      "epoch": 1.239965146674412,
+      "grad_norm": 0.5863944888114929,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 10673
+    },
+    {
+      "epoch": 1.2400813244263724,
+      "grad_norm": 0.5954928994178772,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 10674
+    },
+    {
+      "epoch": 1.240197502178333,
+      "grad_norm": 0.6234539747238159,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 10675
+    },
+    {
+      "epoch": 1.2403136799302934,
+      "grad_norm": 0.5883687734603882,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 10676
+    },
+    {
+      "epoch": 1.240429857682254,
+      "grad_norm": 0.5818089842796326,
+      "learning_rate": 0.0001,
+      "loss": 1.4462,
+      "step": 10677
+    },
+    {
+      "epoch": 1.2405460354342144,
+      "grad_norm": 0.6155632138252258,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 10678
+    },
+    {
+      "epoch": 1.2406622131861749,
+      "grad_norm": 0.5792834162712097,
+      "learning_rate": 0.0001,
+      "loss": 1.4936,
+      "step": 10679
+    },
+    {
+      "epoch": 1.2407783909381354,
+      "grad_norm": 0.5829952359199524,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 10680
+    },
+    {
+      "epoch": 1.2408945686900958,
+      "grad_norm": 0.5667127370834351,
+      "learning_rate": 0.0001,
+      "loss": 1.4372,
+      "step": 10681
+    },
+    {
+      "epoch": 1.2410107464420563,
+      "grad_norm": 0.5373808145523071,
+      "learning_rate": 0.0001,
+      "loss": 1.364,
+      "step": 10682
+    },
+    {
+      "epoch": 1.2411269241940168,
+      "grad_norm": 0.5681789517402649,
+      "learning_rate": 0.0001,
+      "loss": 1.3148,
+      "step": 10683
+    },
+    {
+      "epoch": 1.2412431019459773,
+      "grad_norm": 0.5821676254272461,
+      "learning_rate": 0.0001,
+      "loss": 1.4943,
+      "step": 10684
+    },
+    {
+      "epoch": 1.2413592796979378,
+      "grad_norm": 0.583202600479126,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 10685
+    },
+    {
+      "epoch": 1.2414754574498983,
+      "grad_norm": 0.5483458638191223,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 10686
+    },
+    {
+      "epoch": 1.2415916352018588,
+      "grad_norm": 0.5675621032714844,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 10687
+    },
+    {
+      "epoch": 1.2417078129538193,
+      "grad_norm": 0.5624716877937317,
+      "learning_rate": 0.0001,
+      "loss": 1.4133,
+      "step": 10688
+    },
+    {
+      "epoch": 1.2418239907057798,
+      "grad_norm": 0.6353493332862854,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 10689
+    },
+    {
+      "epoch": 1.2419401684577402,
+      "grad_norm": 0.5259159803390503,
+      "learning_rate": 0.0001,
+      "loss": 1.3333,
+      "step": 10690
+    },
+    {
+      "epoch": 1.2420563462097007,
+      "grad_norm": 0.546409547328949,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 10691
+    },
+    {
+      "epoch": 1.2421725239616612,
+      "grad_norm": 0.5584643483161926,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 10692
+    },
+    {
+      "epoch": 1.242288701713622,
+      "grad_norm": 0.5916255116462708,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 10693
+    },
+    {
+      "epoch": 1.2424048794655824,
+      "grad_norm": 0.6383016705513,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 10694
+    },
+    {
+      "epoch": 1.242521057217543,
+      "grad_norm": 0.5840632319450378,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 10695
+    },
+    {
+      "epoch": 1.2426372349695034,
+      "grad_norm": 0.58888179063797,
+      "learning_rate": 0.0001,
+      "loss": 1.5685,
+      "step": 10696
+    },
+    {
+      "epoch": 1.242753412721464,
+      "grad_norm": 0.5815372467041016,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 10697
+    },
+    {
+      "epoch": 1.2428695904734244,
+      "grad_norm": 0.5951624512672424,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 10698
+    },
+    {
+      "epoch": 1.2429857682253849,
+      "grad_norm": 0.611923098564148,
+      "learning_rate": 0.0001,
+      "loss": 1.3961,
+      "step": 10699
+    },
+    {
+      "epoch": 1.2431019459773454,
+      "grad_norm": 0.5687365531921387,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 10700
+    },
+    {
+      "epoch": 1.2432181237293058,
+      "grad_norm": 0.5626431703567505,
+      "learning_rate": 0.0001,
+      "loss": 1.5305,
+      "step": 10701
+    },
+    {
+      "epoch": 1.2433343014812663,
+      "grad_norm": 0.5979207754135132,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 10702
+    },
+    {
+      "epoch": 1.2434504792332268,
+      "grad_norm": 0.5764636993408203,
+      "learning_rate": 0.0001,
+      "loss": 1.6826,
+      "step": 10703
+    },
+    {
+      "epoch": 1.2435666569851873,
+      "grad_norm": 0.6030092835426331,
+      "learning_rate": 0.0001,
+      "loss": 1.653,
+      "step": 10704
+    },
+    {
+      "epoch": 1.2436828347371478,
+      "grad_norm": 0.5411743521690369,
+      "learning_rate": 0.0001,
+      "loss": 1.3816,
+      "step": 10705
+    },
+    {
+      "epoch": 1.2437990124891083,
+      "grad_norm": 0.5814303755760193,
+      "learning_rate": 0.0001,
+      "loss": 1.6629,
+      "step": 10706
+    },
+    {
+      "epoch": 1.2439151902410688,
+      "grad_norm": 0.5599769949913025,
+      "learning_rate": 0.0001,
+      "loss": 1.3897,
+      "step": 10707
+    },
+    {
+      "epoch": 1.2440313679930293,
+      "grad_norm": 0.5339173674583435,
+      "learning_rate": 0.0001,
+      "loss": 1.4417,
+      "step": 10708
+    },
+    {
+      "epoch": 1.2441475457449898,
+      "grad_norm": 0.5466998815536499,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 10709
+    },
+    {
+      "epoch": 1.2442637234969502,
+      "grad_norm": 0.6054586172103882,
+      "learning_rate": 0.0001,
+      "loss": 1.6259,
+      "step": 10710
+    },
+    {
+      "epoch": 1.244379901248911,
+      "grad_norm": 0.5586958527565002,
+      "learning_rate": 0.0001,
+      "loss": 1.4248,
+      "step": 10711
+    },
+    {
+      "epoch": 1.2444960790008714,
+      "grad_norm": 0.5811334252357483,
+      "learning_rate": 0.0001,
+      "loss": 1.4595,
+      "step": 10712
+    },
+    {
+      "epoch": 1.244612256752832,
+      "grad_norm": 0.5410207509994507,
+      "learning_rate": 0.0001,
+      "loss": 1.418,
+      "step": 10713
+    },
+    {
+      "epoch": 1.2447284345047924,
+      "grad_norm": 0.5457951426506042,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 10714
+    },
+    {
+      "epoch": 1.244844612256753,
+      "grad_norm": 0.6239075660705566,
+      "learning_rate": 0.0001,
+      "loss": 1.5306,
+      "step": 10715
+    },
+    {
+      "epoch": 1.2449607900087134,
+      "grad_norm": 0.6004078984260559,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 10716
+    },
+    {
+      "epoch": 1.245076967760674,
+      "grad_norm": 0.5737530589103699,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 10717
+    },
+    {
+      "epoch": 1.2451931455126344,
+      "grad_norm": 0.5727450847625732,
+      "learning_rate": 0.0001,
+      "loss": 1.6573,
+      "step": 10718
+    },
+    {
+      "epoch": 1.2453093232645949,
+      "grad_norm": 0.5446699261665344,
+      "learning_rate": 0.0001,
+      "loss": 1.2936,
+      "step": 10719
+    },
+    {
+      "epoch": 1.2454255010165554,
+      "grad_norm": 0.5467056035995483,
+      "learning_rate": 0.0001,
+      "loss": 1.3091,
+      "step": 10720
+    },
+    {
+      "epoch": 1.2455416787685158,
+      "grad_norm": 0.5639151334762573,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 10721
+    },
+    {
+      "epoch": 1.2456578565204763,
+      "grad_norm": 0.6179391741752625,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 10722
+    },
+    {
+      "epoch": 1.2457740342724368,
+      "grad_norm": 0.5786750912666321,
+      "learning_rate": 0.0001,
+      "loss": 1.409,
+      "step": 10723
+    },
+    {
+      "epoch": 1.2458902120243973,
+      "grad_norm": 0.5844369530677795,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 10724
+    },
+    {
+      "epoch": 1.2460063897763578,
+      "grad_norm": 0.5636518597602844,
+      "learning_rate": 0.0001,
+      "loss": 1.3696,
+      "step": 10725
+    },
+    {
+      "epoch": 1.2461225675283183,
+      "grad_norm": 0.5851801633834839,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 10726
+    },
+    {
+      "epoch": 1.2462387452802788,
+      "grad_norm": 0.5837157368659973,
+      "learning_rate": 0.0001,
+      "loss": 1.2587,
+      "step": 10727
+    },
+    {
+      "epoch": 1.2463549230322393,
+      "grad_norm": 0.5497305393218994,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 10728
+    },
+    {
+      "epoch": 1.2464711007841998,
+      "grad_norm": 0.6165003180503845,
+      "learning_rate": 0.0001,
+      "loss": 1.637,
+      "step": 10729
+    },
+    {
+      "epoch": 1.2465872785361602,
+      "grad_norm": 0.595329999923706,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 10730
+    },
+    {
+      "epoch": 1.2467034562881207,
+      "grad_norm": 0.5779715180397034,
+      "learning_rate": 0.0001,
+      "loss": 1.5749,
+      "step": 10731
+    },
+    {
+      "epoch": 1.2468196340400812,
+      "grad_norm": 0.6294297575950623,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 10732
+    },
+    {
+      "epoch": 1.2469358117920417,
+      "grad_norm": 0.5893142223358154,
+      "learning_rate": 0.0001,
+      "loss": 1.4439,
+      "step": 10733
+    },
+    {
+      "epoch": 1.2470519895440022,
+      "grad_norm": 0.5959643721580505,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 10734
+    },
+    {
+      "epoch": 1.247168167295963,
+      "grad_norm": 0.6068616509437561,
+      "learning_rate": 0.0001,
+      "loss": 1.4337,
+      "step": 10735
+    },
+    {
+      "epoch": 1.2472843450479234,
+      "grad_norm": 0.5508027076721191,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 10736
+    },
+    {
+      "epoch": 1.247400522799884,
+      "grad_norm": 0.5569624900817871,
+      "learning_rate": 0.0001,
+      "loss": 1.4259,
+      "step": 10737
+    },
+    {
+      "epoch": 1.2475167005518444,
+      "grad_norm": 0.5653142333030701,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 10738
+    },
+    {
+      "epoch": 1.2476328783038049,
+      "grad_norm": 0.6183950304985046,
+      "learning_rate": 0.0001,
+      "loss": 1.5533,
+      "step": 10739
+    },
+    {
+      "epoch": 1.2477490560557654,
+      "grad_norm": 0.5955910086631775,
+      "learning_rate": 0.0001,
+      "loss": 1.3331,
+      "step": 10740
+    },
+    {
+      "epoch": 1.2478652338077258,
+      "grad_norm": 0.5446209907531738,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 10741
+    },
+    {
+      "epoch": 1.2479814115596863,
+      "grad_norm": 0.5655883550643921,
+      "learning_rate": 0.0001,
+      "loss": 1.4624,
+      "step": 10742
+    },
+    {
+      "epoch": 1.2480975893116468,
+      "grad_norm": 0.5753975510597229,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 10743
+    },
+    {
+      "epoch": 1.2482137670636073,
+      "grad_norm": 0.5425572991371155,
+      "learning_rate": 0.0001,
+      "loss": 1.4076,
+      "step": 10744
+    },
+    {
+      "epoch": 1.2483299448155678,
+      "grad_norm": 0.6025961637496948,
+      "learning_rate": 0.0001,
+      "loss": 1.5467,
+      "step": 10745
+    },
+    {
+      "epoch": 1.2484461225675283,
+      "grad_norm": 0.5763283967971802,
+      "learning_rate": 0.0001,
+      "loss": 1.5884,
+      "step": 10746
+    },
+    {
+      "epoch": 1.2485623003194888,
+      "grad_norm": 0.5742851495742798,
+      "learning_rate": 0.0001,
+      "loss": 1.501,
+      "step": 10747
+    },
+    {
+      "epoch": 1.2486784780714493,
+      "grad_norm": 0.5632321834564209,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 10748
+    },
+    {
+      "epoch": 1.2487946558234098,
+      "grad_norm": 0.5582688450813293,
+      "learning_rate": 0.0001,
+      "loss": 1.4634,
+      "step": 10749
+    },
+    {
+      "epoch": 1.2489108335753702,
+      "grad_norm": 0.5706313848495483,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 10750
+    },
+    {
+      "epoch": 1.2490270113273307,
+      "grad_norm": 0.5375720262527466,
+      "learning_rate": 0.0001,
+      "loss": 1.2735,
+      "step": 10751
+    },
+    {
+      "epoch": 1.2491431890792912,
+      "grad_norm": 0.5515146851539612,
+      "learning_rate": 0.0001,
+      "loss": 1.4086,
+      "step": 10752
+    },
+    {
+      "epoch": 1.249259366831252,
+      "grad_norm": 0.5534106492996216,
+      "learning_rate": 0.0001,
+      "loss": 1.284,
+      "step": 10753
+    },
+    {
+      "epoch": 1.2493755445832124,
+      "grad_norm": 0.5403925776481628,
+      "learning_rate": 0.0001,
+      "loss": 1.2556,
+      "step": 10754
+    },
+    {
+      "epoch": 1.249491722335173,
+      "grad_norm": 0.6206772923469543,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 10755
+    },
+    {
+      "epoch": 1.2496079000871334,
+      "grad_norm": 0.6456174850463867,
+      "learning_rate": 0.0001,
+      "loss": 1.467,
+      "step": 10756
+    },
+    {
+      "epoch": 1.249724077839094,
+      "grad_norm": 0.5558960437774658,
+      "learning_rate": 0.0001,
+      "loss": 1.3112,
+      "step": 10757
+    },
+    {
+      "epoch": 1.2498402555910544,
+      "grad_norm": 0.5899320244789124,
+      "learning_rate": 0.0001,
+      "loss": 1.4726,
+      "step": 10758
+    },
+    {
+      "epoch": 1.2499564333430149,
+      "grad_norm": 0.5711237192153931,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 10759
+    },
+    {
+      "epoch": 1.2500726110949754,
+      "grad_norm": 0.5839224457740784,
+      "learning_rate": 0.0001,
+      "loss": 1.607,
+      "step": 10760
+    },
+    {
+      "epoch": 1.2501887888469359,
+      "grad_norm": 0.5762811899185181,
+      "learning_rate": 0.0001,
+      "loss": 1.4382,
+      "step": 10761
+    },
+    {
+      "epoch": 1.2503049665988963,
+      "grad_norm": 0.5910693407058716,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 10762
+    },
+    {
+      "epoch": 1.2504211443508568,
+      "grad_norm": 0.5522376298904419,
+      "learning_rate": 0.0001,
+      "loss": 1.1839,
+      "step": 10763
+    },
+    {
+      "epoch": 1.2505373221028173,
+      "grad_norm": 0.6215783357620239,
+      "learning_rate": 0.0001,
+      "loss": 1.5552,
+      "step": 10764
+    },
+    {
+      "epoch": 1.2506534998547778,
+      "grad_norm": 0.5663979649543762,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 10765
+    },
+    {
+      "epoch": 1.2507696776067383,
+      "grad_norm": 0.5686525702476501,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 10766
+    },
+    {
+      "epoch": 1.2508858553586988,
+      "grad_norm": 0.5687610507011414,
+      "learning_rate": 0.0001,
+      "loss": 1.3635,
+      "step": 10767
+    },
+    {
+      "epoch": 1.2510020331106593,
+      "grad_norm": 0.5779849290847778,
+      "learning_rate": 0.0001,
+      "loss": 1.6153,
+      "step": 10768
+    },
+    {
+      "epoch": 1.2511182108626198,
+      "grad_norm": 0.5521991848945618,
+      "learning_rate": 0.0001,
+      "loss": 1.4001,
+      "step": 10769
+    },
+    {
+      "epoch": 1.2512343886145803,
+      "grad_norm": 0.5512406229972839,
+      "learning_rate": 0.0001,
+      "loss": 1.3389,
+      "step": 10770
+    },
+    {
+      "epoch": 1.2513505663665407,
+      "grad_norm": 0.5759303569793701,
+      "learning_rate": 0.0001,
+      "loss": 1.5774,
+      "step": 10771
+    },
+    {
+      "epoch": 1.2514667441185012,
+      "grad_norm": 0.5932089686393738,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 10772
+    },
+    {
+      "epoch": 1.2515829218704617,
+      "grad_norm": 0.5830219388008118,
+      "learning_rate": 0.0001,
+      "loss": 1.3805,
+      "step": 10773
+    },
+    {
+      "epoch": 1.2516990996224222,
+      "grad_norm": 0.6111432909965515,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 10774
+    },
+    {
+      "epoch": 1.2518152773743827,
+      "grad_norm": 0.554973840713501,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 10775
+    },
+    {
+      "epoch": 1.2519314551263432,
+      "grad_norm": 0.5329660177230835,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 10776
+    },
+    {
+      "epoch": 1.2520476328783037,
+      "grad_norm": 0.5556033849716187,
+      "learning_rate": 0.0001,
+      "loss": 1.3319,
+      "step": 10777
+    },
+    {
+      "epoch": 1.2521638106302644,
+      "grad_norm": 0.5581437349319458,
+      "learning_rate": 0.0001,
+      "loss": 1.4436,
+      "step": 10778
+    },
+    {
+      "epoch": 1.2522799883822249,
+      "grad_norm": 0.5687176585197449,
+      "learning_rate": 0.0001,
+      "loss": 1.4248,
+      "step": 10779
+    },
+    {
+      "epoch": 1.2523961661341854,
+      "grad_norm": 0.6239218711853027,
+      "learning_rate": 0.0001,
+      "loss": 1.5807,
+      "step": 10780
+    },
+    {
+      "epoch": 1.2525123438861459,
+      "grad_norm": 0.542938768863678,
+      "learning_rate": 0.0001,
+      "loss": 1.3693,
+      "step": 10781
+    },
+    {
+      "epoch": 1.2526285216381063,
+      "grad_norm": 0.5767273902893066,
+      "learning_rate": 0.0001,
+      "loss": 1.5495,
+      "step": 10782
+    },
+    {
+      "epoch": 1.2527446993900668,
+      "grad_norm": 0.5635040998458862,
+      "learning_rate": 0.0001,
+      "loss": 1.6863,
+      "step": 10783
+    },
+    {
+      "epoch": 1.2528608771420273,
+      "grad_norm": 0.5268775820732117,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 10784
+    },
+    {
+      "epoch": 1.2529770548939878,
+      "grad_norm": 0.5350583791732788,
+      "learning_rate": 0.0001,
+      "loss": 1.4114,
+      "step": 10785
+    },
+    {
+      "epoch": 1.2530932326459483,
+      "grad_norm": 0.5941740870475769,
+      "learning_rate": 0.0001,
+      "loss": 1.5906,
+      "step": 10786
+    },
+    {
+      "epoch": 1.2532094103979088,
+      "grad_norm": 0.5417436361312866,
+      "learning_rate": 0.0001,
+      "loss": 1.2903,
+      "step": 10787
+    },
+    {
+      "epoch": 1.2533255881498693,
+      "grad_norm": 0.5705372095108032,
+      "learning_rate": 0.0001,
+      "loss": 1.4168,
+      "step": 10788
+    },
+    {
+      "epoch": 1.2534417659018298,
+      "grad_norm": 0.5509761571884155,
+      "learning_rate": 0.0001,
+      "loss": 1.4277,
+      "step": 10789
+    },
+    {
+      "epoch": 1.2535579436537903,
+      "grad_norm": 0.6663389205932617,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 10790
+    },
+    {
+      "epoch": 1.2536741214057507,
+      "grad_norm": 0.6134310960769653,
+      "learning_rate": 0.0001,
+      "loss": 1.4366,
+      "step": 10791
+    },
+    {
+      "epoch": 1.2537902991577112,
+      "grad_norm": 0.5742259621620178,
+      "learning_rate": 0.0001,
+      "loss": 1.4703,
+      "step": 10792
+    },
+    {
+      "epoch": 1.2539064769096717,
+      "grad_norm": 0.5614108443260193,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 10793
+    },
+    {
+      "epoch": 1.2540226546616324,
+      "grad_norm": 0.5708277821540833,
+      "learning_rate": 0.0001,
+      "loss": 1.4222,
+      "step": 10794
+    },
+    {
+      "epoch": 1.254138832413593,
+      "grad_norm": 0.5575094819068909,
+      "learning_rate": 0.0001,
+      "loss": 1.3619,
+      "step": 10795
+    },
+    {
+      "epoch": 1.2542550101655534,
+      "grad_norm": 0.6430823802947998,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 10796
+    },
+    {
+      "epoch": 1.254371187917514,
+      "grad_norm": 0.6255773901939392,
+      "learning_rate": 0.0001,
+      "loss": 1.5583,
+      "step": 10797
+    },
+    {
+      "epoch": 1.2544873656694744,
+      "grad_norm": 0.5493278503417969,
+      "learning_rate": 0.0001,
+      "loss": 1.4425,
+      "step": 10798
+    },
+    {
+      "epoch": 1.2546035434214349,
+      "grad_norm": 0.5868274569511414,
+      "learning_rate": 0.0001,
+      "loss": 1.4117,
+      "step": 10799
+    },
+    {
+      "epoch": 1.2547197211733954,
+      "grad_norm": 0.5614247918128967,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 10800
+    },
+    {
+      "epoch": 1.2548358989253559,
+      "grad_norm": 0.6078110933303833,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 10801
+    },
+    {
+      "epoch": 1.2549520766773163,
+      "grad_norm": 0.5670056939125061,
+      "learning_rate": 0.0001,
+      "loss": 1.3722,
+      "step": 10802
+    },
+    {
+      "epoch": 1.2550682544292768,
+      "grad_norm": 0.5822281837463379,
+      "learning_rate": 0.0001,
+      "loss": 1.3983,
+      "step": 10803
+    },
+    {
+      "epoch": 1.2551844321812373,
+      "grad_norm": 0.5852466821670532,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 10804
+    },
+    {
+      "epoch": 1.2553006099331978,
+      "grad_norm": 0.5783720016479492,
+      "learning_rate": 0.0001,
+      "loss": 1.2861,
+      "step": 10805
+    },
+    {
+      "epoch": 1.2554167876851583,
+      "grad_norm": 0.5860381722450256,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 10806
+    },
+    {
+      "epoch": 1.2555329654371188,
+      "grad_norm": 0.5932544469833374,
+      "learning_rate": 0.0001,
+      "loss": 1.6662,
+      "step": 10807
+    },
+    {
+      "epoch": 1.2556491431890793,
+      "grad_norm": 0.6420629024505615,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 10808
+    },
+    {
+      "epoch": 1.2557653209410398,
+      "grad_norm": 0.6060267090797424,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 10809
+    },
+    {
+      "epoch": 1.2558814986930003,
+      "grad_norm": 0.5844810605049133,
+      "learning_rate": 0.0001,
+      "loss": 1.4504,
+      "step": 10810
+    },
+    {
+      "epoch": 1.2559976764449607,
+      "grad_norm": 0.6256417632102966,
+      "learning_rate": 0.0001,
+      "loss": 1.5777,
+      "step": 10811
+    },
+    {
+      "epoch": 1.2561138541969212,
+      "grad_norm": 0.5611162185668945,
+      "learning_rate": 0.0001,
+      "loss": 1.6017,
+      "step": 10812
+    },
+    {
+      "epoch": 1.2562300319488817,
+      "grad_norm": 0.5997201204299927,
+      "learning_rate": 0.0001,
+      "loss": 1.682,
+      "step": 10813
+    },
+    {
+      "epoch": 1.2563462097008422,
+      "grad_norm": 0.580735981464386,
+      "learning_rate": 0.0001,
+      "loss": 1.5223,
+      "step": 10814
+    },
+    {
+      "epoch": 1.2564623874528027,
+      "grad_norm": 0.5708926916122437,
+      "learning_rate": 0.0001,
+      "loss": 1.3772,
+      "step": 10815
+    },
+    {
+      "epoch": 1.2565785652047632,
+      "grad_norm": 0.5596855282783508,
+      "learning_rate": 0.0001,
+      "loss": 1.3001,
+      "step": 10816
+    },
+    {
+      "epoch": 1.2566947429567237,
+      "grad_norm": 0.6057003736495972,
+      "learning_rate": 0.0001,
+      "loss": 1.4523,
+      "step": 10817
+    },
+    {
+      "epoch": 1.2568109207086842,
+      "grad_norm": 0.6439718008041382,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 10818
+    },
+    {
+      "epoch": 1.2569270984606447,
+      "grad_norm": 0.5926083326339722,
+      "learning_rate": 0.0001,
+      "loss": 1.6552,
+      "step": 10819
+    },
+    {
+      "epoch": 1.2570432762126054,
+      "grad_norm": 0.5531743764877319,
+      "learning_rate": 0.0001,
+      "loss": 1.5071,
+      "step": 10820
+    },
+    {
+      "epoch": 1.2571594539645659,
+      "grad_norm": 0.5597687363624573,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 10821
+    },
+    {
+      "epoch": 1.2572756317165263,
+      "grad_norm": 0.5524884462356567,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 10822
+    },
+    {
+      "epoch": 1.2573918094684868,
+      "grad_norm": 0.6109902262687683,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 10823
+    },
+    {
+      "epoch": 1.2575079872204473,
+      "grad_norm": 0.5356048345565796,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 10824
+    },
+    {
+      "epoch": 1.2576241649724078,
+      "grad_norm": 0.5408940315246582,
+      "learning_rate": 0.0001,
+      "loss": 1.2335,
+      "step": 10825
+    },
+    {
+      "epoch": 1.2577403427243683,
+      "grad_norm": 0.5846429467201233,
+      "learning_rate": 0.0001,
+      "loss": 1.5364,
+      "step": 10826
+    },
+    {
+      "epoch": 1.2578565204763288,
+      "grad_norm": 0.5699112415313721,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 10827
+    },
+    {
+      "epoch": 1.2579726982282893,
+      "grad_norm": 0.6659561395645142,
+      "learning_rate": 0.0001,
+      "loss": 1.6991,
+      "step": 10828
+    },
+    {
+      "epoch": 1.2580888759802498,
+      "grad_norm": 0.6590319871902466,
+      "learning_rate": 0.0001,
+      "loss": 1.6286,
+      "step": 10829
+    },
+    {
+      "epoch": 1.2582050537322103,
+      "grad_norm": 0.633607804775238,
+      "learning_rate": 0.0001,
+      "loss": 1.5275,
+      "step": 10830
+    },
+    {
+      "epoch": 1.2583212314841707,
+      "grad_norm": 0.5824370980262756,
+      "learning_rate": 0.0001,
+      "loss": 1.4908,
+      "step": 10831
+    },
+    {
+      "epoch": 1.2584374092361312,
+      "grad_norm": 0.6036174893379211,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 10832
+    },
+    {
+      "epoch": 1.2585535869880917,
+      "grad_norm": 0.5827155709266663,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 10833
+    },
+    {
+      "epoch": 1.2586697647400522,
+      "grad_norm": 0.579399585723877,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 10834
+    },
+    {
+      "epoch": 1.2587859424920127,
+      "grad_norm": 0.5795319080352783,
+      "learning_rate": 0.0001,
+      "loss": 1.5249,
+      "step": 10835
+    },
+    {
+      "epoch": 1.2589021202439734,
+      "grad_norm": 0.5639935731887817,
+      "learning_rate": 0.0001,
+      "loss": 1.3972,
+      "step": 10836
+    },
+    {
+      "epoch": 1.259018297995934,
+      "grad_norm": 0.5532496571540833,
+      "learning_rate": 0.0001,
+      "loss": 1.465,
+      "step": 10837
+    },
+    {
+      "epoch": 1.2591344757478944,
+      "grad_norm": 0.5931538343429565,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 10838
+    },
+    {
+      "epoch": 1.2592506534998549,
+      "grad_norm": 0.6140592098236084,
+      "learning_rate": 0.0001,
+      "loss": 1.6095,
+      "step": 10839
+    },
+    {
+      "epoch": 1.2593668312518154,
+      "grad_norm": 0.5485202074050903,
+      "learning_rate": 0.0001,
+      "loss": 1.4436,
+      "step": 10840
+    },
+    {
+      "epoch": 1.2594830090037759,
+      "grad_norm": 0.5648629069328308,
+      "learning_rate": 0.0001,
+      "loss": 1.3829,
+      "step": 10841
+    },
+    {
+      "epoch": 1.2595991867557363,
+      "grad_norm": 0.5815029740333557,
+      "learning_rate": 0.0001,
+      "loss": 1.5187,
+      "step": 10842
+    },
+    {
+      "epoch": 1.2597153645076968,
+      "grad_norm": 0.5790690183639526,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 10843
+    },
+    {
+      "epoch": 1.2598315422596573,
+      "grad_norm": 0.6052435636520386,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 10844
+    },
+    {
+      "epoch": 1.2599477200116178,
+      "grad_norm": 0.5813129544258118,
+      "learning_rate": 0.0001,
+      "loss": 1.5912,
+      "step": 10845
+    },
+    {
+      "epoch": 1.2600638977635783,
+      "grad_norm": 0.5918418169021606,
+      "learning_rate": 0.0001,
+      "loss": 1.7113,
+      "step": 10846
+    },
+    {
+      "epoch": 1.2601800755155388,
+      "grad_norm": 0.5554955005645752,
+      "learning_rate": 0.0001,
+      "loss": 1.5289,
+      "step": 10847
+    },
+    {
+      "epoch": 1.2602962532674993,
+      "grad_norm": 0.540164589881897,
+      "learning_rate": 0.0001,
+      "loss": 1.3184,
+      "step": 10848
+    },
+    {
+      "epoch": 1.2604124310194598,
+      "grad_norm": 0.6040856242179871,
+      "learning_rate": 0.0001,
+      "loss": 1.4184,
+      "step": 10849
+    },
+    {
+      "epoch": 1.2605286087714203,
+      "grad_norm": 0.5804588794708252,
+      "learning_rate": 0.0001,
+      "loss": 1.4578,
+      "step": 10850
+    },
+    {
+      "epoch": 1.2606447865233807,
+      "grad_norm": 0.5707676410675049,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 10851
+    },
+    {
+      "epoch": 1.2607609642753412,
+      "grad_norm": 0.647530734539032,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 10852
+    },
+    {
+      "epoch": 1.2608771420273017,
+      "grad_norm": 0.5723570585250854,
+      "learning_rate": 0.0001,
+      "loss": 1.3976,
+      "step": 10853
+    },
+    {
+      "epoch": 1.2609933197792622,
+      "grad_norm": 0.5460187196731567,
+      "learning_rate": 0.0001,
+      "loss": 1.3676,
+      "step": 10854
+    },
+    {
+      "epoch": 1.2611094975312227,
+      "grad_norm": 0.5793501138687134,
+      "learning_rate": 0.0001,
+      "loss": 1.4138,
+      "step": 10855
+    },
+    {
+      "epoch": 1.2612256752831832,
+      "grad_norm": 0.6033415198326111,
+      "learning_rate": 0.0001,
+      "loss": 1.5383,
+      "step": 10856
+    },
+    {
+      "epoch": 1.2613418530351437,
+      "grad_norm": 0.5997538566589355,
+      "learning_rate": 0.0001,
+      "loss": 1.2872,
+      "step": 10857
+    },
+    {
+      "epoch": 1.2614580307871042,
+      "grad_norm": 0.5751636624336243,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 10858
+    },
+    {
+      "epoch": 1.2615742085390647,
+      "grad_norm": 0.6361991167068481,
+      "learning_rate": 0.0001,
+      "loss": 1.4256,
+      "step": 10859
+    },
+    {
+      "epoch": 1.2616903862910251,
+      "grad_norm": 0.5734949707984924,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 10860
+    },
+    {
+      "epoch": 1.2618065640429856,
+      "grad_norm": 0.6365008354187012,
+      "learning_rate": 0.0001,
+      "loss": 1.6581,
+      "step": 10861
+    },
+    {
+      "epoch": 1.2619227417949463,
+      "grad_norm": 0.6318557858467102,
+      "learning_rate": 0.0001,
+      "loss": 1.715,
+      "step": 10862
+    },
+    {
+      "epoch": 1.2620389195469068,
+      "grad_norm": 0.6001318693161011,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 10863
+    },
+    {
+      "epoch": 1.2621550972988673,
+      "grad_norm": 0.6317021250724792,
+      "learning_rate": 0.0001,
+      "loss": 1.6188,
+      "step": 10864
+    },
+    {
+      "epoch": 1.2622712750508278,
+      "grad_norm": 0.5643987059593201,
+      "learning_rate": 0.0001,
+      "loss": 1.4726,
+      "step": 10865
+    },
+    {
+      "epoch": 1.2623874528027883,
+      "grad_norm": 0.5744731426239014,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 10866
+    },
+    {
+      "epoch": 1.2625036305547488,
+      "grad_norm": 0.605496883392334,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 10867
+    },
+    {
+      "epoch": 1.2626198083067093,
+      "grad_norm": 0.5679904222488403,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 10868
+    },
+    {
+      "epoch": 1.2627359860586698,
+      "grad_norm": 0.5534711480140686,
+      "learning_rate": 0.0001,
+      "loss": 1.4292,
+      "step": 10869
+    },
+    {
+      "epoch": 1.2628521638106303,
+      "grad_norm": 0.5767245292663574,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 10870
+    },
+    {
+      "epoch": 1.2629683415625907,
+      "grad_norm": 0.5895627737045288,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 10871
+    },
+    {
+      "epoch": 1.2630845193145512,
+      "grad_norm": 0.5530607104301453,
+      "learning_rate": 0.0001,
+      "loss": 1.3382,
+      "step": 10872
+    },
+    {
+      "epoch": 1.2632006970665117,
+      "grad_norm": 0.5866522789001465,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 10873
+    },
+    {
+      "epoch": 1.2633168748184722,
+      "grad_norm": 0.5559065937995911,
+      "learning_rate": 0.0001,
+      "loss": 1.3051,
+      "step": 10874
+    },
+    {
+      "epoch": 1.2634330525704327,
+      "grad_norm": 0.5616282820701599,
+      "learning_rate": 0.0001,
+      "loss": 1.2669,
+      "step": 10875
+    },
+    {
+      "epoch": 1.2635492303223932,
+      "grad_norm": 0.5756585597991943,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 10876
+    },
+    {
+      "epoch": 1.2636654080743537,
+      "grad_norm": 0.6078899502754211,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 10877
+    },
+    {
+      "epoch": 1.2637815858263144,
+      "grad_norm": 0.5421556830406189,
+      "learning_rate": 0.0001,
+      "loss": 1.3957,
+      "step": 10878
+    },
+    {
+      "epoch": 1.2638977635782749,
+      "grad_norm": 0.6021136045455933,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 10879
+    },
+    {
+      "epoch": 1.2640139413302354,
+      "grad_norm": 0.6049302220344543,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 10880
+    },
+    {
+      "epoch": 1.2641301190821959,
+      "grad_norm": 0.5303308963775635,
+      "learning_rate": 0.0001,
+      "loss": 1.2202,
+      "step": 10881
+    },
+    {
+      "epoch": 1.2642462968341563,
+      "grad_norm": 0.5799589157104492,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 10882
+    },
+    {
+      "epoch": 1.2643624745861168,
+      "grad_norm": 0.5389856696128845,
+      "learning_rate": 0.0001,
+      "loss": 1.2409,
+      "step": 10883
+    },
+    {
+      "epoch": 1.2644786523380773,
+      "grad_norm": 0.5793135166168213,
+      "learning_rate": 0.0001,
+      "loss": 1.3456,
+      "step": 10884
+    },
+    {
+      "epoch": 1.2645948300900378,
+      "grad_norm": 0.665398120880127,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 10885
+    },
+    {
+      "epoch": 1.2647110078419983,
+      "grad_norm": 0.6195827126502991,
+      "learning_rate": 0.0001,
+      "loss": 1.3291,
+      "step": 10886
+    },
+    {
+      "epoch": 1.2648271855939588,
+      "grad_norm": 0.603444516658783,
+      "learning_rate": 0.0001,
+      "loss": 1.4637,
+      "step": 10887
+    },
+    {
+      "epoch": 1.2649433633459193,
+      "grad_norm": 0.6044740080833435,
+      "learning_rate": 0.0001,
+      "loss": 1.2983,
+      "step": 10888
+    },
+    {
+      "epoch": 1.2650595410978798,
+      "grad_norm": 0.5764860510826111,
+      "learning_rate": 0.0001,
+      "loss": 1.4351,
+      "step": 10889
+    },
+    {
+      "epoch": 1.2651757188498403,
+      "grad_norm": 0.5791444182395935,
+      "learning_rate": 0.0001,
+      "loss": 1.3424,
+      "step": 10890
+    },
+    {
+      "epoch": 1.2652918966018007,
+      "grad_norm": 0.6098731160163879,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 10891
+    },
+    {
+      "epoch": 1.2654080743537612,
+      "grad_norm": 0.5369520783424377,
+      "learning_rate": 0.0001,
+      "loss": 1.3948,
+      "step": 10892
+    },
+    {
+      "epoch": 1.2655242521057217,
+      "grad_norm": 0.6197044849395752,
+      "learning_rate": 0.0001,
+      "loss": 1.368,
+      "step": 10893
+    },
+    {
+      "epoch": 1.2656404298576822,
+      "grad_norm": 0.588814377784729,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 10894
+    },
+    {
+      "epoch": 1.2657566076096427,
+      "grad_norm": 0.5804957747459412,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 10895
+    },
+    {
+      "epoch": 1.2658727853616032,
+      "grad_norm": 0.6161913275718689,
+      "learning_rate": 0.0001,
+      "loss": 1.641,
+      "step": 10896
+    },
+    {
+      "epoch": 1.2659889631135637,
+      "grad_norm": 0.6082305312156677,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 10897
+    },
+    {
+      "epoch": 1.2661051408655242,
+      "grad_norm": 0.5396662950515747,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 10898
+    },
+    {
+      "epoch": 1.2662213186174847,
+      "grad_norm": 0.5761926770210266,
+      "learning_rate": 0.0001,
+      "loss": 1.4168,
+      "step": 10899
+    },
+    {
+      "epoch": 1.2663374963694451,
+      "grad_norm": 0.5800683498382568,
+      "learning_rate": 0.0001,
+      "loss": 1.4611,
+      "step": 10900
+    },
+    {
+      "epoch": 1.2664536741214056,
+      "grad_norm": 0.5463969111442566,
+      "learning_rate": 0.0001,
+      "loss": 1.4545,
+      "step": 10901
+    },
+    {
+      "epoch": 1.2665698518733661,
+      "grad_norm": 0.5822511315345764,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 10902
+    },
+    {
+      "epoch": 1.2666860296253266,
+      "grad_norm": 0.6074179410934448,
+      "learning_rate": 0.0001,
+      "loss": 1.7182,
+      "step": 10903
+    },
+    {
+      "epoch": 1.2668022073772873,
+      "grad_norm": 0.6279671788215637,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 10904
+    },
+    {
+      "epoch": 1.2669183851292478,
+      "grad_norm": 0.6327800750732422,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 10905
+    },
+    {
+      "epoch": 1.2670345628812083,
+      "grad_norm": 0.6204458475112915,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 10906
+    },
+    {
+      "epoch": 1.2671507406331688,
+      "grad_norm": 0.5944911241531372,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 10907
+    },
+    {
+      "epoch": 1.2672669183851293,
+      "grad_norm": 0.5533126592636108,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 10908
+    },
+    {
+      "epoch": 1.2673830961370898,
+      "grad_norm": 0.5814772844314575,
+      "learning_rate": 0.0001,
+      "loss": 1.5974,
+      "step": 10909
+    },
+    {
+      "epoch": 1.2674992738890503,
+      "grad_norm": 0.6086983680725098,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 10910
+    },
+    {
+      "epoch": 1.2676154516410107,
+      "grad_norm": 0.6263343691825867,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 10911
+    },
+    {
+      "epoch": 1.2677316293929712,
+      "grad_norm": 0.5537508130073547,
+      "learning_rate": 0.0001,
+      "loss": 1.3299,
+      "step": 10912
+    },
+    {
+      "epoch": 1.2678478071449317,
+      "grad_norm": 0.5584046840667725,
+      "learning_rate": 0.0001,
+      "loss": 1.3079,
+      "step": 10913
+    },
+    {
+      "epoch": 1.2679639848968922,
+      "grad_norm": 0.576492428779602,
+      "learning_rate": 0.0001,
+      "loss": 1.4632,
+      "step": 10914
+    },
+    {
+      "epoch": 1.2680801626488527,
+      "grad_norm": 0.5651898384094238,
+      "learning_rate": 0.0001,
+      "loss": 1.2658,
+      "step": 10915
+    },
+    {
+      "epoch": 1.2681963404008132,
+      "grad_norm": 0.5675121545791626,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 10916
+    },
+    {
+      "epoch": 1.2683125181527737,
+      "grad_norm": 0.612667977809906,
+      "learning_rate": 0.0001,
+      "loss": 1.5688,
+      "step": 10917
+    },
+    {
+      "epoch": 1.2684286959047342,
+      "grad_norm": 0.6280263066291809,
+      "learning_rate": 0.0001,
+      "loss": 1.694,
+      "step": 10918
+    },
+    {
+      "epoch": 1.2685448736566947,
+      "grad_norm": 0.5767539739608765,
+      "learning_rate": 0.0001,
+      "loss": 1.4455,
+      "step": 10919
+    },
+    {
+      "epoch": 1.2686610514086554,
+      "grad_norm": 0.543782114982605,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 10920
+    },
+    {
+      "epoch": 1.2687772291606159,
+      "grad_norm": 0.5753729939460754,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 10921
+    },
+    {
+      "epoch": 1.2688934069125763,
+      "grad_norm": 0.5873640775680542,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 10922
+    },
+    {
+      "epoch": 1.2690095846645368,
+      "grad_norm": 0.556169867515564,
+      "learning_rate": 0.0001,
+      "loss": 1.4408,
+      "step": 10923
+    },
+    {
+      "epoch": 1.2691257624164973,
+      "grad_norm": 0.588584303855896,
+      "learning_rate": 0.0001,
+      "loss": 1.6303,
+      "step": 10924
+    },
+    {
+      "epoch": 1.2692419401684578,
+      "grad_norm": 0.5426774621009827,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 10925
+    },
+    {
+      "epoch": 1.2693581179204183,
+      "grad_norm": 0.5747091174125671,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 10926
+    },
+    {
+      "epoch": 1.2694742956723788,
+      "grad_norm": 0.5184979438781738,
+      "learning_rate": 0.0001,
+      "loss": 1.308,
+      "step": 10927
+    },
+    {
+      "epoch": 1.2695904734243393,
+      "grad_norm": 0.6157345175743103,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 10928
+    },
+    {
+      "epoch": 1.2697066511762998,
+      "grad_norm": 0.6227047443389893,
+      "learning_rate": 0.0001,
+      "loss": 1.4771,
+      "step": 10929
+    },
+    {
+      "epoch": 1.2698228289282603,
+      "grad_norm": 0.5665978193283081,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 10930
+    },
+    {
+      "epoch": 1.2699390066802207,
+      "grad_norm": 0.5821638703346252,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 10931
+    },
+    {
+      "epoch": 1.2700551844321812,
+      "grad_norm": 0.5837578773498535,
+      "learning_rate": 0.0001,
+      "loss": 1.6321,
+      "step": 10932
+    },
+    {
+      "epoch": 1.2701713621841417,
+      "grad_norm": 0.5726567506790161,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 10933
+    },
+    {
+      "epoch": 1.2702875399361022,
+      "grad_norm": 0.5908870697021484,
+      "learning_rate": 0.0001,
+      "loss": 1.5283,
+      "step": 10934
+    },
+    {
+      "epoch": 1.2704037176880627,
+      "grad_norm": 0.5780557990074158,
+      "learning_rate": 0.0001,
+      "loss": 1.539,
+      "step": 10935
+    },
+    {
+      "epoch": 1.2705198954400232,
+      "grad_norm": 0.5335168838500977,
+      "learning_rate": 0.0001,
+      "loss": 1.3698,
+      "step": 10936
+    },
+    {
+      "epoch": 1.2706360731919837,
+      "grad_norm": 0.5760794281959534,
+      "learning_rate": 0.0001,
+      "loss": 1.4409,
+      "step": 10937
+    },
+    {
+      "epoch": 1.2707522509439442,
+      "grad_norm": 0.5667716264724731,
+      "learning_rate": 0.0001,
+      "loss": 1.4142,
+      "step": 10938
+    },
+    {
+      "epoch": 1.2708684286959047,
+      "grad_norm": 0.5758441686630249,
+      "learning_rate": 0.0001,
+      "loss": 1.4093,
+      "step": 10939
+    },
+    {
+      "epoch": 1.2709846064478652,
+      "grad_norm": 0.5902236104011536,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 10940
+    },
+    {
+      "epoch": 1.2711007841998256,
+      "grad_norm": 0.587958037853241,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 10941
+    },
+    {
+      "epoch": 1.2712169619517861,
+      "grad_norm": 0.6102663278579712,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 10942
+    },
+    {
+      "epoch": 1.2713331397037466,
+      "grad_norm": 0.5977857708930969,
+      "learning_rate": 0.0001,
+      "loss": 1.3835,
+      "step": 10943
+    },
+    {
+      "epoch": 1.271449317455707,
+      "grad_norm": 0.5918571352958679,
+      "learning_rate": 0.0001,
+      "loss": 1.5698,
+      "step": 10944
+    },
+    {
+      "epoch": 1.2715654952076676,
+      "grad_norm": 0.5774389505386353,
+      "learning_rate": 0.0001,
+      "loss": 1.3861,
+      "step": 10945
+    },
+    {
+      "epoch": 1.2716816729596283,
+      "grad_norm": 0.6010609865188599,
+      "learning_rate": 0.0001,
+      "loss": 1.4712,
+      "step": 10946
+    },
+    {
+      "epoch": 1.2717978507115888,
+      "grad_norm": 0.5825393795967102,
+      "learning_rate": 0.0001,
+      "loss": 1.6338,
+      "step": 10947
+    },
+    {
+      "epoch": 1.2719140284635493,
+      "grad_norm": 0.6239941120147705,
+      "learning_rate": 0.0001,
+      "loss": 1.4736,
+      "step": 10948
+    },
+    {
+      "epoch": 1.2720302062155098,
+      "grad_norm": 0.5607348680496216,
+      "learning_rate": 0.0001,
+      "loss": 1.3118,
+      "step": 10949
+    },
+    {
+      "epoch": 1.2721463839674703,
+      "grad_norm": 0.6177657842636108,
+      "learning_rate": 0.0001,
+      "loss": 1.7027,
+      "step": 10950
+    },
+    {
+      "epoch": 1.2722625617194308,
+      "grad_norm": 0.5948476791381836,
+      "learning_rate": 0.0001,
+      "loss": 1.5479,
+      "step": 10951
+    },
+    {
+      "epoch": 1.2723787394713912,
+      "grad_norm": 0.5765991806983948,
+      "learning_rate": 0.0001,
+      "loss": 1.6417,
+      "step": 10952
+    },
+    {
+      "epoch": 1.2724949172233517,
+      "grad_norm": 0.5915812253952026,
+      "learning_rate": 0.0001,
+      "loss": 1.5304,
+      "step": 10953
+    },
+    {
+      "epoch": 1.2726110949753122,
+      "grad_norm": 0.5970553755760193,
+      "learning_rate": 0.0001,
+      "loss": 1.6319,
+      "step": 10954
+    },
+    {
+      "epoch": 1.2727272727272727,
+      "grad_norm": 0.5600100755691528,
+      "learning_rate": 0.0001,
+      "loss": 1.3926,
+      "step": 10955
+    },
+    {
+      "epoch": 1.2728434504792332,
+      "grad_norm": 0.5822739005088806,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 10956
+    },
+    {
+      "epoch": 1.2729596282311937,
+      "grad_norm": 0.6493605971336365,
+      "learning_rate": 0.0001,
+      "loss": 1.3364,
+      "step": 10957
+    },
+    {
+      "epoch": 1.2730758059831542,
+      "grad_norm": 0.5897579193115234,
+      "learning_rate": 0.0001,
+      "loss": 1.5439,
+      "step": 10958
+    },
+    {
+      "epoch": 1.2731919837351147,
+      "grad_norm": 0.5744937062263489,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 10959
+    },
+    {
+      "epoch": 1.2733081614870752,
+      "grad_norm": 0.5749931335449219,
+      "learning_rate": 0.0001,
+      "loss": 1.5115,
+      "step": 10960
+    },
+    {
+      "epoch": 1.2734243392390359,
+      "grad_norm": 0.5711995363235474,
+      "learning_rate": 0.0001,
+      "loss": 1.5007,
+      "step": 10961
+    },
+    {
+      "epoch": 1.2735405169909964,
+      "grad_norm": 0.572600781917572,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 10962
+    },
+    {
+      "epoch": 1.2736566947429568,
+      "grad_norm": 0.5981788635253906,
+      "learning_rate": 0.0001,
+      "loss": 1.652,
+      "step": 10963
+    },
+    {
+      "epoch": 1.2737728724949173,
+      "grad_norm": 0.6050378084182739,
+      "learning_rate": 0.0001,
+      "loss": 1.5157,
+      "step": 10964
+    },
+    {
+      "epoch": 1.2738890502468778,
+      "grad_norm": 0.5755563974380493,
+      "learning_rate": 0.0001,
+      "loss": 1.6514,
+      "step": 10965
+    },
+    {
+      "epoch": 1.2740052279988383,
+      "grad_norm": 0.5697473287582397,
+      "learning_rate": 0.0001,
+      "loss": 1.3886,
+      "step": 10966
+    },
+    {
+      "epoch": 1.2741214057507988,
+      "grad_norm": 0.5516510605812073,
+      "learning_rate": 0.0001,
+      "loss": 1.626,
+      "step": 10967
+    },
+    {
+      "epoch": 1.2742375835027593,
+      "grad_norm": 0.5390398502349854,
+      "learning_rate": 0.0001,
+      "loss": 1.2317,
+      "step": 10968
+    },
+    {
+      "epoch": 1.2743537612547198,
+      "grad_norm": 0.5673388838768005,
+      "learning_rate": 0.0001,
+      "loss": 1.3808,
+      "step": 10969
+    },
+    {
+      "epoch": 1.2744699390066803,
+      "grad_norm": 0.5935032367706299,
+      "learning_rate": 0.0001,
+      "loss": 1.5228,
+      "step": 10970
+    },
+    {
+      "epoch": 1.2745861167586408,
+      "grad_norm": 0.604706883430481,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 10971
+    },
+    {
+      "epoch": 1.2747022945106012,
+      "grad_norm": 0.643751859664917,
+      "learning_rate": 0.0001,
+      "loss": 1.5331,
+      "step": 10972
+    },
+    {
+      "epoch": 1.2748184722625617,
+      "grad_norm": 0.6240712404251099,
+      "learning_rate": 0.0001,
+      "loss": 1.6298,
+      "step": 10973
+    },
+    {
+      "epoch": 1.2749346500145222,
+      "grad_norm": 0.5518888831138611,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 10974
+    },
+    {
+      "epoch": 1.2750508277664827,
+      "grad_norm": 0.6049218773841858,
+      "learning_rate": 0.0001,
+      "loss": 1.4118,
+      "step": 10975
+    },
+    {
+      "epoch": 1.2751670055184432,
+      "grad_norm": 0.6087144017219543,
+      "learning_rate": 0.0001,
+      "loss": 1.4068,
+      "step": 10976
+    },
+    {
+      "epoch": 1.2752831832704037,
+      "grad_norm": 0.5613987445831299,
+      "learning_rate": 0.0001,
+      "loss": 1.6216,
+      "step": 10977
+    },
+    {
+      "epoch": 1.2753993610223642,
+      "grad_norm": 0.6147012710571289,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 10978
+    },
+    {
+      "epoch": 1.2755155387743247,
+      "grad_norm": 0.6045247316360474,
+      "learning_rate": 0.0001,
+      "loss": 1.5239,
+      "step": 10979
+    },
+    {
+      "epoch": 1.2756317165262852,
+      "grad_norm": 0.5130501389503479,
+      "learning_rate": 0.0001,
+      "loss": 1.2315,
+      "step": 10980
+    },
+    {
+      "epoch": 1.2757478942782456,
+      "grad_norm": 0.5664576888084412,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 10981
+    },
+    {
+      "epoch": 1.2758640720302061,
+      "grad_norm": 0.5801703929901123,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 10982
+    },
+    {
+      "epoch": 1.2759802497821666,
+      "grad_norm": 0.6036249399185181,
+      "learning_rate": 0.0001,
+      "loss": 1.3984,
+      "step": 10983
+    },
+    {
+      "epoch": 1.276096427534127,
+      "grad_norm": 0.6023778915405273,
+      "learning_rate": 0.0001,
+      "loss": 1.4279,
+      "step": 10984
+    },
+    {
+      "epoch": 1.2762126052860876,
+      "grad_norm": 0.5702974200248718,
+      "learning_rate": 0.0001,
+      "loss": 1.5797,
+      "step": 10985
+    },
+    {
+      "epoch": 1.276328783038048,
+      "grad_norm": 0.5740309357643127,
+      "learning_rate": 0.0001,
+      "loss": 1.2492,
+      "step": 10986
+    },
+    {
+      "epoch": 1.2764449607900086,
+      "grad_norm": 0.6035803556442261,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 10987
+    },
+    {
+      "epoch": 1.2765611385419693,
+      "grad_norm": 0.6096132397651672,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 10988
+    },
+    {
+      "epoch": 1.2766773162939298,
+      "grad_norm": 0.5646277070045471,
+      "learning_rate": 0.0001,
+      "loss": 1.4154,
+      "step": 10989
+    },
+    {
+      "epoch": 1.2767934940458903,
+      "grad_norm": 0.572085440158844,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 10990
+    },
+    {
+      "epoch": 1.2769096717978508,
+      "grad_norm": 0.5817086696624756,
+      "learning_rate": 0.0001,
+      "loss": 1.57,
+      "step": 10991
+    },
+    {
+      "epoch": 1.2770258495498112,
+      "grad_norm": 0.5460660457611084,
+      "learning_rate": 0.0001,
+      "loss": 1.3342,
+      "step": 10992
+    },
+    {
+      "epoch": 1.2771420273017717,
+      "grad_norm": 0.622367262840271,
+      "learning_rate": 0.0001,
+      "loss": 1.6075,
+      "step": 10993
+    },
+    {
+      "epoch": 1.2772582050537322,
+      "grad_norm": 0.5549526214599609,
+      "learning_rate": 0.0001,
+      "loss": 1.2616,
+      "step": 10994
+    },
+    {
+      "epoch": 1.2773743828056927,
+      "grad_norm": 0.5860145688056946,
+      "learning_rate": 0.0001,
+      "loss": 1.4507,
+      "step": 10995
+    },
+    {
+      "epoch": 1.2774905605576532,
+      "grad_norm": 0.6013427972793579,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 10996
+    },
+    {
+      "epoch": 1.2776067383096137,
+      "grad_norm": 0.6159816980361938,
+      "learning_rate": 0.0001,
+      "loss": 1.6728,
+      "step": 10997
+    },
+    {
+      "epoch": 1.2777229160615742,
+      "grad_norm": 0.5887131690979004,
+      "learning_rate": 0.0001,
+      "loss": 1.3728,
+      "step": 10998
+    },
+    {
+      "epoch": 1.2778390938135347,
+      "grad_norm": 0.555867612361908,
+      "learning_rate": 0.0001,
+      "loss": 1.4287,
+      "step": 10999
+    },
+    {
+      "epoch": 1.2779552715654952,
+      "grad_norm": 0.589248538017273,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 11000
+    },
+    {
+      "epoch": 1.2780714493174556,
+      "grad_norm": 0.5850374698638916,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 11001
+    },
+    {
+      "epoch": 1.2781876270694161,
+      "grad_norm": 0.5467734932899475,
+      "learning_rate": 0.0001,
+      "loss": 1.3924,
+      "step": 11002
+    },
+    {
+      "epoch": 1.2783038048213768,
+      "grad_norm": 0.54944908618927,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 11003
+    },
+    {
+      "epoch": 1.2784199825733373,
+      "grad_norm": 0.5763708353042603,
+      "learning_rate": 0.0001,
+      "loss": 1.6101,
+      "step": 11004
+    },
+    {
+      "epoch": 1.2785361603252978,
+      "grad_norm": 0.5769256353378296,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 11005
+    },
+    {
+      "epoch": 1.2786523380772583,
+      "grad_norm": 0.5926720499992371,
+      "learning_rate": 0.0001,
+      "loss": 1.5032,
+      "step": 11006
+    },
+    {
+      "epoch": 1.2787685158292188,
+      "grad_norm": 0.5912777781486511,
+      "learning_rate": 0.0001,
+      "loss": 1.5129,
+      "step": 11007
+    },
+    {
+      "epoch": 1.2788846935811793,
+      "grad_norm": 0.545050859451294,
+      "learning_rate": 0.0001,
+      "loss": 1.3467,
+      "step": 11008
+    },
+    {
+      "epoch": 1.2790008713331398,
+      "grad_norm": 0.5366930365562439,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 11009
+    },
+    {
+      "epoch": 1.2791170490851003,
+      "grad_norm": 0.5812363028526306,
+      "learning_rate": 0.0001,
+      "loss": 1.4036,
+      "step": 11010
+    },
+    {
+      "epoch": 1.2792332268370608,
+      "grad_norm": 0.6034533381462097,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 11011
+    },
+    {
+      "epoch": 1.2793494045890212,
+      "grad_norm": 0.5460754632949829,
+      "learning_rate": 0.0001,
+      "loss": 1.5372,
+      "step": 11012
+    },
+    {
+      "epoch": 1.2794655823409817,
+      "grad_norm": 0.5772432088851929,
+      "learning_rate": 0.0001,
+      "loss": 1.5513,
+      "step": 11013
+    },
+    {
+      "epoch": 1.2795817600929422,
+      "grad_norm": 0.5885570049285889,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 11014
+    },
+    {
+      "epoch": 1.2796979378449027,
+      "grad_norm": 0.6416093707084656,
+      "learning_rate": 0.0001,
+      "loss": 1.6099,
+      "step": 11015
+    },
+    {
+      "epoch": 1.2798141155968632,
+      "grad_norm": 0.576856791973114,
+      "learning_rate": 0.0001,
+      "loss": 1.5727,
+      "step": 11016
+    },
+    {
+      "epoch": 1.2799302933488237,
+      "grad_norm": 0.5491908192634583,
+      "learning_rate": 0.0001,
+      "loss": 1.3706,
+      "step": 11017
+    },
+    {
+      "epoch": 1.2800464711007842,
+      "grad_norm": 0.5981287956237793,
+      "learning_rate": 0.0001,
+      "loss": 1.6112,
+      "step": 11018
+    },
+    {
+      "epoch": 1.2801626488527447,
+      "grad_norm": 0.5656360387802124,
+      "learning_rate": 0.0001,
+      "loss": 1.3962,
+      "step": 11019
+    },
+    {
+      "epoch": 1.2802788266047052,
+      "grad_norm": 0.52436763048172,
+      "learning_rate": 0.0001,
+      "loss": 1.3104,
+      "step": 11020
+    },
+    {
+      "epoch": 1.2803950043566656,
+      "grad_norm": 0.5909287929534912,
+      "learning_rate": 0.0001,
+      "loss": 1.5743,
+      "step": 11021
+    },
+    {
+      "epoch": 1.2805111821086261,
+      "grad_norm": 0.5850455164909363,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 11022
+    },
+    {
+      "epoch": 1.2806273598605866,
+      "grad_norm": 0.6200546026229858,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 11023
+    },
+    {
+      "epoch": 1.2807435376125471,
+      "grad_norm": 0.5266997218132019,
+      "learning_rate": 0.0001,
+      "loss": 1.3889,
+      "step": 11024
+    },
+    {
+      "epoch": 1.2808597153645076,
+      "grad_norm": 0.5400038957595825,
+      "learning_rate": 0.0001,
+      "loss": 1.3844,
+      "step": 11025
+    },
+    {
+      "epoch": 1.280975893116468,
+      "grad_norm": 0.5605980157852173,
+      "learning_rate": 0.0001,
+      "loss": 1.4657,
+      "step": 11026
+    },
+    {
+      "epoch": 1.2810920708684286,
+      "grad_norm": 0.5961278080940247,
+      "learning_rate": 0.0001,
+      "loss": 1.441,
+      "step": 11027
+    },
+    {
+      "epoch": 1.281208248620389,
+      "grad_norm": 0.5594688653945923,
+      "learning_rate": 0.0001,
+      "loss": 1.5488,
+      "step": 11028
+    },
+    {
+      "epoch": 1.2813244263723496,
+      "grad_norm": 0.5864664912223816,
+      "learning_rate": 0.0001,
+      "loss": 1.2833,
+      "step": 11029
+    },
+    {
+      "epoch": 1.2814406041243103,
+      "grad_norm": 0.579258382320404,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 11030
+    },
+    {
+      "epoch": 1.2815567818762708,
+      "grad_norm": 0.5354270339012146,
+      "learning_rate": 0.0001,
+      "loss": 1.3188,
+      "step": 11031
+    },
+    {
+      "epoch": 1.2816729596282312,
+      "grad_norm": 0.5614170432090759,
+      "learning_rate": 0.0001,
+      "loss": 1.589,
+      "step": 11032
+    },
+    {
+      "epoch": 1.2817891373801917,
+      "grad_norm": 0.5893466472625732,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 11033
+    },
+    {
+      "epoch": 1.2819053151321522,
+      "grad_norm": 0.5984696745872498,
+      "learning_rate": 0.0001,
+      "loss": 1.4392,
+      "step": 11034
+    },
+    {
+      "epoch": 1.2820214928841127,
+      "grad_norm": 0.5932953357696533,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 11035
+    },
+    {
+      "epoch": 1.2821376706360732,
+      "grad_norm": 0.5950168371200562,
+      "learning_rate": 0.0001,
+      "loss": 1.5867,
+      "step": 11036
+    },
+    {
+      "epoch": 1.2822538483880337,
+      "grad_norm": 0.5833539962768555,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 11037
+    },
+    {
+      "epoch": 1.2823700261399942,
+      "grad_norm": 0.6518958210945129,
+      "learning_rate": 0.0001,
+      "loss": 1.276,
+      "step": 11038
+    },
+    {
+      "epoch": 1.2824862038919547,
+      "grad_norm": 0.5316762328147888,
+      "learning_rate": 0.0001,
+      "loss": 1.4414,
+      "step": 11039
+    },
+    {
+      "epoch": 1.2826023816439152,
+      "grad_norm": 0.6396839022636414,
+      "learning_rate": 0.0001,
+      "loss": 1.6239,
+      "step": 11040
+    },
+    {
+      "epoch": 1.2827185593958756,
+      "grad_norm": 0.574507474899292,
+      "learning_rate": 0.0001,
+      "loss": 1.496,
+      "step": 11041
+    },
+    {
+      "epoch": 1.2828347371478361,
+      "grad_norm": 0.5620124936103821,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 11042
+    },
+    {
+      "epoch": 1.2829509148997966,
+      "grad_norm": 0.5690659880638123,
+      "learning_rate": 0.0001,
+      "loss": 1.3846,
+      "step": 11043
+    },
+    {
+      "epoch": 1.2830670926517571,
+      "grad_norm": 0.5415745973587036,
+      "learning_rate": 0.0001,
+      "loss": 1.3145,
+      "step": 11044
+    },
+    {
+      "epoch": 1.2831832704037178,
+      "grad_norm": 0.5873833298683167,
+      "learning_rate": 0.0001,
+      "loss": 1.551,
+      "step": 11045
+    },
+    {
+      "epoch": 1.2832994481556783,
+      "grad_norm": 0.5984528064727783,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 11046
+    },
+    {
+      "epoch": 1.2834156259076388,
+      "grad_norm": 0.6160796284675598,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 11047
+    },
+    {
+      "epoch": 1.2835318036595993,
+      "grad_norm": 0.604532778263092,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 11048
+    },
+    {
+      "epoch": 1.2836479814115598,
+      "grad_norm": 0.5491237044334412,
+      "learning_rate": 0.0001,
+      "loss": 1.3932,
+      "step": 11049
+    },
+    {
+      "epoch": 1.2837641591635203,
+      "grad_norm": 0.5465075969696045,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 11050
+    },
+    {
+      "epoch": 1.2838803369154808,
+      "grad_norm": 0.6033710837364197,
+      "learning_rate": 0.0001,
+      "loss": 1.5147,
+      "step": 11051
+    },
+    {
+      "epoch": 1.2839965146674412,
+      "grad_norm": 0.5667850375175476,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 11052
+    },
+    {
+      "epoch": 1.2841126924194017,
+      "grad_norm": 0.5517511367797852,
+      "learning_rate": 0.0001,
+      "loss": 1.389,
+      "step": 11053
+    },
+    {
+      "epoch": 1.2842288701713622,
+      "grad_norm": 0.5938916802406311,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 11054
+    },
+    {
+      "epoch": 1.2843450479233227,
+      "grad_norm": 0.5698368549346924,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 11055
+    },
+    {
+      "epoch": 1.2844612256752832,
+      "grad_norm": 0.5910446643829346,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 11056
+    },
+    {
+      "epoch": 1.2845774034272437,
+      "grad_norm": 0.5896273255348206,
+      "learning_rate": 0.0001,
+      "loss": 1.3875,
+      "step": 11057
+    },
+    {
+      "epoch": 1.2846935811792042,
+      "grad_norm": 0.5363755226135254,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 11058
+    },
+    {
+      "epoch": 1.2848097589311647,
+      "grad_norm": 0.6186480522155762,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 11059
+    },
+    {
+      "epoch": 1.2849259366831252,
+      "grad_norm": 0.5805513858795166,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 11060
+    },
+    {
+      "epoch": 1.2850421144350856,
+      "grad_norm": 0.6029322743415833,
+      "learning_rate": 0.0001,
+      "loss": 1.3933,
+      "step": 11061
+    },
+    {
+      "epoch": 1.2851582921870461,
+      "grad_norm": 0.5690828561782837,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 11062
+    },
+    {
+      "epoch": 1.2852744699390066,
+      "grad_norm": 0.5746386051177979,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 11063
+    },
+    {
+      "epoch": 1.2853906476909671,
+      "grad_norm": 0.5679745674133301,
+      "learning_rate": 0.0001,
+      "loss": 1.3696,
+      "step": 11064
+    },
+    {
+      "epoch": 1.2855068254429276,
+      "grad_norm": 0.6062312126159668,
+      "learning_rate": 0.0001,
+      "loss": 1.5775,
+      "step": 11065
+    },
+    {
+      "epoch": 1.285623003194888,
+      "grad_norm": 0.5830647945404053,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 11066
+    },
+    {
+      "epoch": 1.2857391809468486,
+      "grad_norm": 0.608407199382782,
+      "learning_rate": 0.0001,
+      "loss": 1.6409,
+      "step": 11067
+    },
+    {
+      "epoch": 1.285855358698809,
+      "grad_norm": 0.5620405673980713,
+      "learning_rate": 0.0001,
+      "loss": 1.4291,
+      "step": 11068
+    },
+    {
+      "epoch": 1.2859715364507696,
+      "grad_norm": 0.5535077452659607,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 11069
+    },
+    {
+      "epoch": 1.28608771420273,
+      "grad_norm": 0.5676767230033875,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 11070
+    },
+    {
+      "epoch": 1.2862038919546908,
+      "grad_norm": 0.572475790977478,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 11071
+    },
+    {
+      "epoch": 1.2863200697066512,
+      "grad_norm": 0.5898261666297913,
+      "learning_rate": 0.0001,
+      "loss": 1.3499,
+      "step": 11072
+    },
+    {
+      "epoch": 1.2864362474586117,
+      "grad_norm": 0.5925431251525879,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 11073
+    },
+    {
+      "epoch": 1.2865524252105722,
+      "grad_norm": 0.6155528426170349,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 11074
+    },
+    {
+      "epoch": 1.2866686029625327,
+      "grad_norm": 0.5945338010787964,
+      "learning_rate": 0.0001,
+      "loss": 1.3821,
+      "step": 11075
+    },
+    {
+      "epoch": 1.2867847807144932,
+      "grad_norm": 0.5834474563598633,
+      "learning_rate": 0.0001,
+      "loss": 1.4329,
+      "step": 11076
+    },
+    {
+      "epoch": 1.2869009584664537,
+      "grad_norm": 0.5647174715995789,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 11077
+    },
+    {
+      "epoch": 1.2870171362184142,
+      "grad_norm": 0.6176016926765442,
+      "learning_rate": 0.0001,
+      "loss": 1.5172,
+      "step": 11078
+    },
+    {
+      "epoch": 1.2871333139703747,
+      "grad_norm": 0.5899996161460876,
+      "learning_rate": 0.0001,
+      "loss": 1.55,
+      "step": 11079
+    },
+    {
+      "epoch": 1.2872494917223352,
+      "grad_norm": 0.605802595615387,
+      "learning_rate": 0.0001,
+      "loss": 1.4564,
+      "step": 11080
+    },
+    {
+      "epoch": 1.2873656694742956,
+      "grad_norm": 0.6284874677658081,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 11081
+    },
+    {
+      "epoch": 1.2874818472262561,
+      "grad_norm": 0.5752767324447632,
+      "learning_rate": 0.0001,
+      "loss": 1.3874,
+      "step": 11082
+    },
+    {
+      "epoch": 1.2875980249782166,
+      "grad_norm": 0.5632016062736511,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 11083
+    },
+    {
+      "epoch": 1.2877142027301771,
+      "grad_norm": 0.5655538439750671,
+      "learning_rate": 0.0001,
+      "loss": 1.6066,
+      "step": 11084
+    },
+    {
+      "epoch": 1.2878303804821376,
+      "grad_norm": 0.5820789337158203,
+      "learning_rate": 0.0001,
+      "loss": 1.4042,
+      "step": 11085
+    },
+    {
+      "epoch": 1.287946558234098,
+      "grad_norm": 0.6027318835258484,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 11086
+    },
+    {
+      "epoch": 1.2880627359860588,
+      "grad_norm": 0.6038578152656555,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 11087
+    },
+    {
+      "epoch": 1.2881789137380193,
+      "grad_norm": 0.5578043460845947,
+      "learning_rate": 0.0001,
+      "loss": 1.4933,
+      "step": 11088
+    },
+    {
+      "epoch": 1.2882950914899798,
+      "grad_norm": 0.568867564201355,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 11089
+    },
+    {
+      "epoch": 1.2884112692419403,
+      "grad_norm": 0.6275894641876221,
+      "learning_rate": 0.0001,
+      "loss": 1.6483,
+      "step": 11090
+    },
+    {
+      "epoch": 1.2885274469939008,
+      "grad_norm": 0.5560985207557678,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 11091
+    },
+    {
+      "epoch": 1.2886436247458612,
+      "grad_norm": 0.5904410481452942,
+      "learning_rate": 0.0001,
+      "loss": 1.4185,
+      "step": 11092
+    },
+    {
+      "epoch": 1.2887598024978217,
+      "grad_norm": 0.5531237125396729,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 11093
+    },
+    {
+      "epoch": 1.2888759802497822,
+      "grad_norm": 0.6044683456420898,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 11094
+    },
+    {
+      "epoch": 1.2889921580017427,
+      "grad_norm": 0.5353182554244995,
+      "learning_rate": 0.0001,
+      "loss": 1.5528,
+      "step": 11095
+    },
+    {
+      "epoch": 1.2891083357537032,
+      "grad_norm": 0.5463939905166626,
+      "learning_rate": 0.0001,
+      "loss": 1.3904,
+      "step": 11096
+    },
+    {
+      "epoch": 1.2892245135056637,
+      "grad_norm": 0.6040199995040894,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 11097
+    },
+    {
+      "epoch": 1.2893406912576242,
+      "grad_norm": 0.567561686038971,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 11098
+    },
+    {
+      "epoch": 1.2894568690095847,
+      "grad_norm": 0.5605993866920471,
+      "learning_rate": 0.0001,
+      "loss": 1.298,
+      "step": 11099
+    },
+    {
+      "epoch": 1.2895730467615452,
+      "grad_norm": 0.5997190475463867,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 11100
+    },
+    {
+      "epoch": 1.2896892245135056,
+      "grad_norm": 0.5995270013809204,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 11101
+    },
+    {
+      "epoch": 1.2898054022654661,
+      "grad_norm": 0.5469571948051453,
+      "learning_rate": 0.0001,
+      "loss": 1.3763,
+      "step": 11102
+    },
+    {
+      "epoch": 1.2899215800174266,
+      "grad_norm": 0.5535038709640503,
+      "learning_rate": 0.0001,
+      "loss": 1.4447,
+      "step": 11103
+    },
+    {
+      "epoch": 1.2900377577693871,
+      "grad_norm": 0.5944443941116333,
+      "learning_rate": 0.0001,
+      "loss": 1.3665,
+      "step": 11104
+    },
+    {
+      "epoch": 1.2901539355213476,
+      "grad_norm": 0.5781884789466858,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 11105
+    },
+    {
+      "epoch": 1.290270113273308,
+      "grad_norm": 0.5380309820175171,
+      "learning_rate": 0.0001,
+      "loss": 1.3409,
+      "step": 11106
+    },
+    {
+      "epoch": 1.2903862910252686,
+      "grad_norm": 0.6192278861999512,
+      "learning_rate": 0.0001,
+      "loss": 1.5987,
+      "step": 11107
+    },
+    {
+      "epoch": 1.290502468777229,
+      "grad_norm": 0.5789281129837036,
+      "learning_rate": 0.0001,
+      "loss": 1.4082,
+      "step": 11108
+    },
+    {
+      "epoch": 1.2906186465291896,
+      "grad_norm": 0.53936368227005,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 11109
+    },
+    {
+      "epoch": 1.29073482428115,
+      "grad_norm": 0.5656925439834595,
+      "learning_rate": 0.0001,
+      "loss": 1.403,
+      "step": 11110
+    },
+    {
+      "epoch": 1.2908510020331105,
+      "grad_norm": 0.5613973736763,
+      "learning_rate": 0.0001,
+      "loss": 1.4213,
+      "step": 11111
+    },
+    {
+      "epoch": 1.290967179785071,
+      "grad_norm": 0.5758323073387146,
+      "learning_rate": 0.0001,
+      "loss": 1.435,
+      "step": 11112
+    },
+    {
+      "epoch": 1.2910833575370317,
+      "grad_norm": 0.5894516706466675,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 11113
+    },
+    {
+      "epoch": 1.2911995352889922,
+      "grad_norm": 0.6146057844161987,
+      "learning_rate": 0.0001,
+      "loss": 1.6059,
+      "step": 11114
+    },
+    {
+      "epoch": 1.2913157130409527,
+      "grad_norm": 0.5563119649887085,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 11115
+    },
+    {
+      "epoch": 1.2914318907929132,
+      "grad_norm": 0.5988980531692505,
+      "learning_rate": 0.0001,
+      "loss": 1.3723,
+      "step": 11116
+    },
+    {
+      "epoch": 1.2915480685448737,
+      "grad_norm": 0.5595650672912598,
+      "learning_rate": 0.0001,
+      "loss": 1.3951,
+      "step": 11117
+    },
+    {
+      "epoch": 1.2916642462968342,
+      "grad_norm": 0.6587527990341187,
+      "learning_rate": 0.0001,
+      "loss": 1.4094,
+      "step": 11118
+    },
+    {
+      "epoch": 1.2917804240487947,
+      "grad_norm": 0.5808438062667847,
+      "learning_rate": 0.0001,
+      "loss": 1.475,
+      "step": 11119
+    },
+    {
+      "epoch": 1.2918966018007552,
+      "grad_norm": 0.5954582095146179,
+      "learning_rate": 0.0001,
+      "loss": 1.3649,
+      "step": 11120
+    },
+    {
+      "epoch": 1.2920127795527157,
+      "grad_norm": 0.5856751799583435,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 11121
+    },
+    {
+      "epoch": 1.2921289573046761,
+      "grad_norm": 0.634069561958313,
+      "learning_rate": 0.0001,
+      "loss": 1.6051,
+      "step": 11122
+    },
+    {
+      "epoch": 1.2922451350566366,
+      "grad_norm": 0.5883323550224304,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 11123
+    },
+    {
+      "epoch": 1.2923613128085971,
+      "grad_norm": 0.5847913026809692,
+      "learning_rate": 0.0001,
+      "loss": 1.3908,
+      "step": 11124
+    },
+    {
+      "epoch": 1.2924774905605576,
+      "grad_norm": 0.6062279939651489,
+      "learning_rate": 0.0001,
+      "loss": 1.6381,
+      "step": 11125
+    },
+    {
+      "epoch": 1.292593668312518,
+      "grad_norm": 0.5873641967773438,
+      "learning_rate": 0.0001,
+      "loss": 1.3793,
+      "step": 11126
+    },
+    {
+      "epoch": 1.2927098460644786,
+      "grad_norm": 0.5468490719795227,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 11127
+    },
+    {
+      "epoch": 1.292826023816439,
+      "grad_norm": 0.6617196798324585,
+      "learning_rate": 0.0001,
+      "loss": 1.7155,
+      "step": 11128
+    },
+    {
+      "epoch": 1.2929422015683998,
+      "grad_norm": 0.6220079660415649,
+      "learning_rate": 0.0001,
+      "loss": 1.5208,
+      "step": 11129
+    },
+    {
+      "epoch": 1.2930583793203603,
+      "grad_norm": 0.5805248022079468,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 11130
+    },
+    {
+      "epoch": 1.2931745570723208,
+      "grad_norm": 0.5914593935012817,
+      "learning_rate": 0.0001,
+      "loss": 1.4297,
+      "step": 11131
+    },
+    {
+      "epoch": 1.2932907348242813,
+      "grad_norm": 0.5885531306266785,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 11132
+    },
+    {
+      "epoch": 1.2934069125762417,
+      "grad_norm": 0.5579025149345398,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 11133
+    },
+    {
+      "epoch": 1.2935230903282022,
+      "grad_norm": 0.6001786589622498,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 11134
+    },
+    {
+      "epoch": 1.2936392680801627,
+      "grad_norm": 0.5882461071014404,
+      "learning_rate": 0.0001,
+      "loss": 1.4804,
+      "step": 11135
+    },
+    {
+      "epoch": 1.2937554458321232,
+      "grad_norm": 0.5694596767425537,
+      "learning_rate": 0.0001,
+      "loss": 1.3329,
+      "step": 11136
+    },
+    {
+      "epoch": 1.2938716235840837,
+      "grad_norm": 0.5779051780700684,
+      "learning_rate": 0.0001,
+      "loss": 1.6199,
+      "step": 11137
+    },
+    {
+      "epoch": 1.2939878013360442,
+      "grad_norm": 0.6211315393447876,
+      "learning_rate": 0.0001,
+      "loss": 1.6313,
+      "step": 11138
+    },
+    {
+      "epoch": 1.2941039790880047,
+      "grad_norm": 0.5611022710800171,
+      "learning_rate": 0.0001,
+      "loss": 1.4081,
+      "step": 11139
+    },
+    {
+      "epoch": 1.2942201568399652,
+      "grad_norm": 0.5747177004814148,
+      "learning_rate": 0.0001,
+      "loss": 1.4622,
+      "step": 11140
+    },
+    {
+      "epoch": 1.2943363345919257,
+      "grad_norm": 0.5457985401153564,
+      "learning_rate": 0.0001,
+      "loss": 1.2923,
+      "step": 11141
+    },
+    {
+      "epoch": 1.2944525123438861,
+      "grad_norm": 0.5791705846786499,
+      "learning_rate": 0.0001,
+      "loss": 1.4877,
+      "step": 11142
+    },
+    {
+      "epoch": 1.2945686900958466,
+      "grad_norm": 0.5527358055114746,
+      "learning_rate": 0.0001,
+      "loss": 1.3722,
+      "step": 11143
+    },
+    {
+      "epoch": 1.2946848678478071,
+      "grad_norm": 0.5855079889297485,
+      "learning_rate": 0.0001,
+      "loss": 1.4477,
+      "step": 11144
+    },
+    {
+      "epoch": 1.2948010455997676,
+      "grad_norm": 0.5779101252555847,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 11145
+    },
+    {
+      "epoch": 1.294917223351728,
+      "grad_norm": 0.6360836029052734,
+      "learning_rate": 0.0001,
+      "loss": 1.81,
+      "step": 11146
+    },
+    {
+      "epoch": 1.2950334011036886,
+      "grad_norm": 0.5778235197067261,
+      "learning_rate": 0.0001,
+      "loss": 1.3919,
+      "step": 11147
+    },
+    {
+      "epoch": 1.295149578855649,
+      "grad_norm": 0.5182598233222961,
+      "learning_rate": 0.0001,
+      "loss": 1.2329,
+      "step": 11148
+    },
+    {
+      "epoch": 1.2952657566076096,
+      "grad_norm": 0.5818975567817688,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 11149
+    },
+    {
+      "epoch": 1.29538193435957,
+      "grad_norm": 0.565624475479126,
+      "learning_rate": 0.0001,
+      "loss": 1.3899,
+      "step": 11150
+    },
+    {
+      "epoch": 1.2954981121115305,
+      "grad_norm": 0.5845359563827515,
+      "learning_rate": 0.0001,
+      "loss": 1.6526,
+      "step": 11151
+    },
+    {
+      "epoch": 1.295614289863491,
+      "grad_norm": 0.5842732787132263,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 11152
+    },
+    {
+      "epoch": 1.2957304676154515,
+      "grad_norm": 0.5768247842788696,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 11153
+    },
+    {
+      "epoch": 1.295846645367412,
+      "grad_norm": 0.6080706119537354,
+      "learning_rate": 0.0001,
+      "loss": 1.4585,
+      "step": 11154
+    },
+    {
+      "epoch": 1.2959628231193727,
+      "grad_norm": 0.6032418608665466,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 11155
+    },
+    {
+      "epoch": 1.2960790008713332,
+      "grad_norm": 0.5942699909210205,
+      "learning_rate": 0.0001,
+      "loss": 1.5165,
+      "step": 11156
+    },
+    {
+      "epoch": 1.2961951786232937,
+      "grad_norm": 0.6410212516784668,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 11157
+    },
+    {
+      "epoch": 1.2963113563752542,
+      "grad_norm": 0.576407253742218,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 11158
+    },
+    {
+      "epoch": 1.2964275341272147,
+      "grad_norm": 0.5596346855163574,
+      "learning_rate": 0.0001,
+      "loss": 1.293,
+      "step": 11159
+    },
+    {
+      "epoch": 1.2965437118791752,
+      "grad_norm": 0.5900881886482239,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 11160
+    },
+    {
+      "epoch": 1.2966598896311357,
+      "grad_norm": 0.622340977191925,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 11161
+    },
+    {
+      "epoch": 1.2967760673830961,
+      "grad_norm": 0.5715858936309814,
+      "learning_rate": 0.0001,
+      "loss": 1.2327,
+      "step": 11162
+    },
+    {
+      "epoch": 1.2968922451350566,
+      "grad_norm": 0.5746508836746216,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 11163
+    },
+    {
+      "epoch": 1.2970084228870171,
+      "grad_norm": 0.5897271037101746,
+      "learning_rate": 0.0001,
+      "loss": 1.4323,
+      "step": 11164
+    },
+    {
+      "epoch": 1.2971246006389776,
+      "grad_norm": 0.607289731502533,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 11165
+    },
+    {
+      "epoch": 1.297240778390938,
+      "grad_norm": 0.5573462843894958,
+      "learning_rate": 0.0001,
+      "loss": 1.2812,
+      "step": 11166
+    },
+    {
+      "epoch": 1.2973569561428986,
+      "grad_norm": 0.5533374547958374,
+      "learning_rate": 0.0001,
+      "loss": 1.3657,
+      "step": 11167
+    },
+    {
+      "epoch": 1.297473133894859,
+      "grad_norm": 0.5837119221687317,
+      "learning_rate": 0.0001,
+      "loss": 1.4336,
+      "step": 11168
+    },
+    {
+      "epoch": 1.2975893116468196,
+      "grad_norm": 0.6015486121177673,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 11169
+    },
+    {
+      "epoch": 1.29770548939878,
+      "grad_norm": 0.5753357410430908,
+      "learning_rate": 0.0001,
+      "loss": 1.4456,
+      "step": 11170
+    },
+    {
+      "epoch": 1.2978216671507408,
+      "grad_norm": 0.578898549079895,
+      "learning_rate": 0.0001,
+      "loss": 1.6588,
+      "step": 11171
+    },
+    {
+      "epoch": 1.2979378449027013,
+      "grad_norm": 0.5891339778900146,
+      "learning_rate": 0.0001,
+      "loss": 1.3499,
+      "step": 11172
+    },
+    {
+      "epoch": 1.2980540226546617,
+      "grad_norm": 0.6010352373123169,
+      "learning_rate": 0.0001,
+      "loss": 1.4964,
+      "step": 11173
+    },
+    {
+      "epoch": 1.2981702004066222,
+      "grad_norm": 0.5747548937797546,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 11174
+    },
+    {
+      "epoch": 1.2982863781585827,
+      "grad_norm": 0.5667596459388733,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 11175
+    },
+    {
+      "epoch": 1.2984025559105432,
+      "grad_norm": 0.611515462398529,
+      "learning_rate": 0.0001,
+      "loss": 1.4193,
+      "step": 11176
+    },
+    {
+      "epoch": 1.2985187336625037,
+      "grad_norm": 0.6042912602424622,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 11177
+    },
+    {
+      "epoch": 1.2986349114144642,
+      "grad_norm": 0.5766490697860718,
+      "learning_rate": 0.0001,
+      "loss": 1.4842,
+      "step": 11178
+    },
+    {
+      "epoch": 1.2987510891664247,
+      "grad_norm": 0.5529983043670654,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 11179
+    },
+    {
+      "epoch": 1.2988672669183852,
+      "grad_norm": 0.5562178492546082,
+      "learning_rate": 0.0001,
+      "loss": 1.5143,
+      "step": 11180
+    },
+    {
+      "epoch": 1.2989834446703457,
+      "grad_norm": 0.6044286489486694,
+      "learning_rate": 0.0001,
+      "loss": 1.4856,
+      "step": 11181
+    },
+    {
+      "epoch": 1.2990996224223061,
+      "grad_norm": 0.606543779373169,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 11182
+    },
+    {
+      "epoch": 1.2992158001742666,
+      "grad_norm": 0.562369704246521,
+      "learning_rate": 0.0001,
+      "loss": 1.3999,
+      "step": 11183
+    },
+    {
+      "epoch": 1.2993319779262271,
+      "grad_norm": 0.5866766571998596,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 11184
+    },
+    {
+      "epoch": 1.2994481556781876,
+      "grad_norm": 0.5652108192443848,
+      "learning_rate": 0.0001,
+      "loss": 1.6506,
+      "step": 11185
+    },
+    {
+      "epoch": 1.299564333430148,
+      "grad_norm": 0.5331071615219116,
+      "learning_rate": 0.0001,
+      "loss": 1.3049,
+      "step": 11186
+    },
+    {
+      "epoch": 1.2996805111821086,
+      "grad_norm": 0.5862208604812622,
+      "learning_rate": 0.0001,
+      "loss": 1.395,
+      "step": 11187
+    },
+    {
+      "epoch": 1.299796688934069,
+      "grad_norm": 0.5830875635147095,
+      "learning_rate": 0.0001,
+      "loss": 1.6407,
+      "step": 11188
+    },
+    {
+      "epoch": 1.2999128666860296,
+      "grad_norm": 0.5540904402732849,
+      "learning_rate": 0.0001,
+      "loss": 1.3349,
+      "step": 11189
+    },
+    {
+      "epoch": 1.30002904443799,
+      "grad_norm": 0.5949186086654663,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 11190
+    },
+    {
+      "epoch": 1.3001452221899505,
+      "grad_norm": 0.6121503710746765,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 11191
+    },
+    {
+      "epoch": 1.300261399941911,
+      "grad_norm": 0.5854485034942627,
+      "learning_rate": 0.0001,
+      "loss": 1.4894,
+      "step": 11192
+    },
+    {
+      "epoch": 1.3003775776938715,
+      "grad_norm": 0.5966970920562744,
+      "learning_rate": 0.0001,
+      "loss": 1.4434,
+      "step": 11193
+    },
+    {
+      "epoch": 1.300493755445832,
+      "grad_norm": 0.567981481552124,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 11194
+    },
+    {
+      "epoch": 1.3006099331977925,
+      "grad_norm": 0.568736732006073,
+      "learning_rate": 0.0001,
+      "loss": 1.5138,
+      "step": 11195
+    },
+    {
+      "epoch": 1.300726110949753,
+      "grad_norm": 0.5789809823036194,
+      "learning_rate": 0.0001,
+      "loss": 1.4117,
+      "step": 11196
+    },
+    {
+      "epoch": 1.3008422887017137,
+      "grad_norm": 0.592512845993042,
+      "learning_rate": 0.0001,
+      "loss": 1.4149,
+      "step": 11197
+    },
+    {
+      "epoch": 1.3009584664536742,
+      "grad_norm": 0.564712405204773,
+      "learning_rate": 0.0001,
+      "loss": 1.355,
+      "step": 11198
+    },
+    {
+      "epoch": 1.3010746442056347,
+      "grad_norm": 0.5966187715530396,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 11199
+    },
+    {
+      "epoch": 1.3011908219575952,
+      "grad_norm": 0.602142870426178,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 11200
+    },
+    {
+      "epoch": 1.3013069997095557,
+      "grad_norm": 0.5634971857070923,
+      "learning_rate": 0.0001,
+      "loss": 1.338,
+      "step": 11201
+    },
+    {
+      "epoch": 1.3014231774615161,
+      "grad_norm": 0.549117922782898,
+      "learning_rate": 0.0001,
+      "loss": 1.4056,
+      "step": 11202
+    },
+    {
+      "epoch": 1.3015393552134766,
+      "grad_norm": 0.5619381070137024,
+      "learning_rate": 0.0001,
+      "loss": 1.5767,
+      "step": 11203
+    },
+    {
+      "epoch": 1.3016555329654371,
+      "grad_norm": 0.5681606531143188,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 11204
+    },
+    {
+      "epoch": 1.3017717107173976,
+      "grad_norm": 0.5451623797416687,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 11205
+    },
+    {
+      "epoch": 1.301887888469358,
+      "grad_norm": 0.550369381904602,
+      "learning_rate": 0.0001,
+      "loss": 1.4309,
+      "step": 11206
+    },
+    {
+      "epoch": 1.3020040662213186,
+      "grad_norm": 0.5932458639144897,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 11207
+    },
+    {
+      "epoch": 1.302120243973279,
+      "grad_norm": 0.529190719127655,
+      "learning_rate": 0.0001,
+      "loss": 1.3753,
+      "step": 11208
+    },
+    {
+      "epoch": 1.3022364217252396,
+      "grad_norm": 0.584304928779602,
+      "learning_rate": 0.0001,
+      "loss": 1.3129,
+      "step": 11209
+    },
+    {
+      "epoch": 1.3023525994772,
+      "grad_norm": 0.6115213632583618,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 11210
+    },
+    {
+      "epoch": 1.3024687772291605,
+      "grad_norm": 0.5591283440589905,
+      "learning_rate": 0.0001,
+      "loss": 1.4011,
+      "step": 11211
+    },
+    {
+      "epoch": 1.302584954981121,
+      "grad_norm": 0.6138569116592407,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 11212
+    },
+    {
+      "epoch": 1.3027011327330817,
+      "grad_norm": 0.5965496897697449,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 11213
+    },
+    {
+      "epoch": 1.3028173104850422,
+      "grad_norm": 0.5974039435386658,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 11214
+    },
+    {
+      "epoch": 1.3029334882370027,
+      "grad_norm": 0.5517904162406921,
+      "learning_rate": 0.0001,
+      "loss": 1.3946,
+      "step": 11215
+    },
+    {
+      "epoch": 1.3030496659889632,
+      "grad_norm": 0.6423289775848389,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 11216
+    },
+    {
+      "epoch": 1.3031658437409237,
+      "grad_norm": 0.5717339515686035,
+      "learning_rate": 0.0001,
+      "loss": 1.4246,
+      "step": 11217
+    },
+    {
+      "epoch": 1.3032820214928842,
+      "grad_norm": 0.5621806383132935,
+      "learning_rate": 0.0001,
+      "loss": 1.457,
+      "step": 11218
+    },
+    {
+      "epoch": 1.3033981992448447,
+      "grad_norm": 0.5442423224449158,
+      "learning_rate": 0.0001,
+      "loss": 1.3835,
+      "step": 11219
+    },
+    {
+      "epoch": 1.3035143769968052,
+      "grad_norm": 0.6146840453147888,
+      "learning_rate": 0.0001,
+      "loss": 1.6248,
+      "step": 11220
+    },
+    {
+      "epoch": 1.3036305547487657,
+      "grad_norm": 0.54494708776474,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 11221
+    },
+    {
+      "epoch": 1.3037467325007261,
+      "grad_norm": 0.5648915767669678,
+      "learning_rate": 0.0001,
+      "loss": 1.346,
+      "step": 11222
+    },
+    {
+      "epoch": 1.3038629102526866,
+      "grad_norm": 0.5836517810821533,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 11223
+    },
+    {
+      "epoch": 1.3039790880046471,
+      "grad_norm": 0.6305371522903442,
+      "learning_rate": 0.0001,
+      "loss": 1.5957,
+      "step": 11224
+    },
+    {
+      "epoch": 1.3040952657566076,
+      "grad_norm": 0.5962154865264893,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 11225
+    },
+    {
+      "epoch": 1.304211443508568,
+      "grad_norm": 0.5581847429275513,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 11226
+    },
+    {
+      "epoch": 1.3043276212605286,
+      "grad_norm": 0.5733591318130493,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 11227
+    },
+    {
+      "epoch": 1.304443799012489,
+      "grad_norm": 0.5814220309257507,
+      "learning_rate": 0.0001,
+      "loss": 1.4996,
+      "step": 11228
+    },
+    {
+      "epoch": 1.3045599767644496,
+      "grad_norm": 0.5474358201026917,
+      "learning_rate": 0.0001,
+      "loss": 1.4572,
+      "step": 11229
+    },
+    {
+      "epoch": 1.30467615451641,
+      "grad_norm": 0.5901806354522705,
+      "learning_rate": 0.0001,
+      "loss": 1.4194,
+      "step": 11230
+    },
+    {
+      "epoch": 1.3047923322683705,
+      "grad_norm": 0.6499736309051514,
+      "learning_rate": 0.0001,
+      "loss": 1.6479,
+      "step": 11231
+    },
+    {
+      "epoch": 1.304908510020331,
+      "grad_norm": 0.5674834251403809,
+      "learning_rate": 0.0001,
+      "loss": 1.2315,
+      "step": 11232
+    },
+    {
+      "epoch": 1.3050246877722915,
+      "grad_norm": 0.5817140340805054,
+      "learning_rate": 0.0001,
+      "loss": 1.5433,
+      "step": 11233
+    },
+    {
+      "epoch": 1.305140865524252,
+      "grad_norm": 0.5742542147636414,
+      "learning_rate": 0.0001,
+      "loss": 1.5561,
+      "step": 11234
+    },
+    {
+      "epoch": 1.3052570432762125,
+      "grad_norm": 0.5343637466430664,
+      "learning_rate": 0.0001,
+      "loss": 1.3944,
+      "step": 11235
+    },
+    {
+      "epoch": 1.305373221028173,
+      "grad_norm": 0.623912513256073,
+      "learning_rate": 0.0001,
+      "loss": 1.5761,
+      "step": 11236
+    },
+    {
+      "epoch": 1.3054893987801335,
+      "grad_norm": 0.5666230916976929,
+      "learning_rate": 0.0001,
+      "loss": 1.3874,
+      "step": 11237
+    },
+    {
+      "epoch": 1.305605576532094,
+      "grad_norm": 0.5873108506202698,
+      "learning_rate": 0.0001,
+      "loss": 1.5621,
+      "step": 11238
+    },
+    {
+      "epoch": 1.3057217542840547,
+      "grad_norm": 0.5752112865447998,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 11239
+    },
+    {
+      "epoch": 1.3058379320360152,
+      "grad_norm": 0.528434693813324,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 11240
+    },
+    {
+      "epoch": 1.3059541097879757,
+      "grad_norm": 0.5661198496818542,
+      "learning_rate": 0.0001,
+      "loss": 1.3099,
+      "step": 11241
+    },
+    {
+      "epoch": 1.3060702875399361,
+      "grad_norm": 0.5762214064598083,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 11242
+    },
+    {
+      "epoch": 1.3061864652918966,
+      "grad_norm": 0.5601356029510498,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 11243
+    },
+    {
+      "epoch": 1.3063026430438571,
+      "grad_norm": 0.5520750284194946,
+      "learning_rate": 0.0001,
+      "loss": 1.3123,
+      "step": 11244
+    },
+    {
+      "epoch": 1.3064188207958176,
+      "grad_norm": 0.5668884515762329,
+      "learning_rate": 0.0001,
+      "loss": 1.3559,
+      "step": 11245
+    },
+    {
+      "epoch": 1.306534998547778,
+      "grad_norm": 0.5902964472770691,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 11246
+    },
+    {
+      "epoch": 1.3066511762997386,
+      "grad_norm": 0.6026374697685242,
+      "learning_rate": 0.0001,
+      "loss": 1.3018,
+      "step": 11247
+    },
+    {
+      "epoch": 1.306767354051699,
+      "grad_norm": 0.5816320776939392,
+      "learning_rate": 0.0001,
+      "loss": 1.2819,
+      "step": 11248
+    },
+    {
+      "epoch": 1.3068835318036596,
+      "grad_norm": 0.5546746850013733,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 11249
+    },
+    {
+      "epoch": 1.30699970955562,
+      "grad_norm": 0.657646119594574,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 11250
+    },
+    {
+      "epoch": 1.3071158873075805,
+      "grad_norm": 0.5970759391784668,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 11251
+    },
+    {
+      "epoch": 1.307232065059541,
+      "grad_norm": 0.6167639493942261,
+      "learning_rate": 0.0001,
+      "loss": 1.665,
+      "step": 11252
+    },
+    {
+      "epoch": 1.3073482428115015,
+      "grad_norm": 0.5807110667228699,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 11253
+    },
+    {
+      "epoch": 1.307464420563462,
+      "grad_norm": 0.5864242315292358,
+      "learning_rate": 0.0001,
+      "loss": 1.4656,
+      "step": 11254
+    },
+    {
+      "epoch": 1.3075805983154227,
+      "grad_norm": 0.5699091553688049,
+      "learning_rate": 0.0001,
+      "loss": 1.4961,
+      "step": 11255
+    },
+    {
+      "epoch": 1.3076967760673832,
+      "grad_norm": 0.5962982177734375,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 11256
+    },
+    {
+      "epoch": 1.3078129538193437,
+      "grad_norm": 0.5681212544441223,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 11257
+    },
+    {
+      "epoch": 1.3079291315713042,
+      "grad_norm": 0.6166642308235168,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 11258
+    },
+    {
+      "epoch": 1.3080453093232647,
+      "grad_norm": 0.5670874714851379,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 11259
+    },
+    {
+      "epoch": 1.3081614870752252,
+      "grad_norm": 0.5790867209434509,
+      "learning_rate": 0.0001,
+      "loss": 1.4262,
+      "step": 11260
+    },
+    {
+      "epoch": 1.3082776648271857,
+      "grad_norm": 0.612891674041748,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 11261
+    },
+    {
+      "epoch": 1.3083938425791461,
+      "grad_norm": 0.5698578357696533,
+      "learning_rate": 0.0001,
+      "loss": 1.424,
+      "step": 11262
+    },
+    {
+      "epoch": 1.3085100203311066,
+      "grad_norm": 0.5675930380821228,
+      "learning_rate": 0.0001,
+      "loss": 1.6215,
+      "step": 11263
+    },
+    {
+      "epoch": 1.3086261980830671,
+      "grad_norm": 0.5446262359619141,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 11264
+    },
+    {
+      "epoch": 1.3087423758350276,
+      "grad_norm": 0.5941632986068726,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 11265
+    },
+    {
+      "epoch": 1.308858553586988,
+      "grad_norm": 0.6017045378684998,
+      "learning_rate": 0.0001,
+      "loss": 1.3836,
+      "step": 11266
+    },
+    {
+      "epoch": 1.3089747313389486,
+      "grad_norm": 0.6072211265563965,
+      "learning_rate": 0.0001,
+      "loss": 1.7059,
+      "step": 11267
+    },
+    {
+      "epoch": 1.309090909090909,
+      "grad_norm": 0.5675247311592102,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 11268
+    },
+    {
+      "epoch": 1.3092070868428696,
+      "grad_norm": 0.6296728849411011,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 11269
+    },
+    {
+      "epoch": 1.30932326459483,
+      "grad_norm": 0.5690588355064392,
+      "learning_rate": 0.0001,
+      "loss": 1.3506,
+      "step": 11270
+    },
+    {
+      "epoch": 1.3094394423467905,
+      "grad_norm": 0.5606512427330017,
+      "learning_rate": 0.0001,
+      "loss": 1.5288,
+      "step": 11271
+    },
+    {
+      "epoch": 1.309555620098751,
+      "grad_norm": 0.6174575686454773,
+      "learning_rate": 0.0001,
+      "loss": 1.4779,
+      "step": 11272
+    },
+    {
+      "epoch": 1.3096717978507115,
+      "grad_norm": 0.5738336443901062,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 11273
+    },
+    {
+      "epoch": 1.309787975602672,
+      "grad_norm": 0.6263105869293213,
+      "learning_rate": 0.0001,
+      "loss": 1.5608,
+      "step": 11274
+    },
+    {
+      "epoch": 1.3099041533546325,
+      "grad_norm": 0.5497341156005859,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 11275
+    },
+    {
+      "epoch": 1.310020331106593,
+      "grad_norm": 0.5970504879951477,
+      "learning_rate": 0.0001,
+      "loss": 1.3028,
+      "step": 11276
+    },
+    {
+      "epoch": 1.3101365088585535,
+      "grad_norm": 0.5593211054801941,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 11277
+    },
+    {
+      "epoch": 1.310252686610514,
+      "grad_norm": 0.5131186246871948,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 11278
+    },
+    {
+      "epoch": 1.3103688643624745,
+      "grad_norm": 0.5773268938064575,
+      "learning_rate": 0.0001,
+      "loss": 1.5205,
+      "step": 11279
+    },
+    {
+      "epoch": 1.310485042114435,
+      "grad_norm": 0.564262330532074,
+      "learning_rate": 0.0001,
+      "loss": 1.5415,
+      "step": 11280
+    },
+    {
+      "epoch": 1.3106012198663957,
+      "grad_norm": 0.5280408263206482,
+      "learning_rate": 0.0001,
+      "loss": 1.4061,
+      "step": 11281
+    },
+    {
+      "epoch": 1.3107173976183562,
+      "grad_norm": 0.5752106308937073,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 11282
+    },
+    {
+      "epoch": 1.3108335753703166,
+      "grad_norm": 0.5846773982048035,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 11283
+    },
+    {
+      "epoch": 1.3109497531222771,
+      "grad_norm": 0.594781756401062,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 11284
+    },
+    {
+      "epoch": 1.3110659308742376,
+      "grad_norm": 0.5700458288192749,
+      "learning_rate": 0.0001,
+      "loss": 1.4275,
+      "step": 11285
+    },
+    {
+      "epoch": 1.311182108626198,
+      "grad_norm": 0.550640344619751,
+      "learning_rate": 0.0001,
+      "loss": 1.2759,
+      "step": 11286
+    },
+    {
+      "epoch": 1.3112982863781586,
+      "grad_norm": 0.5917354822158813,
+      "learning_rate": 0.0001,
+      "loss": 1.5185,
+      "step": 11287
+    },
+    {
+      "epoch": 1.311414464130119,
+      "grad_norm": 0.6752248406410217,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 11288
+    },
+    {
+      "epoch": 1.3115306418820796,
+      "grad_norm": 0.6515881419181824,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 11289
+    },
+    {
+      "epoch": 1.31164681963404,
+      "grad_norm": 0.6039806008338928,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 11290
+    },
+    {
+      "epoch": 1.3117629973860006,
+      "grad_norm": 0.6334456205368042,
+      "learning_rate": 0.0001,
+      "loss": 1.3471,
+      "step": 11291
+    },
+    {
+      "epoch": 1.311879175137961,
+      "grad_norm": 0.574119508266449,
+      "learning_rate": 0.0001,
+      "loss": 1.3821,
+      "step": 11292
+    },
+    {
+      "epoch": 1.3119953528899215,
+      "grad_norm": 0.5512613654136658,
+      "learning_rate": 0.0001,
+      "loss": 1.3874,
+      "step": 11293
+    },
+    {
+      "epoch": 1.312111530641882,
+      "grad_norm": 0.5976804494857788,
+      "learning_rate": 0.0001,
+      "loss": 1.4928,
+      "step": 11294
+    },
+    {
+      "epoch": 1.3122277083938425,
+      "grad_norm": 0.6032594442367554,
+      "learning_rate": 0.0001,
+      "loss": 1.7967,
+      "step": 11295
+    },
+    {
+      "epoch": 1.312343886145803,
+      "grad_norm": 0.5688466429710388,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 11296
+    },
+    {
+      "epoch": 1.3124600638977637,
+      "grad_norm": 0.5861960053443909,
+      "learning_rate": 0.0001,
+      "loss": 1.4174,
+      "step": 11297
+    },
+    {
+      "epoch": 1.3125762416497242,
+      "grad_norm": 0.5523131489753723,
+      "learning_rate": 0.0001,
+      "loss": 1.3868,
+      "step": 11298
+    },
+    {
+      "epoch": 1.3126924194016847,
+      "grad_norm": 0.5910392999649048,
+      "learning_rate": 0.0001,
+      "loss": 1.3825,
+      "step": 11299
+    },
+    {
+      "epoch": 1.3128085971536452,
+      "grad_norm": 0.574103832244873,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 11300
+    },
+    {
+      "epoch": 1.3129247749056057,
+      "grad_norm": 0.5659858584403992,
+      "learning_rate": 0.0001,
+      "loss": 1.3085,
+      "step": 11301
+    },
+    {
+      "epoch": 1.3130409526575662,
+      "grad_norm": 0.5777335166931152,
+      "learning_rate": 0.0001,
+      "loss": 1.3941,
+      "step": 11302
+    },
+    {
+      "epoch": 1.3131571304095266,
+      "grad_norm": 0.6105539202690125,
+      "learning_rate": 0.0001,
+      "loss": 1.6012,
+      "step": 11303
+    },
+    {
+      "epoch": 1.3132733081614871,
+      "grad_norm": 0.605769693851471,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 11304
+    },
+    {
+      "epoch": 1.3133894859134476,
+      "grad_norm": 0.5957057476043701,
+      "learning_rate": 0.0001,
+      "loss": 1.412,
+      "step": 11305
+    },
+    {
+      "epoch": 1.313505663665408,
+      "grad_norm": 0.5396880507469177,
+      "learning_rate": 0.0001,
+      "loss": 1.2279,
+      "step": 11306
+    },
+    {
+      "epoch": 1.3136218414173686,
+      "grad_norm": 0.6096924543380737,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 11307
+    },
+    {
+      "epoch": 1.313738019169329,
+      "grad_norm": 0.5338277816772461,
+      "learning_rate": 0.0001,
+      "loss": 1.3593,
+      "step": 11308
+    },
+    {
+      "epoch": 1.3138541969212896,
+      "grad_norm": 0.5803439617156982,
+      "learning_rate": 0.0001,
+      "loss": 1.537,
+      "step": 11309
+    },
+    {
+      "epoch": 1.31397037467325,
+      "grad_norm": 0.624612033367157,
+      "learning_rate": 0.0001,
+      "loss": 1.4565,
+      "step": 11310
+    },
+    {
+      "epoch": 1.3140865524252106,
+      "grad_norm": 0.6359199285507202,
+      "learning_rate": 0.0001,
+      "loss": 1.4757,
+      "step": 11311
+    },
+    {
+      "epoch": 1.314202730177171,
+      "grad_norm": 0.5825530290603638,
+      "learning_rate": 0.0001,
+      "loss": 1.6943,
+      "step": 11312
+    },
+    {
+      "epoch": 1.3143189079291315,
+      "grad_norm": 0.5994022488594055,
+      "learning_rate": 0.0001,
+      "loss": 1.3063,
+      "step": 11313
+    },
+    {
+      "epoch": 1.314435085681092,
+      "grad_norm": 0.5629523992538452,
+      "learning_rate": 0.0001,
+      "loss": 1.2914,
+      "step": 11314
+    },
+    {
+      "epoch": 1.3145512634330525,
+      "grad_norm": 0.5352174639701843,
+      "learning_rate": 0.0001,
+      "loss": 1.3611,
+      "step": 11315
+    },
+    {
+      "epoch": 1.314667441185013,
+      "grad_norm": 0.5785360336303711,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 11316
+    },
+    {
+      "epoch": 1.3147836189369735,
+      "grad_norm": 0.5326957106590271,
+      "learning_rate": 0.0001,
+      "loss": 1.3276,
+      "step": 11317
+    },
+    {
+      "epoch": 1.314899796688934,
+      "grad_norm": 0.5812473297119141,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 11318
+    },
+    {
+      "epoch": 1.3150159744408945,
+      "grad_norm": 0.567494809627533,
+      "learning_rate": 0.0001,
+      "loss": 1.5323,
+      "step": 11319
+    },
+    {
+      "epoch": 1.315132152192855,
+      "grad_norm": 0.6070596575737,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 11320
+    },
+    {
+      "epoch": 1.3152483299448154,
+      "grad_norm": 0.6242623925209045,
+      "learning_rate": 0.0001,
+      "loss": 1.5356,
+      "step": 11321
+    },
+    {
+      "epoch": 1.315364507696776,
+      "grad_norm": 0.5908821821212769,
+      "learning_rate": 0.0001,
+      "loss": 1.1826,
+      "step": 11322
+    },
+    {
+      "epoch": 1.3154806854487366,
+      "grad_norm": 0.6052210927009583,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 11323
+    },
+    {
+      "epoch": 1.3155968632006971,
+      "grad_norm": 0.5847655534744263,
+      "learning_rate": 0.0001,
+      "loss": 1.3685,
+      "step": 11324
+    },
+    {
+      "epoch": 1.3157130409526576,
+      "grad_norm": 0.5557252168655396,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 11325
+    },
+    {
+      "epoch": 1.315829218704618,
+      "grad_norm": 0.6291106939315796,
+      "learning_rate": 0.0001,
+      "loss": 1.4808,
+      "step": 11326
+    },
+    {
+      "epoch": 1.3159453964565786,
+      "grad_norm": 0.5765290260314941,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 11327
+    },
+    {
+      "epoch": 1.316061574208539,
+      "grad_norm": 0.5529724359512329,
+      "learning_rate": 0.0001,
+      "loss": 1.3843,
+      "step": 11328
+    },
+    {
+      "epoch": 1.3161777519604996,
+      "grad_norm": 0.5492607355117798,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 11329
+    },
+    {
+      "epoch": 1.31629392971246,
+      "grad_norm": 0.6425765156745911,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 11330
+    },
+    {
+      "epoch": 1.3164101074644206,
+      "grad_norm": 0.5800068974494934,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 11331
+    },
+    {
+      "epoch": 1.316526285216381,
+      "grad_norm": 0.5752959251403809,
+      "learning_rate": 0.0001,
+      "loss": 1.5756,
+      "step": 11332
+    },
+    {
+      "epoch": 1.3166424629683415,
+      "grad_norm": 0.5906381607055664,
+      "learning_rate": 0.0001,
+      "loss": 1.5202,
+      "step": 11333
+    },
+    {
+      "epoch": 1.316758640720302,
+      "grad_norm": 0.6053020358085632,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 11334
+    },
+    {
+      "epoch": 1.3168748184722625,
+      "grad_norm": 0.578601598739624,
+      "learning_rate": 0.0001,
+      "loss": 1.5683,
+      "step": 11335
+    },
+    {
+      "epoch": 1.316990996224223,
+      "grad_norm": 0.5814728736877441,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 11336
+    },
+    {
+      "epoch": 1.3171071739761835,
+      "grad_norm": 0.5758288502693176,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 11337
+    },
+    {
+      "epoch": 1.3172233517281442,
+      "grad_norm": 0.5635868906974792,
+      "learning_rate": 0.0001,
+      "loss": 1.4216,
+      "step": 11338
+    },
+    {
+      "epoch": 1.3173395294801047,
+      "grad_norm": 0.5569899082183838,
+      "learning_rate": 0.0001,
+      "loss": 1.3649,
+      "step": 11339
+    },
+    {
+      "epoch": 1.3174557072320652,
+      "grad_norm": 0.5895545482635498,
+      "learning_rate": 0.0001,
+      "loss": 1.4655,
+      "step": 11340
+    },
+    {
+      "epoch": 1.3175718849840257,
+      "grad_norm": 0.5705216526985168,
+      "learning_rate": 0.0001,
+      "loss": 1.4427,
+      "step": 11341
+    },
+    {
+      "epoch": 1.3176880627359862,
+      "grad_norm": 0.5997105240821838,
+      "learning_rate": 0.0001,
+      "loss": 1.5919,
+      "step": 11342
+    },
+    {
+      "epoch": 1.3178042404879466,
+      "grad_norm": 0.566246509552002,
+      "learning_rate": 0.0001,
+      "loss": 1.3959,
+      "step": 11343
+    },
+    {
+      "epoch": 1.3179204182399071,
+      "grad_norm": 0.5906617641448975,
+      "learning_rate": 0.0001,
+      "loss": 1.4819,
+      "step": 11344
+    },
+    {
+      "epoch": 1.3180365959918676,
+      "grad_norm": 0.6158729791641235,
+      "learning_rate": 0.0001,
+      "loss": 1.4863,
+      "step": 11345
+    },
+    {
+      "epoch": 1.318152773743828,
+      "grad_norm": 0.6189556121826172,
+      "learning_rate": 0.0001,
+      "loss": 1.2961,
+      "step": 11346
+    },
+    {
+      "epoch": 1.3182689514957886,
+      "grad_norm": 0.6094973683357239,
+      "learning_rate": 0.0001,
+      "loss": 1.5633,
+      "step": 11347
+    },
+    {
+      "epoch": 1.318385129247749,
+      "grad_norm": 0.579227864742279,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 11348
+    },
+    {
+      "epoch": 1.3185013069997096,
+      "grad_norm": 0.5919841527938843,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 11349
+    },
+    {
+      "epoch": 1.31861748475167,
+      "grad_norm": 0.6161413192749023,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 11350
+    },
+    {
+      "epoch": 1.3187336625036306,
+      "grad_norm": 0.6276363730430603,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 11351
+    },
+    {
+      "epoch": 1.318849840255591,
+      "grad_norm": 0.5806869864463806,
+      "learning_rate": 0.0001,
+      "loss": 1.4119,
+      "step": 11352
+    },
+    {
+      "epoch": 1.3189660180075515,
+      "grad_norm": 0.5905710458755493,
+      "learning_rate": 0.0001,
+      "loss": 1.3384,
+      "step": 11353
+    },
+    {
+      "epoch": 1.319082195759512,
+      "grad_norm": 0.6194556355476379,
+      "learning_rate": 0.0001,
+      "loss": 1.6933,
+      "step": 11354
+    },
+    {
+      "epoch": 1.3191983735114725,
+      "grad_norm": 0.6072413921356201,
+      "learning_rate": 0.0001,
+      "loss": 1.5567,
+      "step": 11355
+    },
+    {
+      "epoch": 1.319314551263433,
+      "grad_norm": 0.6501614451408386,
+      "learning_rate": 0.0001,
+      "loss": 1.5474,
+      "step": 11356
+    },
+    {
+      "epoch": 1.3194307290153935,
+      "grad_norm": 0.6481568217277527,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 11357
+    },
+    {
+      "epoch": 1.319546906767354,
+      "grad_norm": 0.5640507340431213,
+      "learning_rate": 0.0001,
+      "loss": 1.4518,
+      "step": 11358
+    },
+    {
+      "epoch": 1.3196630845193145,
+      "grad_norm": 0.5995705127716064,
+      "learning_rate": 0.0001,
+      "loss": 1.4165,
+      "step": 11359
+    },
+    {
+      "epoch": 1.319779262271275,
+      "grad_norm": 0.561647355556488,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 11360
+    },
+    {
+      "epoch": 1.3198954400232354,
+      "grad_norm": 0.5589419603347778,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 11361
+    },
+    {
+      "epoch": 1.320011617775196,
+      "grad_norm": 0.5721113085746765,
+      "learning_rate": 0.0001,
+      "loss": 1.3061,
+      "step": 11362
+    },
+    {
+      "epoch": 1.3201277955271564,
+      "grad_norm": 0.6529408097267151,
+      "learning_rate": 0.0001,
+      "loss": 1.6187,
+      "step": 11363
+    },
+    {
+      "epoch": 1.320243973279117,
+      "grad_norm": 0.5891768932342529,
+      "learning_rate": 0.0001,
+      "loss": 1.4214,
+      "step": 11364
+    },
+    {
+      "epoch": 1.3203601510310776,
+      "grad_norm": 0.5529371500015259,
+      "learning_rate": 0.0001,
+      "loss": 1.2607,
+      "step": 11365
+    },
+    {
+      "epoch": 1.3204763287830381,
+      "grad_norm": 0.6112219095230103,
+      "learning_rate": 0.0001,
+      "loss": 1.6024,
+      "step": 11366
+    },
+    {
+      "epoch": 1.3205925065349986,
+      "grad_norm": 0.559587299823761,
+      "learning_rate": 0.0001,
+      "loss": 1.3874,
+      "step": 11367
+    },
+    {
+      "epoch": 1.320708684286959,
+      "grad_norm": 0.5865263938903809,
+      "learning_rate": 0.0001,
+      "loss": 1.4663,
+      "step": 11368
+    },
+    {
+      "epoch": 1.3208248620389196,
+      "grad_norm": 0.5650449991226196,
+      "learning_rate": 0.0001,
+      "loss": 1.19,
+      "step": 11369
+    },
+    {
+      "epoch": 1.32094103979088,
+      "grad_norm": 0.5587924122810364,
+      "learning_rate": 0.0001,
+      "loss": 1.5052,
+      "step": 11370
+    },
+    {
+      "epoch": 1.3210572175428406,
+      "grad_norm": 0.6297246217727661,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 11371
+    },
+    {
+      "epoch": 1.321173395294801,
+      "grad_norm": 0.59564608335495,
+      "learning_rate": 0.0001,
+      "loss": 1.3343,
+      "step": 11372
+    },
+    {
+      "epoch": 1.3212895730467615,
+      "grad_norm": 0.5825412273406982,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 11373
+    },
+    {
+      "epoch": 1.321405750798722,
+      "grad_norm": 0.6541181206703186,
+      "learning_rate": 0.0001,
+      "loss": 1.6079,
+      "step": 11374
+    },
+    {
+      "epoch": 1.3215219285506825,
+      "grad_norm": 0.5667305588722229,
+      "learning_rate": 0.0001,
+      "loss": 1.4322,
+      "step": 11375
+    },
+    {
+      "epoch": 1.321638106302643,
+      "grad_norm": 0.6084561347961426,
+      "learning_rate": 0.0001,
+      "loss": 1.4976,
+      "step": 11376
+    },
+    {
+      "epoch": 1.3217542840546035,
+      "grad_norm": 0.6270078420639038,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 11377
+    },
+    {
+      "epoch": 1.321870461806564,
+      "grad_norm": 0.576206386089325,
+      "learning_rate": 0.0001,
+      "loss": 1.4318,
+      "step": 11378
+    },
+    {
+      "epoch": 1.3219866395585245,
+      "grad_norm": 0.6498734354972839,
+      "learning_rate": 0.0001,
+      "loss": 1.6348,
+      "step": 11379
+    },
+    {
+      "epoch": 1.3221028173104852,
+      "grad_norm": 0.5948904752731323,
+      "learning_rate": 0.0001,
+      "loss": 1.6058,
+      "step": 11380
+    },
+    {
+      "epoch": 1.3222189950624457,
+      "grad_norm": 0.5957643389701843,
+      "learning_rate": 0.0001,
+      "loss": 1.5562,
+      "step": 11381
+    },
+    {
+      "epoch": 1.3223351728144062,
+      "grad_norm": 0.5807846784591675,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 11382
+    },
+    {
+      "epoch": 1.3224513505663666,
+      "grad_norm": 0.5830207467079163,
+      "learning_rate": 0.0001,
+      "loss": 1.4322,
+      "step": 11383
+    },
+    {
+      "epoch": 1.3225675283183271,
+      "grad_norm": 0.5690310001373291,
+      "learning_rate": 0.0001,
+      "loss": 1.5376,
+      "step": 11384
+    },
+    {
+      "epoch": 1.3226837060702876,
+      "grad_norm": 0.5671119689941406,
+      "learning_rate": 0.0001,
+      "loss": 1.445,
+      "step": 11385
+    },
+    {
+      "epoch": 1.3227998838222481,
+      "grad_norm": 0.5723904967308044,
+      "learning_rate": 0.0001,
+      "loss": 1.3052,
+      "step": 11386
+    },
+    {
+      "epoch": 1.3229160615742086,
+      "grad_norm": 0.5799258947372437,
+      "learning_rate": 0.0001,
+      "loss": 1.3406,
+      "step": 11387
+    },
+    {
+      "epoch": 1.323032239326169,
+      "grad_norm": 0.6347507238388062,
+      "learning_rate": 0.0001,
+      "loss": 1.6213,
+      "step": 11388
+    },
+    {
+      "epoch": 1.3231484170781296,
+      "grad_norm": 0.6038704514503479,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 11389
+    },
+    {
+      "epoch": 1.32326459483009,
+      "grad_norm": 0.5528449416160583,
+      "learning_rate": 0.0001,
+      "loss": 1.4312,
+      "step": 11390
+    },
+    {
+      "epoch": 1.3233807725820506,
+      "grad_norm": 0.6076419949531555,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 11391
+    },
+    {
+      "epoch": 1.323496950334011,
+      "grad_norm": 0.5536520481109619,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 11392
+    },
+    {
+      "epoch": 1.3236131280859715,
+      "grad_norm": 0.58294278383255,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 11393
+    },
+    {
+      "epoch": 1.323729305837932,
+      "grad_norm": 0.6274382472038269,
+      "learning_rate": 0.0001,
+      "loss": 1.6235,
+      "step": 11394
+    },
+    {
+      "epoch": 1.3238454835898925,
+      "grad_norm": 0.5923705101013184,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 11395
+    },
+    {
+      "epoch": 1.323961661341853,
+      "grad_norm": 0.6048977375030518,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 11396
+    },
+    {
+      "epoch": 1.3240778390938135,
+      "grad_norm": 0.5796459913253784,
+      "learning_rate": 0.0001,
+      "loss": 1.5118,
+      "step": 11397
+    },
+    {
+      "epoch": 1.324194016845774,
+      "grad_norm": 0.632941484451294,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 11398
+    },
+    {
+      "epoch": 1.3243101945977345,
+      "grad_norm": 0.582012951374054,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 11399
+    },
+    {
+      "epoch": 1.324426372349695,
+      "grad_norm": 0.5736321210861206,
+      "learning_rate": 0.0001,
+      "loss": 1.333,
+      "step": 11400
+    },
+    {
+      "epoch": 1.3245425501016554,
+      "grad_norm": 0.5718238949775696,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 11401
+    },
+    {
+      "epoch": 1.324658727853616,
+      "grad_norm": 0.6074345707893372,
+      "learning_rate": 0.0001,
+      "loss": 1.526,
+      "step": 11402
+    },
+    {
+      "epoch": 1.3247749056055764,
+      "grad_norm": 0.6050236225128174,
+      "learning_rate": 0.0001,
+      "loss": 1.4498,
+      "step": 11403
+    },
+    {
+      "epoch": 1.324891083357537,
+      "grad_norm": 0.5850708484649658,
+      "learning_rate": 0.0001,
+      "loss": 1.422,
+      "step": 11404
+    },
+    {
+      "epoch": 1.3250072611094974,
+      "grad_norm": 0.6051907539367676,
+      "learning_rate": 0.0001,
+      "loss": 1.2887,
+      "step": 11405
+    },
+    {
+      "epoch": 1.3251234388614581,
+      "grad_norm": 0.5894731283187866,
+      "learning_rate": 0.0001,
+      "loss": 1.2812,
+      "step": 11406
+    },
+    {
+      "epoch": 1.3252396166134186,
+      "grad_norm": 0.5793226957321167,
+      "learning_rate": 0.0001,
+      "loss": 1.5097,
+      "step": 11407
+    },
+    {
+      "epoch": 1.325355794365379,
+      "grad_norm": 0.564224898815155,
+      "learning_rate": 0.0001,
+      "loss": 1.2043,
+      "step": 11408
+    },
+    {
+      "epoch": 1.3254719721173396,
+      "grad_norm": 0.5819063782691956,
+      "learning_rate": 0.0001,
+      "loss": 1.4239,
+      "step": 11409
+    },
+    {
+      "epoch": 1.3255881498693,
+      "grad_norm": 0.539932370185852,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 11410
+    },
+    {
+      "epoch": 1.3257043276212606,
+      "grad_norm": 0.5646186470985413,
+      "learning_rate": 0.0001,
+      "loss": 1.2979,
+      "step": 11411
+    },
+    {
+      "epoch": 1.325820505373221,
+      "grad_norm": 0.5890675783157349,
+      "learning_rate": 0.0001,
+      "loss": 1.5787,
+      "step": 11412
+    },
+    {
+      "epoch": 1.3259366831251815,
+      "grad_norm": 0.5870331525802612,
+      "learning_rate": 0.0001,
+      "loss": 1.5525,
+      "step": 11413
+    },
+    {
+      "epoch": 1.326052860877142,
+      "grad_norm": 0.5778072476387024,
+      "learning_rate": 0.0001,
+      "loss": 1.4487,
+      "step": 11414
+    },
+    {
+      "epoch": 1.3261690386291025,
+      "grad_norm": 0.5781219601631165,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 11415
+    },
+    {
+      "epoch": 1.326285216381063,
+      "grad_norm": 0.6099265217781067,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 11416
+    },
+    {
+      "epoch": 1.3264013941330235,
+      "grad_norm": 0.5225432515144348,
+      "learning_rate": 0.0001,
+      "loss": 1.2422,
+      "step": 11417
+    },
+    {
+      "epoch": 1.326517571884984,
+      "grad_norm": 0.630908191204071,
+      "learning_rate": 0.0001,
+      "loss": 1.5711,
+      "step": 11418
+    },
+    {
+      "epoch": 1.3266337496369445,
+      "grad_norm": 0.5389137268066406,
+      "learning_rate": 0.0001,
+      "loss": 1.2754,
+      "step": 11419
+    },
+    {
+      "epoch": 1.326749927388905,
+      "grad_norm": 0.6575531959533691,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 11420
+    },
+    {
+      "epoch": 1.3268661051408654,
+      "grad_norm": 0.6499255299568176,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 11421
+    },
+    {
+      "epoch": 1.3269822828928262,
+      "grad_norm": 0.5644387006759644,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 11422
+    },
+    {
+      "epoch": 1.3270984606447866,
+      "grad_norm": 0.6073735356330872,
+      "learning_rate": 0.0001,
+      "loss": 1.5635,
+      "step": 11423
+    },
+    {
+      "epoch": 1.3272146383967471,
+      "grad_norm": 0.6502189636230469,
+      "learning_rate": 0.0001,
+      "loss": 1.5142,
+      "step": 11424
+    },
+    {
+      "epoch": 1.3273308161487076,
+      "grad_norm": 0.6316292881965637,
+      "learning_rate": 0.0001,
+      "loss": 1.5855,
+      "step": 11425
+    },
+    {
+      "epoch": 1.3274469939006681,
+      "grad_norm": 0.5228843092918396,
+      "learning_rate": 0.0001,
+      "loss": 1.3864,
+      "step": 11426
+    },
+    {
+      "epoch": 1.3275631716526286,
+      "grad_norm": 0.5909907221794128,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 11427
+    },
+    {
+      "epoch": 1.327679349404589,
+      "grad_norm": 0.6318853497505188,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 11428
+    },
+    {
+      "epoch": 1.3277955271565496,
+      "grad_norm": 0.5925455689430237,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 11429
+    },
+    {
+      "epoch": 1.32791170490851,
+      "grad_norm": 0.6002850532531738,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 11430
+    },
+    {
+      "epoch": 1.3280278826604706,
+      "grad_norm": 0.6215521693229675,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 11431
+    },
+    {
+      "epoch": 1.328144060412431,
+      "grad_norm": 0.6084332466125488,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 11432
+    },
+    {
+      "epoch": 1.3282602381643915,
+      "grad_norm": 0.6313285231590271,
+      "learning_rate": 0.0001,
+      "loss": 1.5212,
+      "step": 11433
+    },
+    {
+      "epoch": 1.328376415916352,
+      "grad_norm": 0.5687292814254761,
+      "learning_rate": 0.0001,
+      "loss": 1.4264,
+      "step": 11434
+    },
+    {
+      "epoch": 1.3284925936683125,
+      "grad_norm": 0.6608239412307739,
+      "learning_rate": 0.0001,
+      "loss": 1.6383,
+      "step": 11435
+    },
+    {
+      "epoch": 1.328608771420273,
+      "grad_norm": 0.6085755825042725,
+      "learning_rate": 0.0001,
+      "loss": 1.4592,
+      "step": 11436
+    },
+    {
+      "epoch": 1.3287249491722335,
+      "grad_norm": 0.62114417552948,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 11437
+    },
+    {
+      "epoch": 1.328841126924194,
+      "grad_norm": 0.5876638889312744,
+      "learning_rate": 0.0001,
+      "loss": 1.5137,
+      "step": 11438
+    },
+    {
+      "epoch": 1.3289573046761545,
+      "grad_norm": 0.5959676504135132,
+      "learning_rate": 0.0001,
+      "loss": 1.5303,
+      "step": 11439
+    },
+    {
+      "epoch": 1.329073482428115,
+      "grad_norm": 0.569511890411377,
+      "learning_rate": 0.0001,
+      "loss": 1.4759,
+      "step": 11440
+    },
+    {
+      "epoch": 1.3291896601800754,
+      "grad_norm": 0.5644763112068176,
+      "learning_rate": 0.0001,
+      "loss": 1.5487,
+      "step": 11441
+    },
+    {
+      "epoch": 1.329305837932036,
+      "grad_norm": 0.5564660429954529,
+      "learning_rate": 0.0001,
+      "loss": 1.5077,
+      "step": 11442
+    },
+    {
+      "epoch": 1.3294220156839964,
+      "grad_norm": 0.5818626284599304,
+      "learning_rate": 0.0001,
+      "loss": 1.6736,
+      "step": 11443
+    },
+    {
+      "epoch": 1.329538193435957,
+      "grad_norm": 0.6205581426620483,
+      "learning_rate": 0.0001,
+      "loss": 1.3026,
+      "step": 11444
+    },
+    {
+      "epoch": 1.3296543711879174,
+      "grad_norm": 0.5764113068580627,
+      "learning_rate": 0.0001,
+      "loss": 1.414,
+      "step": 11445
+    },
+    {
+      "epoch": 1.329770548939878,
+      "grad_norm": 0.5862987637519836,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 11446
+    },
+    {
+      "epoch": 1.3298867266918384,
+      "grad_norm": 0.5654096007347107,
+      "learning_rate": 0.0001,
+      "loss": 1.2999,
+      "step": 11447
+    },
+    {
+      "epoch": 1.330002904443799,
+      "grad_norm": 0.6210341453552246,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 11448
+    },
+    {
+      "epoch": 1.3301190821957596,
+      "grad_norm": 0.6105040311813354,
+      "learning_rate": 0.0001,
+      "loss": 1.6378,
+      "step": 11449
+    },
+    {
+      "epoch": 1.33023525994772,
+      "grad_norm": 0.5692093372344971,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 11450
+    },
+    {
+      "epoch": 1.3303514376996806,
+      "grad_norm": 0.5587416887283325,
+      "learning_rate": 0.0001,
+      "loss": 1.3777,
+      "step": 11451
+    },
+    {
+      "epoch": 1.330467615451641,
+      "grad_norm": 0.5672914981842041,
+      "learning_rate": 0.0001,
+      "loss": 1.3449,
+      "step": 11452
+    },
+    {
+      "epoch": 1.3305837932036015,
+      "grad_norm": 0.5590872764587402,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 11453
+    },
+    {
+      "epoch": 1.330699970955562,
+      "grad_norm": 0.5552334189414978,
+      "learning_rate": 0.0001,
+      "loss": 1.649,
+      "step": 11454
+    },
+    {
+      "epoch": 1.3308161487075225,
+      "grad_norm": 0.6139841079711914,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 11455
+    },
+    {
+      "epoch": 1.330932326459483,
+      "grad_norm": 0.5881022810935974,
+      "learning_rate": 0.0001,
+      "loss": 1.3873,
+      "step": 11456
+    },
+    {
+      "epoch": 1.3310485042114435,
+      "grad_norm": 0.5796955227851868,
+      "learning_rate": 0.0001,
+      "loss": 1.598,
+      "step": 11457
+    },
+    {
+      "epoch": 1.331164681963404,
+      "grad_norm": 0.5995549559593201,
+      "learning_rate": 0.0001,
+      "loss": 1.437,
+      "step": 11458
+    },
+    {
+      "epoch": 1.3312808597153645,
+      "grad_norm": 0.5792880654335022,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 11459
+    },
+    {
+      "epoch": 1.331397037467325,
+      "grad_norm": 0.5754793882369995,
+      "learning_rate": 0.0001,
+      "loss": 1.6644,
+      "step": 11460
+    },
+    {
+      "epoch": 1.3315132152192855,
+      "grad_norm": 0.593312680721283,
+      "learning_rate": 0.0001,
+      "loss": 1.5568,
+      "step": 11461
+    },
+    {
+      "epoch": 1.331629392971246,
+      "grad_norm": 0.5615260601043701,
+      "learning_rate": 0.0001,
+      "loss": 1.3415,
+      "step": 11462
+    },
+    {
+      "epoch": 1.3317455707232064,
+      "grad_norm": 0.5913121104240417,
+      "learning_rate": 0.0001,
+      "loss": 1.2549,
+      "step": 11463
+    },
+    {
+      "epoch": 1.3318617484751671,
+      "grad_norm": 0.55987948179245,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 11464
+    },
+    {
+      "epoch": 1.3319779262271276,
+      "grad_norm": 0.6145794987678528,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 11465
+    },
+    {
+      "epoch": 1.3320941039790881,
+      "grad_norm": 0.5672405362129211,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 11466
+    },
+    {
+      "epoch": 1.3322102817310486,
+      "grad_norm": 0.6223773956298828,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 11467
+    },
+    {
+      "epoch": 1.332326459483009,
+      "grad_norm": 0.5505383014678955,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 11468
+    },
+    {
+      "epoch": 1.3324426372349696,
+      "grad_norm": 0.6319714784622192,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 11469
+    },
+    {
+      "epoch": 1.33255881498693,
+      "grad_norm": 0.5813085436820984,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 11470
+    },
+    {
+      "epoch": 1.3326749927388906,
+      "grad_norm": 0.5962019562721252,
+      "learning_rate": 0.0001,
+      "loss": 1.3574,
+      "step": 11471
+    },
+    {
+      "epoch": 1.332791170490851,
+      "grad_norm": 0.5934303402900696,
+      "learning_rate": 0.0001,
+      "loss": 1.5318,
+      "step": 11472
+    },
+    {
+      "epoch": 1.3329073482428115,
+      "grad_norm": 0.6068841218948364,
+      "learning_rate": 0.0001,
+      "loss": 1.3101,
+      "step": 11473
+    },
+    {
+      "epoch": 1.333023525994772,
+      "grad_norm": 0.5842049717903137,
+      "learning_rate": 0.0001,
+      "loss": 1.522,
+      "step": 11474
+    },
+    {
+      "epoch": 1.3331397037467325,
+      "grad_norm": 0.6370680928230286,
+      "learning_rate": 0.0001,
+      "loss": 1.4767,
+      "step": 11475
+    },
+    {
+      "epoch": 1.333255881498693,
+      "grad_norm": 0.6403306722640991,
+      "learning_rate": 0.0001,
+      "loss": 1.4926,
+      "step": 11476
+    },
+    {
+      "epoch": 1.3333720592506535,
+      "grad_norm": 0.5705578327178955,
+      "learning_rate": 0.0001,
+      "loss": 1.387,
+      "step": 11477
+    },
+    {
+      "epoch": 1.333488237002614,
+      "grad_norm": 0.5864787101745605,
+      "learning_rate": 0.0001,
+      "loss": 1.623,
+      "step": 11478
+    },
+    {
+      "epoch": 1.3336044147545745,
+      "grad_norm": 0.5673004388809204,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 11479
+    },
+    {
+      "epoch": 1.333720592506535,
+      "grad_norm": 0.5407872200012207,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 11480
+    },
+    {
+      "epoch": 1.3338367702584955,
+      "grad_norm": 0.53462815284729,
+      "learning_rate": 0.0001,
+      "loss": 1.3134,
+      "step": 11481
+    },
+    {
+      "epoch": 1.333952948010456,
+      "grad_norm": 0.5789211392402649,
+      "learning_rate": 0.0001,
+      "loss": 1.3455,
+      "step": 11482
+    },
+    {
+      "epoch": 1.3340691257624164,
+      "grad_norm": 0.6079903244972229,
+      "learning_rate": 0.0001,
+      "loss": 1.3866,
+      "step": 11483
+    },
+    {
+      "epoch": 1.334185303514377,
+      "grad_norm": 0.5825889110565186,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 11484
+    },
+    {
+      "epoch": 1.3343014812663374,
+      "grad_norm": 0.565239429473877,
+      "learning_rate": 0.0001,
+      "loss": 1.4349,
+      "step": 11485
+    },
+    {
+      "epoch": 1.334417659018298,
+      "grad_norm": 0.5922324061393738,
+      "learning_rate": 0.0001,
+      "loss": 1.4571,
+      "step": 11486
+    },
+    {
+      "epoch": 1.3345338367702584,
+      "grad_norm": 0.5655604600906372,
+      "learning_rate": 0.0001,
+      "loss": 1.3975,
+      "step": 11487
+    },
+    {
+      "epoch": 1.3346500145222189,
+      "grad_norm": 0.6273969411849976,
+      "learning_rate": 0.0001,
+      "loss": 1.7001,
+      "step": 11488
+    },
+    {
+      "epoch": 1.3347661922741794,
+      "grad_norm": 0.6035481691360474,
+      "learning_rate": 0.0001,
+      "loss": 1.5982,
+      "step": 11489
+    },
+    {
+      "epoch": 1.33488237002614,
+      "grad_norm": 0.5946334004402161,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 11490
+    },
+    {
+      "epoch": 1.3349985477781006,
+      "grad_norm": 0.587891697883606,
+      "learning_rate": 0.0001,
+      "loss": 1.558,
+      "step": 11491
+    },
+    {
+      "epoch": 1.335114725530061,
+      "grad_norm": 0.6165737509727478,
+      "learning_rate": 0.0001,
+      "loss": 1.5527,
+      "step": 11492
+    },
+    {
+      "epoch": 1.3352309032820215,
+      "grad_norm": 0.6015834808349609,
+      "learning_rate": 0.0001,
+      "loss": 1.5087,
+      "step": 11493
+    },
+    {
+      "epoch": 1.335347081033982,
+      "grad_norm": 0.5921751856803894,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 11494
+    },
+    {
+      "epoch": 1.3354632587859425,
+      "grad_norm": 0.602714478969574,
+      "learning_rate": 0.0001,
+      "loss": 1.2711,
+      "step": 11495
+    },
+    {
+      "epoch": 1.335579436537903,
+      "grad_norm": 0.5740492939949036,
+      "learning_rate": 0.0001,
+      "loss": 1.5967,
+      "step": 11496
+    },
+    {
+      "epoch": 1.3356956142898635,
+      "grad_norm": 0.5856906771659851,
+      "learning_rate": 0.0001,
+      "loss": 1.6184,
+      "step": 11497
+    },
+    {
+      "epoch": 1.335811792041824,
+      "grad_norm": 0.5492340922355652,
+      "learning_rate": 0.0001,
+      "loss": 1.413,
+      "step": 11498
+    },
+    {
+      "epoch": 1.3359279697937845,
+      "grad_norm": 0.6142228245735168,
+      "learning_rate": 0.0001,
+      "loss": 1.4071,
+      "step": 11499
+    },
+    {
+      "epoch": 1.336044147545745,
+      "grad_norm": 0.5769044160842896,
+      "learning_rate": 0.0001,
+      "loss": 1.4111,
+      "step": 11500
+    },
+    {
+      "epoch": 1.3361603252977055,
+      "grad_norm": 0.6063992381095886,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 11501
+    },
+    {
+      "epoch": 1.336276503049666,
+      "grad_norm": 0.5918049812316895,
+      "learning_rate": 0.0001,
+      "loss": 1.4288,
+      "step": 11502
+    },
+    {
+      "epoch": 1.3363926808016264,
+      "grad_norm": 0.6081083416938782,
+      "learning_rate": 0.0001,
+      "loss": 1.4468,
+      "step": 11503
+    },
+    {
+      "epoch": 1.336508858553587,
+      "grad_norm": 0.567497730255127,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 11504
+    },
+    {
+      "epoch": 1.3366250363055474,
+      "grad_norm": 0.5840210318565369,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 11505
+    },
+    {
+      "epoch": 1.3367412140575081,
+      "grad_norm": 0.5692864656448364,
+      "learning_rate": 0.0001,
+      "loss": 1.3308,
+      "step": 11506
+    },
+    {
+      "epoch": 1.3368573918094686,
+      "grad_norm": 0.567046046257019,
+      "learning_rate": 0.0001,
+      "loss": 1.4432,
+      "step": 11507
+    },
+    {
+      "epoch": 1.336973569561429,
+      "grad_norm": 0.5424807667732239,
+      "learning_rate": 0.0001,
+      "loss": 1.3719,
+      "step": 11508
+    },
+    {
+      "epoch": 1.3370897473133896,
+      "grad_norm": 0.5703045725822449,
+      "learning_rate": 0.0001,
+      "loss": 1.5068,
+      "step": 11509
+    },
+    {
+      "epoch": 1.33720592506535,
+      "grad_norm": 0.5481063723564148,
+      "learning_rate": 0.0001,
+      "loss": 1.3803,
+      "step": 11510
+    },
+    {
+      "epoch": 1.3373221028173106,
+      "grad_norm": 0.5333367586135864,
+      "learning_rate": 0.0001,
+      "loss": 1.3449,
+      "step": 11511
+    },
+    {
+      "epoch": 1.337438280569271,
+      "grad_norm": 0.6329618096351624,
+      "learning_rate": 0.0001,
+      "loss": 1.5148,
+      "step": 11512
+    },
+    {
+      "epoch": 1.3375544583212315,
+      "grad_norm": 0.5826229453086853,
+      "learning_rate": 0.0001,
+      "loss": 1.3026,
+      "step": 11513
+    },
+    {
+      "epoch": 1.337670636073192,
+      "grad_norm": 0.5669860243797302,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 11514
+    },
+    {
+      "epoch": 1.3377868138251525,
+      "grad_norm": 0.5839524865150452,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 11515
+    },
+    {
+      "epoch": 1.337902991577113,
+      "grad_norm": 0.6274355053901672,
+      "learning_rate": 0.0001,
+      "loss": 1.6241,
+      "step": 11516
+    },
+    {
+      "epoch": 1.3380191693290735,
+      "grad_norm": 0.5699935555458069,
+      "learning_rate": 0.0001,
+      "loss": 1.3266,
+      "step": 11517
+    },
+    {
+      "epoch": 1.338135347081034,
+      "grad_norm": 0.5473787188529968,
+      "learning_rate": 0.0001,
+      "loss": 1.3531,
+      "step": 11518
+    },
+    {
+      "epoch": 1.3382515248329945,
+      "grad_norm": 0.605440080165863,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 11519
+    },
+    {
+      "epoch": 1.338367702584955,
+      "grad_norm": 0.558005690574646,
+      "learning_rate": 0.0001,
+      "loss": 1.3049,
+      "step": 11520
+    },
+    {
+      "epoch": 1.3384838803369155,
+      "grad_norm": 0.6115478873252869,
+      "learning_rate": 0.0001,
+      "loss": 1.5498,
+      "step": 11521
+    },
+    {
+      "epoch": 1.338600058088876,
+      "grad_norm": 0.590194821357727,
+      "learning_rate": 0.0001,
+      "loss": 1.4244,
+      "step": 11522
+    },
+    {
+      "epoch": 1.3387162358408364,
+      "grad_norm": 0.5935150980949402,
+      "learning_rate": 0.0001,
+      "loss": 1.4641,
+      "step": 11523
+    },
+    {
+      "epoch": 1.338832413592797,
+      "grad_norm": 0.6085209846496582,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 11524
+    },
+    {
+      "epoch": 1.3389485913447574,
+      "grad_norm": 0.5541259050369263,
+      "learning_rate": 0.0001,
+      "loss": 1.332,
+      "step": 11525
+    },
+    {
+      "epoch": 1.339064769096718,
+      "grad_norm": 0.5870988368988037,
+      "learning_rate": 0.0001,
+      "loss": 1.5353,
+      "step": 11526
+    },
+    {
+      "epoch": 1.3391809468486784,
+      "grad_norm": 0.5646886825561523,
+      "learning_rate": 0.0001,
+      "loss": 1.5916,
+      "step": 11527
+    },
+    {
+      "epoch": 1.3392971246006389,
+      "grad_norm": 0.5684593319892883,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 11528
+    },
+    {
+      "epoch": 1.3394133023525994,
+      "grad_norm": 0.6310387849807739,
+      "learning_rate": 0.0001,
+      "loss": 1.7766,
+      "step": 11529
+    },
+    {
+      "epoch": 1.3395294801045599,
+      "grad_norm": 0.6121505498886108,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 11530
+    },
+    {
+      "epoch": 1.3396456578565203,
+      "grad_norm": 0.5295522212982178,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 11531
+    },
+    {
+      "epoch": 1.339761835608481,
+      "grad_norm": 0.5718926191329956,
+      "learning_rate": 0.0001,
+      "loss": 1.3464,
+      "step": 11532
+    },
+    {
+      "epoch": 1.3398780133604415,
+      "grad_norm": 0.598277747631073,
+      "learning_rate": 0.0001,
+      "loss": 1.5392,
+      "step": 11533
+    },
+    {
+      "epoch": 1.339994191112402,
+      "grad_norm": 0.5410280227661133,
+      "learning_rate": 0.0001,
+      "loss": 1.3229,
+      "step": 11534
+    },
+    {
+      "epoch": 1.3401103688643625,
+      "grad_norm": 0.545367956161499,
+      "learning_rate": 0.0001,
+      "loss": 1.5294,
+      "step": 11535
+    },
+    {
+      "epoch": 1.340226546616323,
+      "grad_norm": 0.5809006690979004,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 11536
+    },
+    {
+      "epoch": 1.3403427243682835,
+      "grad_norm": 0.6075916886329651,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 11537
+    },
+    {
+      "epoch": 1.340458902120244,
+      "grad_norm": 0.5859774351119995,
+      "learning_rate": 0.0001,
+      "loss": 1.214,
+      "step": 11538
+    },
+    {
+      "epoch": 1.3405750798722045,
+      "grad_norm": 0.5794239640235901,
+      "learning_rate": 0.0001,
+      "loss": 1.3904,
+      "step": 11539
+    },
+    {
+      "epoch": 1.340691257624165,
+      "grad_norm": 0.5825338959693909,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 11540
+    },
+    {
+      "epoch": 1.3408074353761255,
+      "grad_norm": 0.5686221718788147,
+      "learning_rate": 0.0001,
+      "loss": 1.2162,
+      "step": 11541
+    },
+    {
+      "epoch": 1.340923613128086,
+      "grad_norm": 0.5766289234161377,
+      "learning_rate": 0.0001,
+      "loss": 1.5009,
+      "step": 11542
+    },
+    {
+      "epoch": 1.3410397908800464,
+      "grad_norm": 0.6284175515174866,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 11543
+    },
+    {
+      "epoch": 1.341155968632007,
+      "grad_norm": 0.6178475022315979,
+      "learning_rate": 0.0001,
+      "loss": 1.261,
+      "step": 11544
+    },
+    {
+      "epoch": 1.3412721463839674,
+      "grad_norm": 0.6055488586425781,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 11545
+    },
+    {
+      "epoch": 1.341388324135928,
+      "grad_norm": 0.6022990345954895,
+      "learning_rate": 0.0001,
+      "loss": 1.7137,
+      "step": 11546
+    },
+    {
+      "epoch": 1.3415045018878884,
+      "grad_norm": 0.6357380151748657,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 11547
+    },
+    {
+      "epoch": 1.341620679639849,
+      "grad_norm": 0.582905650138855,
+      "learning_rate": 0.0001,
+      "loss": 1.405,
+      "step": 11548
+    },
+    {
+      "epoch": 1.3417368573918096,
+      "grad_norm": 0.5897303223609924,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 11549
+    },
+    {
+      "epoch": 1.34185303514377,
+      "grad_norm": 0.6201091408729553,
+      "learning_rate": 0.0001,
+      "loss": 1.305,
+      "step": 11550
+    },
+    {
+      "epoch": 1.3419692128957306,
+      "grad_norm": 0.5758824348449707,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 11551
+    },
+    {
+      "epoch": 1.342085390647691,
+      "grad_norm": 0.5709869861602783,
+      "learning_rate": 0.0001,
+      "loss": 1.3256,
+      "step": 11552
+    },
+    {
+      "epoch": 1.3422015683996515,
+      "grad_norm": 0.6379136443138123,
+      "learning_rate": 0.0001,
+      "loss": 1.4303,
+      "step": 11553
+    },
+    {
+      "epoch": 1.342317746151612,
+      "grad_norm": 0.5988776087760925,
+      "learning_rate": 0.0001,
+      "loss": 1.4281,
+      "step": 11554
+    },
+    {
+      "epoch": 1.3424339239035725,
+      "grad_norm": 0.597923994064331,
+      "learning_rate": 0.0001,
+      "loss": 1.4695,
+      "step": 11555
+    },
+    {
+      "epoch": 1.342550101655533,
+      "grad_norm": 0.578001081943512,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 11556
+    },
+    {
+      "epoch": 1.3426662794074935,
+      "grad_norm": 0.5900245308876038,
+      "learning_rate": 0.0001,
+      "loss": 1.5779,
+      "step": 11557
+    },
+    {
+      "epoch": 1.342782457159454,
+      "grad_norm": 0.5991559028625488,
+      "learning_rate": 0.0001,
+      "loss": 1.5152,
+      "step": 11558
+    },
+    {
+      "epoch": 1.3428986349114145,
+      "grad_norm": 0.5901554822921753,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 11559
+    },
+    {
+      "epoch": 1.343014812663375,
+      "grad_norm": 0.5958110094070435,
+      "learning_rate": 0.0001,
+      "loss": 1.4813,
+      "step": 11560
+    },
+    {
+      "epoch": 1.3431309904153355,
+      "grad_norm": 0.6428405046463013,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 11561
+    },
+    {
+      "epoch": 1.343247168167296,
+      "grad_norm": 0.6413325071334839,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 11562
+    },
+    {
+      "epoch": 1.3433633459192564,
+      "grad_norm": 0.596393346786499,
+      "learning_rate": 0.0001,
+      "loss": 1.4102,
+      "step": 11563
+    },
+    {
+      "epoch": 1.343479523671217,
+      "grad_norm": 0.6022195219993591,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 11564
+    },
+    {
+      "epoch": 1.3435957014231774,
+      "grad_norm": 0.5690363049507141,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 11565
+    },
+    {
+      "epoch": 1.343711879175138,
+      "grad_norm": 0.5736088156700134,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 11566
+    },
+    {
+      "epoch": 1.3438280569270984,
+      "grad_norm": 0.551696240901947,
+      "learning_rate": 0.0001,
+      "loss": 1.2572,
+      "step": 11567
+    },
+    {
+      "epoch": 1.3439442346790589,
+      "grad_norm": 0.5662961602210999,
+      "learning_rate": 0.0001,
+      "loss": 1.5853,
+      "step": 11568
+    },
+    {
+      "epoch": 1.3440604124310194,
+      "grad_norm": 0.5956161022186279,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 11569
+    },
+    {
+      "epoch": 1.3441765901829799,
+      "grad_norm": 0.5635676383972168,
+      "learning_rate": 0.0001,
+      "loss": 1.5237,
+      "step": 11570
+    },
+    {
+      "epoch": 1.3442927679349403,
+      "grad_norm": 0.5952593684196472,
+      "learning_rate": 0.0001,
+      "loss": 1.3867,
+      "step": 11571
+    },
+    {
+      "epoch": 1.3444089456869008,
+      "grad_norm": 0.5866329073905945,
+      "learning_rate": 0.0001,
+      "loss": 1.5998,
+      "step": 11572
+    },
+    {
+      "epoch": 1.3445251234388613,
+      "grad_norm": 0.6334074139595032,
+      "learning_rate": 0.0001,
+      "loss": 1.6183,
+      "step": 11573
+    },
+    {
+      "epoch": 1.344641301190822,
+      "grad_norm": 0.5864160656929016,
+      "learning_rate": 0.0001,
+      "loss": 1.3639,
+      "step": 11574
+    },
+    {
+      "epoch": 1.3447574789427825,
+      "grad_norm": 0.5770210027694702,
+      "learning_rate": 0.0001,
+      "loss": 1.3357,
+      "step": 11575
+    },
+    {
+      "epoch": 1.344873656694743,
+      "grad_norm": 0.5954274535179138,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 11576
+    },
+    {
+      "epoch": 1.3449898344467035,
+      "grad_norm": 0.6258078813552856,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 11577
+    },
+    {
+      "epoch": 1.345106012198664,
+      "grad_norm": 0.62962406873703,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 11578
+    },
+    {
+      "epoch": 1.3452221899506245,
+      "grad_norm": 0.572242259979248,
+      "learning_rate": 0.0001,
+      "loss": 1.2881,
+      "step": 11579
+    },
+    {
+      "epoch": 1.345338367702585,
+      "grad_norm": 0.5981161594390869,
+      "learning_rate": 0.0001,
+      "loss": 1.5231,
+      "step": 11580
+    },
+    {
+      "epoch": 1.3454545454545455,
+      "grad_norm": 0.5858927369117737,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 11581
+    },
+    {
+      "epoch": 1.345570723206506,
+      "grad_norm": 0.5862258672714233,
+      "learning_rate": 0.0001,
+      "loss": 1.456,
+      "step": 11582
+    },
+    {
+      "epoch": 1.3456869009584664,
+      "grad_norm": 0.5800897479057312,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 11583
+    },
+    {
+      "epoch": 1.345803078710427,
+      "grad_norm": 0.5962756276130676,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 11584
+    },
+    {
+      "epoch": 1.3459192564623874,
+      "grad_norm": 0.5521575808525085,
+      "learning_rate": 0.0001,
+      "loss": 1.3611,
+      "step": 11585
+    },
+    {
+      "epoch": 1.346035434214348,
+      "grad_norm": 0.5704815983772278,
+      "learning_rate": 0.0001,
+      "loss": 1.4965,
+      "step": 11586
+    },
+    {
+      "epoch": 1.3461516119663084,
+      "grad_norm": 0.55373615026474,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 11587
+    },
+    {
+      "epoch": 1.3462677897182689,
+      "grad_norm": 0.5447104573249817,
+      "learning_rate": 0.0001,
+      "loss": 1.3546,
+      "step": 11588
+    },
+    {
+      "epoch": 1.3463839674702294,
+      "grad_norm": 0.6218578815460205,
+      "learning_rate": 0.0001,
+      "loss": 1.4718,
+      "step": 11589
+    },
+    {
+      "epoch": 1.34650014522219,
+      "grad_norm": 0.5952721238136292,
+      "learning_rate": 0.0001,
+      "loss": 1.3838,
+      "step": 11590
+    },
+    {
+      "epoch": 1.3466163229741506,
+      "grad_norm": 0.5739622712135315,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 11591
+    },
+    {
+      "epoch": 1.346732500726111,
+      "grad_norm": 0.5628431439399719,
+      "learning_rate": 0.0001,
+      "loss": 1.3305,
+      "step": 11592
+    },
+    {
+      "epoch": 1.3468486784780715,
+      "grad_norm": 0.5847616195678711,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 11593
+    },
+    {
+      "epoch": 1.346964856230032,
+      "grad_norm": 0.5959730744361877,
+      "learning_rate": 0.0001,
+      "loss": 1.4444,
+      "step": 11594
+    },
+    {
+      "epoch": 1.3470810339819925,
+      "grad_norm": 0.5944980382919312,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 11595
+    },
+    {
+      "epoch": 1.347197211733953,
+      "grad_norm": 0.6205131411552429,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 11596
+    },
+    {
+      "epoch": 1.3473133894859135,
+      "grad_norm": 0.6303896903991699,
+      "learning_rate": 0.0001,
+      "loss": 1.6072,
+      "step": 11597
+    },
+    {
+      "epoch": 1.347429567237874,
+      "grad_norm": 0.5928091406822205,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 11598
+    },
+    {
+      "epoch": 1.3475457449898345,
+      "grad_norm": 0.6262045502662659,
+      "learning_rate": 0.0001,
+      "loss": 1.5543,
+      "step": 11599
+    },
+    {
+      "epoch": 1.347661922741795,
+      "grad_norm": 0.6027948260307312,
+      "learning_rate": 0.0001,
+      "loss": 1.4472,
+      "step": 11600
+    },
+    {
+      "epoch": 1.3477781004937555,
+      "grad_norm": 0.5992240905761719,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 11601
+    },
+    {
+      "epoch": 1.347894278245716,
+      "grad_norm": 0.5964482426643372,
+      "learning_rate": 0.0001,
+      "loss": 1.4014,
+      "step": 11602
+    },
+    {
+      "epoch": 1.3480104559976764,
+      "grad_norm": 0.5847633481025696,
+      "learning_rate": 0.0001,
+      "loss": 1.5462,
+      "step": 11603
+    },
+    {
+      "epoch": 1.348126633749637,
+      "grad_norm": 0.5737441778182983,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 11604
+    },
+    {
+      "epoch": 1.3482428115015974,
+      "grad_norm": 0.5500698685646057,
+      "learning_rate": 0.0001,
+      "loss": 1.4647,
+      "step": 11605
+    },
+    {
+      "epoch": 1.348358989253558,
+      "grad_norm": 0.5695959329605103,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 11606
+    },
+    {
+      "epoch": 1.3484751670055184,
+      "grad_norm": 0.5738762617111206,
+      "learning_rate": 0.0001,
+      "loss": 1.6238,
+      "step": 11607
+    },
+    {
+      "epoch": 1.3485913447574789,
+      "grad_norm": 0.5387170314788818,
+      "learning_rate": 0.0001,
+      "loss": 1.402,
+      "step": 11608
+    },
+    {
+      "epoch": 1.3487075225094394,
+      "grad_norm": 0.5509867072105408,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 11609
+    },
+    {
+      "epoch": 1.3488237002613999,
+      "grad_norm": 0.5982716679573059,
+      "learning_rate": 0.0001,
+      "loss": 1.4475,
+      "step": 11610
+    },
+    {
+      "epoch": 1.3489398780133603,
+      "grad_norm": 0.5490851402282715,
+      "learning_rate": 0.0001,
+      "loss": 1.4278,
+      "step": 11611
+    },
+    {
+      "epoch": 1.3490560557653208,
+      "grad_norm": 0.5994747877120972,
+      "learning_rate": 0.0001,
+      "loss": 1.4825,
+      "step": 11612
+    },
+    {
+      "epoch": 1.3491722335172813,
+      "grad_norm": 0.6025500297546387,
+      "learning_rate": 0.0001,
+      "loss": 1.6634,
+      "step": 11613
+    },
+    {
+      "epoch": 1.3492884112692418,
+      "grad_norm": 0.5859069228172302,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 11614
+    },
+    {
+      "epoch": 1.3494045890212023,
+      "grad_norm": 0.6145164370536804,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 11615
+    },
+    {
+      "epoch": 1.349520766773163,
+      "grad_norm": 0.6078627109527588,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 11616
+    },
+    {
+      "epoch": 1.3496369445251235,
+      "grad_norm": 0.5715606808662415,
+      "learning_rate": 0.0001,
+      "loss": 1.5395,
+      "step": 11617
+    },
+    {
+      "epoch": 1.349753122277084,
+      "grad_norm": 0.5815566778182983,
+      "learning_rate": 0.0001,
+      "loss": 1.403,
+      "step": 11618
+    },
+    {
+      "epoch": 1.3498693000290445,
+      "grad_norm": 0.5756155252456665,
+      "learning_rate": 0.0001,
+      "loss": 1.3829,
+      "step": 11619
+    },
+    {
+      "epoch": 1.349985477781005,
+      "grad_norm": 0.6013368368148804,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 11620
+    },
+    {
+      "epoch": 1.3501016555329655,
+      "grad_norm": 0.5688734650611877,
+      "learning_rate": 0.0001,
+      "loss": 1.3319,
+      "step": 11621
+    },
+    {
+      "epoch": 1.350217833284926,
+      "grad_norm": 0.5575860738754272,
+      "learning_rate": 0.0001,
+      "loss": 1.3737,
+      "step": 11622
+    },
+    {
+      "epoch": 1.3503340110368864,
+      "grad_norm": 0.6282293200492859,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 11623
+    },
+    {
+      "epoch": 1.350450188788847,
+      "grad_norm": 0.5837210416793823,
+      "learning_rate": 0.0001,
+      "loss": 1.3822,
+      "step": 11624
+    },
+    {
+      "epoch": 1.3505663665408074,
+      "grad_norm": 0.6083855032920837,
+      "learning_rate": 0.0001,
+      "loss": 1.5915,
+      "step": 11625
+    },
+    {
+      "epoch": 1.350682544292768,
+      "grad_norm": 0.5760353803634644,
+      "learning_rate": 0.0001,
+      "loss": 1.4134,
+      "step": 11626
+    },
+    {
+      "epoch": 1.3507987220447284,
+      "grad_norm": 0.5956183075904846,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 11627
+    },
+    {
+      "epoch": 1.3509148997966889,
+      "grad_norm": 0.5973549485206604,
+      "learning_rate": 0.0001,
+      "loss": 1.6133,
+      "step": 11628
+    },
+    {
+      "epoch": 1.3510310775486494,
+      "grad_norm": 0.5945565104484558,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 11629
+    },
+    {
+      "epoch": 1.3511472553006099,
+      "grad_norm": 0.5747148394584656,
+      "learning_rate": 0.0001,
+      "loss": 1.383,
+      "step": 11630
+    },
+    {
+      "epoch": 1.3512634330525704,
+      "grad_norm": 0.6089474558830261,
+      "learning_rate": 0.0001,
+      "loss": 1.4009,
+      "step": 11631
+    },
+    {
+      "epoch": 1.351379610804531,
+      "grad_norm": 0.5638538599014282,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 11632
+    },
+    {
+      "epoch": 1.3514957885564916,
+      "grad_norm": 0.5719699263572693,
+      "learning_rate": 0.0001,
+      "loss": 1.3883,
+      "step": 11633
+    },
+    {
+      "epoch": 1.351611966308452,
+      "grad_norm": 0.6159690618515015,
+      "learning_rate": 0.0001,
+      "loss": 1.4904,
+      "step": 11634
+    },
+    {
+      "epoch": 1.3517281440604125,
+      "grad_norm": 0.6014747023582458,
+      "learning_rate": 0.0001,
+      "loss": 1.5219,
+      "step": 11635
+    },
+    {
+      "epoch": 1.351844321812373,
+      "grad_norm": 0.5920547246932983,
+      "learning_rate": 0.0001,
+      "loss": 1.4401,
+      "step": 11636
+    },
+    {
+      "epoch": 1.3519604995643335,
+      "grad_norm": 0.596462607383728,
+      "learning_rate": 0.0001,
+      "loss": 1.4317,
+      "step": 11637
+    },
+    {
+      "epoch": 1.352076677316294,
+      "grad_norm": 0.5985187888145447,
+      "learning_rate": 0.0001,
+      "loss": 1.4105,
+      "step": 11638
+    },
+    {
+      "epoch": 1.3521928550682545,
+      "grad_norm": 0.556800901889801,
+      "learning_rate": 0.0001,
+      "loss": 1.31,
+      "step": 11639
+    },
+    {
+      "epoch": 1.352309032820215,
+      "grad_norm": 0.5883356332778931,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 11640
+    },
+    {
+      "epoch": 1.3524252105721755,
+      "grad_norm": 0.5282686352729797,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 11641
+    },
+    {
+      "epoch": 1.352541388324136,
+      "grad_norm": 0.5937612056732178,
+      "learning_rate": 0.0001,
+      "loss": 1.4411,
+      "step": 11642
+    },
+    {
+      "epoch": 1.3526575660760964,
+      "grad_norm": 0.6212623715400696,
+      "learning_rate": 0.0001,
+      "loss": 1.5084,
+      "step": 11643
+    },
+    {
+      "epoch": 1.352773743828057,
+      "grad_norm": 0.5802718997001648,
+      "learning_rate": 0.0001,
+      "loss": 1.3355,
+      "step": 11644
+    },
+    {
+      "epoch": 1.3528899215800174,
+      "grad_norm": 0.5913172364234924,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 11645
+    },
+    {
+      "epoch": 1.353006099331978,
+      "grad_norm": 0.5834310054779053,
+      "learning_rate": 0.0001,
+      "loss": 1.4286,
+      "step": 11646
+    },
+    {
+      "epoch": 1.3531222770839384,
+      "grad_norm": 0.5490809679031372,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 11647
+    },
+    {
+      "epoch": 1.3532384548358989,
+      "grad_norm": 0.5814580321311951,
+      "learning_rate": 0.0001,
+      "loss": 1.304,
+      "step": 11648
+    },
+    {
+      "epoch": 1.3533546325878594,
+      "grad_norm": 0.6068845391273499,
+      "learning_rate": 0.0001,
+      "loss": 1.4916,
+      "step": 11649
+    },
+    {
+      "epoch": 1.3534708103398199,
+      "grad_norm": 0.6316379308700562,
+      "learning_rate": 0.0001,
+      "loss": 1.2805,
+      "step": 11650
+    },
+    {
+      "epoch": 1.3535869880917804,
+      "grad_norm": 0.5487270951271057,
+      "learning_rate": 0.0001,
+      "loss": 1.536,
+      "step": 11651
+    },
+    {
+      "epoch": 1.3537031658437408,
+      "grad_norm": 0.6253399848937988,
+      "learning_rate": 0.0001,
+      "loss": 1.609,
+      "step": 11652
+    },
+    {
+      "epoch": 1.3538193435957013,
+      "grad_norm": 0.5893374085426331,
+      "learning_rate": 0.0001,
+      "loss": 1.5526,
+      "step": 11653
+    },
+    {
+      "epoch": 1.3539355213476618,
+      "grad_norm": 0.5669167637825012,
+      "learning_rate": 0.0001,
+      "loss": 1.51,
+      "step": 11654
+    },
+    {
+      "epoch": 1.3540516990996223,
+      "grad_norm": 0.601037323474884,
+      "learning_rate": 0.0001,
+      "loss": 1.3515,
+      "step": 11655
+    },
+    {
+      "epoch": 1.3541678768515828,
+      "grad_norm": 0.5730093717575073,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 11656
+    },
+    {
+      "epoch": 1.3542840546035433,
+      "grad_norm": 0.5829418301582336,
+      "learning_rate": 0.0001,
+      "loss": 1.486,
+      "step": 11657
+    },
+    {
+      "epoch": 1.354400232355504,
+      "grad_norm": 0.5747588872909546,
+      "learning_rate": 0.0001,
+      "loss": 1.5405,
+      "step": 11658
+    },
+    {
+      "epoch": 1.3545164101074645,
+      "grad_norm": 0.6172161102294922,
+      "learning_rate": 0.0001,
+      "loss": 1.5367,
+      "step": 11659
+    },
+    {
+      "epoch": 1.354632587859425,
+      "grad_norm": 0.5914645195007324,
+      "learning_rate": 0.0001,
+      "loss": 1.37,
+      "step": 11660
+    },
+    {
+      "epoch": 1.3547487656113855,
+      "grad_norm": 0.6524379253387451,
+      "learning_rate": 0.0001,
+      "loss": 1.5559,
+      "step": 11661
+    },
+    {
+      "epoch": 1.354864943363346,
+      "grad_norm": 0.588579535484314,
+      "learning_rate": 0.0001,
+      "loss": 1.38,
+      "step": 11662
+    },
+    {
+      "epoch": 1.3549811211153064,
+      "grad_norm": 0.5805348753929138,
+      "learning_rate": 0.0001,
+      "loss": 1.6023,
+      "step": 11663
+    },
+    {
+      "epoch": 1.355097298867267,
+      "grad_norm": 0.5591554641723633,
+      "learning_rate": 0.0001,
+      "loss": 1.4325,
+      "step": 11664
+    },
+    {
+      "epoch": 1.3552134766192274,
+      "grad_norm": 0.545655369758606,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 11665
+    },
+    {
+      "epoch": 1.355329654371188,
+      "grad_norm": 0.5984359383583069,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 11666
+    },
+    {
+      "epoch": 1.3554458321231484,
+      "grad_norm": 0.5707201361656189,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 11667
+    },
+    {
+      "epoch": 1.3555620098751089,
+      "grad_norm": 0.6246615648269653,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 11668
+    },
+    {
+      "epoch": 1.3556781876270694,
+      "grad_norm": 0.5760742425918579,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 11669
+    },
+    {
+      "epoch": 1.3557943653790299,
+      "grad_norm": 0.5490787029266357,
+      "learning_rate": 0.0001,
+      "loss": 1.2988,
+      "step": 11670
+    },
+    {
+      "epoch": 1.3559105431309904,
+      "grad_norm": 0.5468339920043945,
+      "learning_rate": 0.0001,
+      "loss": 1.2894,
+      "step": 11671
+    },
+    {
+      "epoch": 1.3560267208829508,
+      "grad_norm": 0.6399328708648682,
+      "learning_rate": 0.0001,
+      "loss": 1.4589,
+      "step": 11672
+    },
+    {
+      "epoch": 1.3561428986349113,
+      "grad_norm": 0.6269949078559875,
+      "learning_rate": 0.0001,
+      "loss": 1.6827,
+      "step": 11673
+    },
+    {
+      "epoch": 1.356259076386872,
+      "grad_norm": 0.5965935587882996,
+      "learning_rate": 0.0001,
+      "loss": 1.4121,
+      "step": 11674
+    },
+    {
+      "epoch": 1.3563752541388325,
+      "grad_norm": 0.586742639541626,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 11675
+    },
+    {
+      "epoch": 1.356491431890793,
+      "grad_norm": 0.5911440253257751,
+      "learning_rate": 0.0001,
+      "loss": 1.5497,
+      "step": 11676
+    },
+    {
+      "epoch": 1.3566076096427535,
+      "grad_norm": 0.6567511558532715,
+      "learning_rate": 0.0001,
+      "loss": 1.6401,
+      "step": 11677
+    },
+    {
+      "epoch": 1.356723787394714,
+      "grad_norm": 0.5498983263969421,
+      "learning_rate": 0.0001,
+      "loss": 1.269,
+      "step": 11678
+    },
+    {
+      "epoch": 1.3568399651466745,
+      "grad_norm": 0.5766180753707886,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 11679
+    },
+    {
+      "epoch": 1.356956142898635,
+      "grad_norm": 0.5323266386985779,
+      "learning_rate": 0.0001,
+      "loss": 1.2326,
+      "step": 11680
+    },
+    {
+      "epoch": 1.3570723206505955,
+      "grad_norm": 0.5497226119041443,
+      "learning_rate": 0.0001,
+      "loss": 1.5357,
+      "step": 11681
+    },
+    {
+      "epoch": 1.357188498402556,
+      "grad_norm": 0.5257942080497742,
+      "learning_rate": 0.0001,
+      "loss": 1.2964,
+      "step": 11682
+    },
+    {
+      "epoch": 1.3573046761545164,
+      "grad_norm": 0.6501601338386536,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 11683
+    },
+    {
+      "epoch": 1.357420853906477,
+      "grad_norm": 0.5762137174606323,
+      "learning_rate": 0.0001,
+      "loss": 1.3364,
+      "step": 11684
+    },
+    {
+      "epoch": 1.3575370316584374,
+      "grad_norm": 0.6714363694190979,
+      "learning_rate": 0.0001,
+      "loss": 1.5958,
+      "step": 11685
+    },
+    {
+      "epoch": 1.357653209410398,
+      "grad_norm": 0.614554762840271,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 11686
+    },
+    {
+      "epoch": 1.3577693871623584,
+      "grad_norm": 0.5594595670700073,
+      "learning_rate": 0.0001,
+      "loss": 1.3527,
+      "step": 11687
+    },
+    {
+      "epoch": 1.3578855649143189,
+      "grad_norm": 0.6071390509605408,
+      "learning_rate": 0.0001,
+      "loss": 1.4608,
+      "step": 11688
+    },
+    {
+      "epoch": 1.3580017426662794,
+      "grad_norm": 0.5696654319763184,
+      "learning_rate": 0.0001,
+      "loss": 1.5251,
+      "step": 11689
+    },
+    {
+      "epoch": 1.3581179204182399,
+      "grad_norm": 0.5189816355705261,
+      "learning_rate": 0.0001,
+      "loss": 1.2097,
+      "step": 11690
+    },
+    {
+      "epoch": 1.3582340981702004,
+      "grad_norm": 0.5831829905509949,
+      "learning_rate": 0.0001,
+      "loss": 1.5446,
+      "step": 11691
+    },
+    {
+      "epoch": 1.3583502759221608,
+      "grad_norm": 0.6040063500404358,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 11692
+    },
+    {
+      "epoch": 1.3584664536741213,
+      "grad_norm": 0.5630427598953247,
+      "learning_rate": 0.0001,
+      "loss": 1.384,
+      "step": 11693
+    },
+    {
+      "epoch": 1.3585826314260818,
+      "grad_norm": 0.6305066347122192,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 11694
+    },
+    {
+      "epoch": 1.3586988091780423,
+      "grad_norm": 0.614304780960083,
+      "learning_rate": 0.0001,
+      "loss": 1.4363,
+      "step": 11695
+    },
+    {
+      "epoch": 1.3588149869300028,
+      "grad_norm": 0.5854483842849731,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 11696
+    },
+    {
+      "epoch": 1.3589311646819633,
+      "grad_norm": 0.5789347290992737,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 11697
+    },
+    {
+      "epoch": 1.3590473424339238,
+      "grad_norm": 0.5860118269920349,
+      "learning_rate": 0.0001,
+      "loss": 1.4509,
+      "step": 11698
+    },
+    {
+      "epoch": 1.3591635201858843,
+      "grad_norm": 0.5637704730033875,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 11699
+    },
+    {
+      "epoch": 1.359279697937845,
+      "grad_norm": 0.6265964508056641,
+      "learning_rate": 0.0001,
+      "loss": 1.5091,
+      "step": 11700
+    },
+    {
+      "epoch": 1.3593958756898055,
+      "grad_norm": 0.6148953437805176,
+      "learning_rate": 0.0001,
+      "loss": 1.4862,
+      "step": 11701
+    },
+    {
+      "epoch": 1.359512053441766,
+      "grad_norm": 0.6008301973342896,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 11702
+    },
+    {
+      "epoch": 1.3596282311937264,
+      "grad_norm": 0.6362335681915283,
+      "learning_rate": 0.0001,
+      "loss": 1.4398,
+      "step": 11703
+    },
+    {
+      "epoch": 1.359744408945687,
+      "grad_norm": 0.6056876182556152,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 11704
+    },
+    {
+      "epoch": 1.3598605866976474,
+      "grad_norm": 0.625529944896698,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 11705
+    },
+    {
+      "epoch": 1.359976764449608,
+      "grad_norm": 0.6198188662528992,
+      "learning_rate": 0.0001,
+      "loss": 1.3204,
+      "step": 11706
+    },
+    {
+      "epoch": 1.3600929422015684,
+      "grad_norm": 0.540850818157196,
+      "learning_rate": 0.0001,
+      "loss": 1.3004,
+      "step": 11707
+    },
+    {
+      "epoch": 1.360209119953529,
+      "grad_norm": 0.6461174488067627,
+      "learning_rate": 0.0001,
+      "loss": 1.5201,
+      "step": 11708
+    },
+    {
+      "epoch": 1.3603252977054894,
+      "grad_norm": 0.5790330767631531,
+      "learning_rate": 0.0001,
+      "loss": 1.4575,
+      "step": 11709
+    },
+    {
+      "epoch": 1.3604414754574499,
+      "grad_norm": 0.6312645673751831,
+      "learning_rate": 0.0001,
+      "loss": 1.6026,
+      "step": 11710
+    },
+    {
+      "epoch": 1.3605576532094104,
+      "grad_norm": 0.5561741590499878,
+      "learning_rate": 0.0001,
+      "loss": 1.2689,
+      "step": 11711
+    },
+    {
+      "epoch": 1.3606738309613708,
+      "grad_norm": 0.5601104497909546,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 11712
+    },
+    {
+      "epoch": 1.3607900087133313,
+      "grad_norm": 0.6619279384613037,
+      "learning_rate": 0.0001,
+      "loss": 1.6191,
+      "step": 11713
+    },
+    {
+      "epoch": 1.3609061864652918,
+      "grad_norm": 0.6291702389717102,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 11714
+    },
+    {
+      "epoch": 1.3610223642172525,
+      "grad_norm": 0.6017669439315796,
+      "learning_rate": 0.0001,
+      "loss": 1.3569,
+      "step": 11715
+    },
+    {
+      "epoch": 1.361138541969213,
+      "grad_norm": 0.5894443988800049,
+      "learning_rate": 0.0001,
+      "loss": 1.5589,
+      "step": 11716
+    },
+    {
+      "epoch": 1.3612547197211735,
+      "grad_norm": 0.6237130165100098,
+      "learning_rate": 0.0001,
+      "loss": 1.5952,
+      "step": 11717
+    },
+    {
+      "epoch": 1.361370897473134,
+      "grad_norm": 0.5698156952857971,
+      "learning_rate": 0.0001,
+      "loss": 1.3084,
+      "step": 11718
+    },
+    {
+      "epoch": 1.3614870752250945,
+      "grad_norm": 0.5625483393669128,
+      "learning_rate": 0.0001,
+      "loss": 1.3964,
+      "step": 11719
+    },
+    {
+      "epoch": 1.361603252977055,
+      "grad_norm": 0.6011530756950378,
+      "learning_rate": 0.0001,
+      "loss": 1.5401,
+      "step": 11720
+    },
+    {
+      "epoch": 1.3617194307290155,
+      "grad_norm": 0.5694208741188049,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 11721
+    },
+    {
+      "epoch": 1.361835608480976,
+      "grad_norm": 0.6437340974807739,
+      "learning_rate": 0.0001,
+      "loss": 1.5627,
+      "step": 11722
+    },
+    {
+      "epoch": 1.3619517862329364,
+      "grad_norm": 0.5433085560798645,
+      "learning_rate": 0.0001,
+      "loss": 1.2774,
+      "step": 11723
+    },
+    {
+      "epoch": 1.362067963984897,
+      "grad_norm": 0.5646154880523682,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 11724
+    },
+    {
+      "epoch": 1.3621841417368574,
+      "grad_norm": 0.5664541125297546,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 11725
+    },
+    {
+      "epoch": 1.362300319488818,
+      "grad_norm": 0.536392867565155,
+      "learning_rate": 0.0001,
+      "loss": 1.4264,
+      "step": 11726
+    },
+    {
+      "epoch": 1.3624164972407784,
+      "grad_norm": 0.578348696231842,
+      "learning_rate": 0.0001,
+      "loss": 1.4304,
+      "step": 11727
+    },
+    {
+      "epoch": 1.362532674992739,
+      "grad_norm": 0.633996307849884,
+      "learning_rate": 0.0001,
+      "loss": 1.5121,
+      "step": 11728
+    },
+    {
+      "epoch": 1.3626488527446994,
+      "grad_norm": 0.5468116998672485,
+      "learning_rate": 0.0001,
+      "loss": 1.4625,
+      "step": 11729
+    },
+    {
+      "epoch": 1.3627650304966599,
+      "grad_norm": 0.5580371618270874,
+      "learning_rate": 0.0001,
+      "loss": 1.3462,
+      "step": 11730
+    },
+    {
+      "epoch": 1.3628812082486204,
+      "grad_norm": 0.5752867460250854,
+      "learning_rate": 0.0001,
+      "loss": 1.3912,
+      "step": 11731
+    },
+    {
+      "epoch": 1.3629973860005808,
+      "grad_norm": 0.626487135887146,
+      "learning_rate": 0.0001,
+      "loss": 1.5946,
+      "step": 11732
+    },
+    {
+      "epoch": 1.3631135637525413,
+      "grad_norm": 0.5772953033447266,
+      "learning_rate": 0.0001,
+      "loss": 1.4676,
+      "step": 11733
+    },
+    {
+      "epoch": 1.3632297415045018,
+      "grad_norm": 0.5664294362068176,
+      "learning_rate": 0.0001,
+      "loss": 1.3619,
+      "step": 11734
+    },
+    {
+      "epoch": 1.3633459192564623,
+      "grad_norm": 0.5935901999473572,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 11735
+    },
+    {
+      "epoch": 1.3634620970084228,
+      "grad_norm": 0.6294174194335938,
+      "learning_rate": 0.0001,
+      "loss": 1.6044,
+      "step": 11736
+    },
+    {
+      "epoch": 1.3635782747603833,
+      "grad_norm": 0.6048177480697632,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 11737
+    },
+    {
+      "epoch": 1.3636944525123438,
+      "grad_norm": 0.6142577528953552,
+      "learning_rate": 0.0001,
+      "loss": 1.4884,
+      "step": 11738
+    },
+    {
+      "epoch": 1.3638106302643043,
+      "grad_norm": 0.549647867679596,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 11739
+    },
+    {
+      "epoch": 1.3639268080162648,
+      "grad_norm": 0.6002432703971863,
+      "learning_rate": 0.0001,
+      "loss": 1.557,
+      "step": 11740
+    },
+    {
+      "epoch": 1.3640429857682252,
+      "grad_norm": 0.5926917791366577,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 11741
+    },
+    {
+      "epoch": 1.364159163520186,
+      "grad_norm": 0.5604010224342346,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 11742
+    },
+    {
+      "epoch": 1.3642753412721464,
+      "grad_norm": 0.5525916218757629,
+      "learning_rate": 0.0001,
+      "loss": 1.3864,
+      "step": 11743
+    },
+    {
+      "epoch": 1.364391519024107,
+      "grad_norm": 0.5697430968284607,
+      "learning_rate": 0.0001,
+      "loss": 1.3873,
+      "step": 11744
+    },
+    {
+      "epoch": 1.3645076967760674,
+      "grad_norm": 0.5901938080787659,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 11745
+    },
+    {
+      "epoch": 1.364623874528028,
+      "grad_norm": 0.6322489976882935,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 11746
+    },
+    {
+      "epoch": 1.3647400522799884,
+      "grad_norm": 0.563325822353363,
+      "learning_rate": 0.0001,
+      "loss": 1.4027,
+      "step": 11747
+    },
+    {
+      "epoch": 1.364856230031949,
+      "grad_norm": 0.5789667963981628,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 11748
+    },
+    {
+      "epoch": 1.3649724077839094,
+      "grad_norm": 0.58510422706604,
+      "learning_rate": 0.0001,
+      "loss": 1.4739,
+      "step": 11749
+    },
+    {
+      "epoch": 1.3650885855358699,
+      "grad_norm": 0.5673305988311768,
+      "learning_rate": 0.0001,
+      "loss": 1.3469,
+      "step": 11750
+    },
+    {
+      "epoch": 1.3652047632878304,
+      "grad_norm": 0.6077521443367004,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 11751
+    },
+    {
+      "epoch": 1.3653209410397908,
+      "grad_norm": 0.6002272963523865,
+      "learning_rate": 0.0001,
+      "loss": 1.6788,
+      "step": 11752
+    },
+    {
+      "epoch": 1.3654371187917513,
+      "grad_norm": 0.5982224345207214,
+      "learning_rate": 0.0001,
+      "loss": 1.4948,
+      "step": 11753
+    },
+    {
+      "epoch": 1.3655532965437118,
+      "grad_norm": 0.5681218504905701,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 11754
+    },
+    {
+      "epoch": 1.3656694742956723,
+      "grad_norm": 0.6169049739837646,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 11755
+    },
+    {
+      "epoch": 1.3657856520476328,
+      "grad_norm": 0.5828377604484558,
+      "learning_rate": 0.0001,
+      "loss": 1.4077,
+      "step": 11756
+    },
+    {
+      "epoch": 1.3659018297995935,
+      "grad_norm": 0.5715786814689636,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 11757
+    },
+    {
+      "epoch": 1.366018007551554,
+      "grad_norm": 0.5703264474868774,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 11758
+    },
+    {
+      "epoch": 1.3661341853035145,
+      "grad_norm": 0.6128681302070618,
+      "learning_rate": 0.0001,
+      "loss": 1.532,
+      "step": 11759
+    },
+    {
+      "epoch": 1.366250363055475,
+      "grad_norm": 0.5742857456207275,
+      "learning_rate": 0.0001,
+      "loss": 1.3648,
+      "step": 11760
+    },
+    {
+      "epoch": 1.3663665408074355,
+      "grad_norm": 0.5583314895629883,
+      "learning_rate": 0.0001,
+      "loss": 1.4086,
+      "step": 11761
+    },
+    {
+      "epoch": 1.366482718559396,
+      "grad_norm": 0.5679374933242798,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 11762
+    },
+    {
+      "epoch": 1.3665988963113564,
+      "grad_norm": 0.5741711258888245,
+      "learning_rate": 0.0001,
+      "loss": 1.4876,
+      "step": 11763
+    },
+    {
+      "epoch": 1.366715074063317,
+      "grad_norm": 0.5635258555412292,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 11764
+    },
+    {
+      "epoch": 1.3668312518152774,
+      "grad_norm": 0.5916835069656372,
+      "learning_rate": 0.0001,
+      "loss": 1.3363,
+      "step": 11765
+    },
+    {
+      "epoch": 1.366947429567238,
+      "grad_norm": 0.5715609192848206,
+      "learning_rate": 0.0001,
+      "loss": 1.3299,
+      "step": 11766
+    },
+    {
+      "epoch": 1.3670636073191984,
+      "grad_norm": 0.6283421516418457,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 11767
+    },
+    {
+      "epoch": 1.367179785071159,
+      "grad_norm": 0.557995617389679,
+      "learning_rate": 0.0001,
+      "loss": 1.3089,
+      "step": 11768
+    },
+    {
+      "epoch": 1.3672959628231194,
+      "grad_norm": 0.5753653049468994,
+      "learning_rate": 0.0001,
+      "loss": 1.4653,
+      "step": 11769
+    },
+    {
+      "epoch": 1.3674121405750799,
+      "grad_norm": 0.6110243797302246,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 11770
+    },
+    {
+      "epoch": 1.3675283183270404,
+      "grad_norm": 0.642266571521759,
+      "learning_rate": 0.0001,
+      "loss": 1.4771,
+      "step": 11771
+    },
+    {
+      "epoch": 1.3676444960790008,
+      "grad_norm": 0.5759993195533752,
+      "learning_rate": 0.0001,
+      "loss": 1.3839,
+      "step": 11772
+    },
+    {
+      "epoch": 1.3677606738309613,
+      "grad_norm": 0.5989052653312683,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 11773
+    },
+    {
+      "epoch": 1.3678768515829218,
+      "grad_norm": 0.5996981263160706,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 11774
+    },
+    {
+      "epoch": 1.3679930293348823,
+      "grad_norm": 0.5985528826713562,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 11775
+    },
+    {
+      "epoch": 1.3681092070868428,
+      "grad_norm": 0.5674298405647278,
+      "learning_rate": 0.0001,
+      "loss": 1.3278,
+      "step": 11776
+    },
+    {
+      "epoch": 1.3682253848388033,
+      "grad_norm": 0.6040525436401367,
+      "learning_rate": 0.0001,
+      "loss": 1.5492,
+      "step": 11777
+    },
+    {
+      "epoch": 1.3683415625907638,
+      "grad_norm": 0.5432946085929871,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 11778
+    },
+    {
+      "epoch": 1.3684577403427243,
+      "grad_norm": 0.5686826109886169,
+      "learning_rate": 0.0001,
+      "loss": 1.4459,
+      "step": 11779
+    },
+    {
+      "epoch": 1.3685739180946848,
+      "grad_norm": 0.5497419238090515,
+      "learning_rate": 0.0001,
+      "loss": 1.2312,
+      "step": 11780
+    },
+    {
+      "epoch": 1.3686900958466452,
+      "grad_norm": 0.6301116943359375,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 11781
+    },
+    {
+      "epoch": 1.3688062735986057,
+      "grad_norm": 0.562529981136322,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 11782
+    },
+    {
+      "epoch": 1.3689224513505664,
+      "grad_norm": 0.6351971626281738,
+      "learning_rate": 0.0001,
+      "loss": 1.584,
+      "step": 11783
+    },
+    {
+      "epoch": 1.369038629102527,
+      "grad_norm": 0.5856718420982361,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 11784
+    },
+    {
+      "epoch": 1.3691548068544874,
+      "grad_norm": 0.5878714323043823,
+      "learning_rate": 0.0001,
+      "loss": 1.5263,
+      "step": 11785
+    },
+    {
+      "epoch": 1.369270984606448,
+      "grad_norm": 0.5866144299507141,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 11786
+    },
+    {
+      "epoch": 1.3693871623584084,
+      "grad_norm": 0.5930546522140503,
+      "learning_rate": 0.0001,
+      "loss": 1.436,
+      "step": 11787
+    },
+    {
+      "epoch": 1.369503340110369,
+      "grad_norm": 0.5738999247550964,
+      "learning_rate": 0.0001,
+      "loss": 1.3887,
+      "step": 11788
+    },
+    {
+      "epoch": 1.3696195178623294,
+      "grad_norm": 0.6546767950057983,
+      "learning_rate": 0.0001,
+      "loss": 1.7115,
+      "step": 11789
+    },
+    {
+      "epoch": 1.3697356956142899,
+      "grad_norm": 0.5498051643371582,
+      "learning_rate": 0.0001,
+      "loss": 1.3995,
+      "step": 11790
+    },
+    {
+      "epoch": 1.3698518733662504,
+      "grad_norm": 0.5807487368583679,
+      "learning_rate": 0.0001,
+      "loss": 1.4775,
+      "step": 11791
+    },
+    {
+      "epoch": 1.3699680511182108,
+      "grad_norm": 0.582394003868103,
+      "learning_rate": 0.0001,
+      "loss": 1.4096,
+      "step": 11792
+    },
+    {
+      "epoch": 1.3700842288701713,
+      "grad_norm": 0.5830745100975037,
+      "learning_rate": 0.0001,
+      "loss": 1.4392,
+      "step": 11793
+    },
+    {
+      "epoch": 1.3702004066221318,
+      "grad_norm": 0.6170597672462463,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 11794
+    },
+    {
+      "epoch": 1.3703165843740923,
+      "grad_norm": 0.5893037915229797,
+      "learning_rate": 0.0001,
+      "loss": 1.4202,
+      "step": 11795
+    },
+    {
+      "epoch": 1.3704327621260528,
+      "grad_norm": 0.618791401386261,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 11796
+    },
+    {
+      "epoch": 1.3705489398780133,
+      "grad_norm": 0.6027114987373352,
+      "learning_rate": 0.0001,
+      "loss": 1.4953,
+      "step": 11797
+    },
+    {
+      "epoch": 1.3706651176299738,
+      "grad_norm": 0.5932108163833618,
+      "learning_rate": 0.0001,
+      "loss": 1.5312,
+      "step": 11798
+    },
+    {
+      "epoch": 1.3707812953819345,
+      "grad_norm": 0.6085308790206909,
+      "learning_rate": 0.0001,
+      "loss": 1.4098,
+      "step": 11799
+    },
+    {
+      "epoch": 1.370897473133895,
+      "grad_norm": 0.6194184422492981,
+      "learning_rate": 0.0001,
+      "loss": 1.4491,
+      "step": 11800
+    },
+    {
+      "epoch": 1.3710136508858555,
+      "grad_norm": 0.5471352934837341,
+      "learning_rate": 0.0001,
+      "loss": 1.2728,
+      "step": 11801
+    },
+    {
+      "epoch": 1.371129828637816,
+      "grad_norm": 0.5235902667045593,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 11802
+    },
+    {
+      "epoch": 1.3712460063897765,
+      "grad_norm": 0.6246318817138672,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 11803
+    },
+    {
+      "epoch": 1.371362184141737,
+      "grad_norm": 0.6138882637023926,
+      "learning_rate": 0.0001,
+      "loss": 1.2501,
+      "step": 11804
+    },
+    {
+      "epoch": 1.3714783618936974,
+      "grad_norm": 0.6174538731575012,
+      "learning_rate": 0.0001,
+      "loss": 1.34,
+      "step": 11805
+    },
+    {
+      "epoch": 1.371594539645658,
+      "grad_norm": 0.6327789425849915,
+      "learning_rate": 0.0001,
+      "loss": 1.5747,
+      "step": 11806
+    },
+    {
+      "epoch": 1.3717107173976184,
+      "grad_norm": 0.5819458365440369,
+      "learning_rate": 0.0001,
+      "loss": 1.4555,
+      "step": 11807
+    },
+    {
+      "epoch": 1.371826895149579,
+      "grad_norm": 0.5566579699516296,
+      "learning_rate": 0.0001,
+      "loss": 1.3598,
+      "step": 11808
+    },
+    {
+      "epoch": 1.3719430729015394,
+      "grad_norm": 0.608745813369751,
+      "learning_rate": 0.0001,
+      "loss": 1.7057,
+      "step": 11809
+    },
+    {
+      "epoch": 1.3720592506534999,
+      "grad_norm": 0.6066353917121887,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 11810
+    },
+    {
+      "epoch": 1.3721754284054604,
+      "grad_norm": 0.6526094079017639,
+      "learning_rate": 0.0001,
+      "loss": 1.6568,
+      "step": 11811
+    },
+    {
+      "epoch": 1.3722916061574209,
+      "grad_norm": 0.5477768778800964,
+      "learning_rate": 0.0001,
+      "loss": 1.4258,
+      "step": 11812
+    },
+    {
+      "epoch": 1.3724077839093813,
+      "grad_norm": 0.5935622453689575,
+      "learning_rate": 0.0001,
+      "loss": 1.4168,
+      "step": 11813
+    },
+    {
+      "epoch": 1.3725239616613418,
+      "grad_norm": 0.6308596730232239,
+      "learning_rate": 0.0001,
+      "loss": 1.3882,
+      "step": 11814
+    },
+    {
+      "epoch": 1.3726401394133023,
+      "grad_norm": 0.5651733875274658,
+      "learning_rate": 0.0001,
+      "loss": 1.2528,
+      "step": 11815
+    },
+    {
+      "epoch": 1.3727563171652628,
+      "grad_norm": 0.5938041806221008,
+      "learning_rate": 0.0001,
+      "loss": 1.4036,
+      "step": 11816
+    },
+    {
+      "epoch": 1.3728724949172233,
+      "grad_norm": 0.584787905216217,
+      "learning_rate": 0.0001,
+      "loss": 1.5447,
+      "step": 11817
+    },
+    {
+      "epoch": 1.3729886726691838,
+      "grad_norm": 0.5955517292022705,
+      "learning_rate": 0.0001,
+      "loss": 1.3351,
+      "step": 11818
+    },
+    {
+      "epoch": 1.3731048504211443,
+      "grad_norm": 0.595584511756897,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 11819
+    },
+    {
+      "epoch": 1.3732210281731048,
+      "grad_norm": 0.5970519781112671,
+      "learning_rate": 0.0001,
+      "loss": 1.4131,
+      "step": 11820
+    },
+    {
+      "epoch": 1.3733372059250653,
+      "grad_norm": 0.579200029373169,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 11821
+    },
+    {
+      "epoch": 1.3734533836770257,
+      "grad_norm": 0.6147415637969971,
+      "learning_rate": 0.0001,
+      "loss": 1.5784,
+      "step": 11822
+    },
+    {
+      "epoch": 1.3735695614289862,
+      "grad_norm": 0.6013969779014587,
+      "learning_rate": 0.0001,
+      "loss": 1.5465,
+      "step": 11823
+    },
+    {
+      "epoch": 1.3736857391809467,
+      "grad_norm": 0.5508737564086914,
+      "learning_rate": 0.0001,
+      "loss": 1.4251,
+      "step": 11824
+    },
+    {
+      "epoch": 1.3738019169329074,
+      "grad_norm": 0.5326734185218811,
+      "learning_rate": 0.0001,
+      "loss": 1.3177,
+      "step": 11825
+    },
+    {
+      "epoch": 1.373918094684868,
+      "grad_norm": 0.5824387073516846,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 11826
+    },
+    {
+      "epoch": 1.3740342724368284,
+      "grad_norm": 0.5780125856399536,
+      "learning_rate": 0.0001,
+      "loss": 1.3633,
+      "step": 11827
+    },
+    {
+      "epoch": 1.374150450188789,
+      "grad_norm": 0.56535804271698,
+      "learning_rate": 0.0001,
+      "loss": 1.5986,
+      "step": 11828
+    },
+    {
+      "epoch": 1.3742666279407494,
+      "grad_norm": 0.6138978600502014,
+      "learning_rate": 0.0001,
+      "loss": 1.5314,
+      "step": 11829
+    },
+    {
+      "epoch": 1.3743828056927099,
+      "grad_norm": 0.6302936673164368,
+      "learning_rate": 0.0001,
+      "loss": 1.5483,
+      "step": 11830
+    },
+    {
+      "epoch": 1.3744989834446704,
+      "grad_norm": 0.5425077080726624,
+      "learning_rate": 0.0001,
+      "loss": 1.2737,
+      "step": 11831
+    },
+    {
+      "epoch": 1.3746151611966309,
+      "grad_norm": 0.5782576203346252,
+      "learning_rate": 0.0001,
+      "loss": 1.3394,
+      "step": 11832
+    },
+    {
+      "epoch": 1.3747313389485913,
+      "grad_norm": 0.5924588441848755,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 11833
+    },
+    {
+      "epoch": 1.3748475167005518,
+      "grad_norm": 0.6358031034469604,
+      "learning_rate": 0.0001,
+      "loss": 1.5883,
+      "step": 11834
+    },
+    {
+      "epoch": 1.3749636944525123,
+      "grad_norm": 0.5872380137443542,
+      "learning_rate": 0.0001,
+      "loss": 1.3631,
+      "step": 11835
+    },
+    {
+      "epoch": 1.3750798722044728,
+      "grad_norm": 0.5860925316810608,
+      "learning_rate": 0.0001,
+      "loss": 1.5267,
+      "step": 11836
+    },
+    {
+      "epoch": 1.3751960499564333,
+      "grad_norm": 0.5498318076133728,
+      "learning_rate": 0.0001,
+      "loss": 1.4259,
+      "step": 11837
+    },
+    {
+      "epoch": 1.3753122277083938,
+      "grad_norm": 0.5745248198509216,
+      "learning_rate": 0.0001,
+      "loss": 1.5403,
+      "step": 11838
+    },
+    {
+      "epoch": 1.3754284054603543,
+      "grad_norm": 0.6191520690917969,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 11839
+    },
+    {
+      "epoch": 1.3755445832123148,
+      "grad_norm": 0.5659325122833252,
+      "learning_rate": 0.0001,
+      "loss": 1.3464,
+      "step": 11840
+    },
+    {
+      "epoch": 1.3756607609642755,
+      "grad_norm": 0.5831794738769531,
+      "learning_rate": 0.0001,
+      "loss": 1.3686,
+      "step": 11841
+    },
+    {
+      "epoch": 1.375776938716236,
+      "grad_norm": 0.6090126633644104,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 11842
+    },
+    {
+      "epoch": 1.3758931164681965,
+      "grad_norm": 0.5565717816352844,
+      "learning_rate": 0.0001,
+      "loss": 1.4553,
+      "step": 11843
+    },
+    {
+      "epoch": 1.376009294220157,
+      "grad_norm": 0.6339083909988403,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 11844
+    },
+    {
+      "epoch": 1.3761254719721174,
+      "grad_norm": 0.5814673900604248,
+      "learning_rate": 0.0001,
+      "loss": 1.3132,
+      "step": 11845
+    },
+    {
+      "epoch": 1.376241649724078,
+      "grad_norm": 0.5672376155853271,
+      "learning_rate": 0.0001,
+      "loss": 1.3554,
+      "step": 11846
+    },
+    {
+      "epoch": 1.3763578274760384,
+      "grad_norm": 0.6251681447029114,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 11847
+    },
+    {
+      "epoch": 1.376474005227999,
+      "grad_norm": 0.5934876203536987,
+      "learning_rate": 0.0001,
+      "loss": 1.5827,
+      "step": 11848
+    },
+    {
+      "epoch": 1.3765901829799594,
+      "grad_norm": 0.6062328219413757,
+      "learning_rate": 0.0001,
+      "loss": 1.5786,
+      "step": 11849
+    },
+    {
+      "epoch": 1.3767063607319199,
+      "grad_norm": 0.6141518354415894,
+      "learning_rate": 0.0001,
+      "loss": 1.7519,
+      "step": 11850
+    },
+    {
+      "epoch": 1.3768225384838804,
+      "grad_norm": 0.5661596059799194,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 11851
+    },
+    {
+      "epoch": 1.3769387162358409,
+      "grad_norm": 0.5913867354393005,
+      "learning_rate": 0.0001,
+      "loss": 1.523,
+      "step": 11852
+    },
+    {
+      "epoch": 1.3770548939878013,
+      "grad_norm": 0.6132528185844421,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 11853
+    },
+    {
+      "epoch": 1.3771710717397618,
+      "grad_norm": 0.5519713163375854,
+      "learning_rate": 0.0001,
+      "loss": 1.2775,
+      "step": 11854
+    },
+    {
+      "epoch": 1.3772872494917223,
+      "grad_norm": 0.586379885673523,
+      "learning_rate": 0.0001,
+      "loss": 1.3913,
+      "step": 11855
+    },
+    {
+      "epoch": 1.3774034272436828,
+      "grad_norm": 0.6136939525604248,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 11856
+    },
+    {
+      "epoch": 1.3775196049956433,
+      "grad_norm": 0.6094126105308533,
+      "learning_rate": 0.0001,
+      "loss": 1.4171,
+      "step": 11857
+    },
+    {
+      "epoch": 1.3776357827476038,
+      "grad_norm": 0.6217323541641235,
+      "learning_rate": 0.0001,
+      "loss": 1.4317,
+      "step": 11858
+    },
+    {
+      "epoch": 1.3777519604995643,
+      "grad_norm": 0.6349629163742065,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 11859
+    },
+    {
+      "epoch": 1.3778681382515248,
+      "grad_norm": 0.565229058265686,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 11860
+    },
+    {
+      "epoch": 1.3779843160034853,
+      "grad_norm": 0.6026855707168579,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 11861
+    },
+    {
+      "epoch": 1.3781004937554457,
+      "grad_norm": 0.5912261009216309,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 11862
+    },
+    {
+      "epoch": 1.3782166715074062,
+      "grad_norm": 0.5594475865364075,
+      "learning_rate": 0.0001,
+      "loss": 1.5227,
+      "step": 11863
+    },
+    {
+      "epoch": 1.3783328492593667,
+      "grad_norm": 0.5987929701805115,
+      "learning_rate": 0.0001,
+      "loss": 1.4032,
+      "step": 11864
+    },
+    {
+      "epoch": 1.3784490270113272,
+      "grad_norm": 0.5730629563331604,
+      "learning_rate": 0.0001,
+      "loss": 1.338,
+      "step": 11865
+    },
+    {
+      "epoch": 1.3785652047632877,
+      "grad_norm": 0.5603897571563721,
+      "learning_rate": 0.0001,
+      "loss": 1.4159,
+      "step": 11866
+    },
+    {
+      "epoch": 1.3786813825152484,
+      "grad_norm": 0.6142343878746033,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 11867
+    },
+    {
+      "epoch": 1.378797560267209,
+      "grad_norm": 0.6077200174331665,
+      "learning_rate": 0.0001,
+      "loss": 1.516,
+      "step": 11868
+    },
+    {
+      "epoch": 1.3789137380191694,
+      "grad_norm": 0.5924538373947144,
+      "learning_rate": 0.0001,
+      "loss": 1.5671,
+      "step": 11869
+    },
+    {
+      "epoch": 1.3790299157711299,
+      "grad_norm": 0.572593092918396,
+      "learning_rate": 0.0001,
+      "loss": 1.3483,
+      "step": 11870
+    },
+    {
+      "epoch": 1.3791460935230904,
+      "grad_norm": 0.5781865119934082,
+      "learning_rate": 0.0001,
+      "loss": 1.2929,
+      "step": 11871
+    },
+    {
+      "epoch": 1.3792622712750509,
+      "grad_norm": 0.6432904005050659,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 11872
+    },
+    {
+      "epoch": 1.3793784490270113,
+      "grad_norm": 0.5920875072479248,
+      "learning_rate": 0.0001,
+      "loss": 1.3776,
+      "step": 11873
+    },
+    {
+      "epoch": 1.3794946267789718,
+      "grad_norm": 0.5507972836494446,
+      "learning_rate": 0.0001,
+      "loss": 1.3918,
+      "step": 11874
+    },
+    {
+      "epoch": 1.3796108045309323,
+      "grad_norm": 0.5901596546173096,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 11875
+    },
+    {
+      "epoch": 1.3797269822828928,
+      "grad_norm": 0.6390935182571411,
+      "learning_rate": 0.0001,
+      "loss": 1.6274,
+      "step": 11876
+    },
+    {
+      "epoch": 1.3798431600348533,
+      "grad_norm": 0.6109213829040527,
+      "learning_rate": 0.0001,
+      "loss": 1.5234,
+      "step": 11877
+    },
+    {
+      "epoch": 1.3799593377868138,
+      "grad_norm": 0.5569701790809631,
+      "learning_rate": 0.0001,
+      "loss": 1.3981,
+      "step": 11878
+    },
+    {
+      "epoch": 1.3800755155387743,
+      "grad_norm": 0.5873339772224426,
+      "learning_rate": 0.0001,
+      "loss": 1.4033,
+      "step": 11879
+    },
+    {
+      "epoch": 1.3801916932907348,
+      "grad_norm": 0.6170241236686707,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 11880
+    },
+    {
+      "epoch": 1.3803078710426953,
+      "grad_norm": 0.5810169577598572,
+      "learning_rate": 0.0001,
+      "loss": 1.3362,
+      "step": 11881
+    },
+    {
+      "epoch": 1.3804240487946557,
+      "grad_norm": 0.6244444251060486,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 11882
+    },
+    {
+      "epoch": 1.3805402265466165,
+      "grad_norm": 0.5936790704727173,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 11883
+    },
+    {
+      "epoch": 1.380656404298577,
+      "grad_norm": 0.5871090888977051,
+      "learning_rate": 0.0001,
+      "loss": 1.4629,
+      "step": 11884
+    },
+    {
+      "epoch": 1.3807725820505374,
+      "grad_norm": 0.6491077542304993,
+      "learning_rate": 0.0001,
+      "loss": 1.3812,
+      "step": 11885
+    },
+    {
+      "epoch": 1.380888759802498,
+      "grad_norm": 0.5600960850715637,
+      "learning_rate": 0.0001,
+      "loss": 1.2137,
+      "step": 11886
+    },
+    {
+      "epoch": 1.3810049375544584,
+      "grad_norm": 0.6018544435501099,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 11887
+    },
+    {
+      "epoch": 1.381121115306419,
+      "grad_norm": 0.58489990234375,
+      "learning_rate": 0.0001,
+      "loss": 1.3715,
+      "step": 11888
+    },
+    {
+      "epoch": 1.3812372930583794,
+      "grad_norm": 0.5836353302001953,
+      "learning_rate": 0.0001,
+      "loss": 1.2618,
+      "step": 11889
+    },
+    {
+      "epoch": 1.3813534708103399,
+      "grad_norm": 0.6115884184837341,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 11890
+    },
+    {
+      "epoch": 1.3814696485623004,
+      "grad_norm": 0.6390863060951233,
+      "learning_rate": 0.0001,
+      "loss": 1.4521,
+      "step": 11891
+    },
+    {
+      "epoch": 1.3815858263142609,
+      "grad_norm": 0.5852669477462769,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 11892
+    },
+    {
+      "epoch": 1.3817020040662213,
+      "grad_norm": 0.6284686923027039,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 11893
+    },
+    {
+      "epoch": 1.3818181818181818,
+      "grad_norm": 0.6020316481590271,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 11894
+    },
+    {
+      "epoch": 1.3819343595701423,
+      "grad_norm": 0.650677502155304,
+      "learning_rate": 0.0001,
+      "loss": 1.7152,
+      "step": 11895
+    },
+    {
+      "epoch": 1.3820505373221028,
+      "grad_norm": 0.5915738940238953,
+      "learning_rate": 0.0001,
+      "loss": 1.4075,
+      "step": 11896
+    },
+    {
+      "epoch": 1.3821667150740633,
+      "grad_norm": 0.6179349422454834,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 11897
+    },
+    {
+      "epoch": 1.3822828928260238,
+      "grad_norm": 0.6398385167121887,
+      "learning_rate": 0.0001,
+      "loss": 1.2884,
+      "step": 11898
+    },
+    {
+      "epoch": 1.3823990705779843,
+      "grad_norm": 0.6043710112571716,
+      "learning_rate": 0.0001,
+      "loss": 1.4017,
+      "step": 11899
+    },
+    {
+      "epoch": 1.3825152483299448,
+      "grad_norm": 0.5860773921012878,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 11900
+    },
+    {
+      "epoch": 1.3826314260819053,
+      "grad_norm": 0.6333348751068115,
+      "learning_rate": 0.0001,
+      "loss": 1.3162,
+      "step": 11901
+    },
+    {
+      "epoch": 1.3827476038338657,
+      "grad_norm": 0.6222668290138245,
+      "learning_rate": 0.0001,
+      "loss": 1.4695,
+      "step": 11902
+    },
+    {
+      "epoch": 1.3828637815858262,
+      "grad_norm": 0.5919185280799866,
+      "learning_rate": 0.0001,
+      "loss": 1.3576,
+      "step": 11903
+    },
+    {
+      "epoch": 1.3829799593377867,
+      "grad_norm": 0.5540260076522827,
+      "learning_rate": 0.0001,
+      "loss": 1.2607,
+      "step": 11904
+    },
+    {
+      "epoch": 1.3830961370897472,
+      "grad_norm": 0.6326545476913452,
+      "learning_rate": 0.0001,
+      "loss": 1.5169,
+      "step": 11905
+    },
+    {
+      "epoch": 1.3832123148417077,
+      "grad_norm": 0.5651580691337585,
+      "learning_rate": 0.0001,
+      "loss": 1.3546,
+      "step": 11906
+    },
+    {
+      "epoch": 1.3833284925936682,
+      "grad_norm": 0.5845093131065369,
+      "learning_rate": 0.0001,
+      "loss": 1.7022,
+      "step": 11907
+    },
+    {
+      "epoch": 1.3834446703456287,
+      "grad_norm": 0.6123130321502686,
+      "learning_rate": 0.0001,
+      "loss": 1.672,
+      "step": 11908
+    },
+    {
+      "epoch": 1.3835608480975894,
+      "grad_norm": 0.5754287242889404,
+      "learning_rate": 0.0001,
+      "loss": 1.4662,
+      "step": 11909
+    },
+    {
+      "epoch": 1.3836770258495499,
+      "grad_norm": 0.6103564500808716,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 11910
+    },
+    {
+      "epoch": 1.3837932036015104,
+      "grad_norm": 0.6126376390457153,
+      "learning_rate": 0.0001,
+      "loss": 1.4635,
+      "step": 11911
+    },
+    {
+      "epoch": 1.3839093813534709,
+      "grad_norm": 0.5791416168212891,
+      "learning_rate": 0.0001,
+      "loss": 1.4217,
+      "step": 11912
+    },
+    {
+      "epoch": 1.3840255591054313,
+      "grad_norm": 0.5979540944099426,
+      "learning_rate": 0.0001,
+      "loss": 1.3736,
+      "step": 11913
+    },
+    {
+      "epoch": 1.3841417368573918,
+      "grad_norm": 0.5859628319740295,
+      "learning_rate": 0.0001,
+      "loss": 1.2924,
+      "step": 11914
+    },
+    {
+      "epoch": 1.3842579146093523,
+      "grad_norm": 0.6160323619842529,
+      "learning_rate": 0.0001,
+      "loss": 1.4061,
+      "step": 11915
+    },
+    {
+      "epoch": 1.3843740923613128,
+      "grad_norm": 0.6810671091079712,
+      "learning_rate": 0.0001,
+      "loss": 1.3882,
+      "step": 11916
+    },
+    {
+      "epoch": 1.3844902701132733,
+      "grad_norm": 0.6706360578536987,
+      "learning_rate": 0.0001,
+      "loss": 1.5949,
+      "step": 11917
+    },
+    {
+      "epoch": 1.3846064478652338,
+      "grad_norm": 0.5953517556190491,
+      "learning_rate": 0.0001,
+      "loss": 1.3044,
+      "step": 11918
+    },
+    {
+      "epoch": 1.3847226256171943,
+      "grad_norm": 0.5882486701011658,
+      "learning_rate": 0.0001,
+      "loss": 1.4569,
+      "step": 11919
+    },
+    {
+      "epoch": 1.3848388033691548,
+      "grad_norm": 0.6317164301872253,
+      "learning_rate": 0.0001,
+      "loss": 1.5054,
+      "step": 11920
+    },
+    {
+      "epoch": 1.3849549811211153,
+      "grad_norm": 0.6241341233253479,
+      "learning_rate": 0.0001,
+      "loss": 1.5809,
+      "step": 11921
+    },
+    {
+      "epoch": 1.3850711588730757,
+      "grad_norm": 0.6015408635139465,
+      "learning_rate": 0.0001,
+      "loss": 1.4114,
+      "step": 11922
+    },
+    {
+      "epoch": 1.3851873366250362,
+      "grad_norm": 0.5517604351043701,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 11923
+    },
+    {
+      "epoch": 1.3853035143769967,
+      "grad_norm": 0.6593002080917358,
+      "learning_rate": 0.0001,
+      "loss": 1.6226,
+      "step": 11924
+    },
+    {
+      "epoch": 1.3854196921289574,
+      "grad_norm": 0.5856340527534485,
+      "learning_rate": 0.0001,
+      "loss": 1.5272,
+      "step": 11925
+    },
+    {
+      "epoch": 1.385535869880918,
+      "grad_norm": 0.6336721181869507,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 11926
+    },
+    {
+      "epoch": 1.3856520476328784,
+      "grad_norm": 0.5610465407371521,
+      "learning_rate": 0.0001,
+      "loss": 1.6593,
+      "step": 11927
+    },
+    {
+      "epoch": 1.385768225384839,
+      "grad_norm": 0.6065878868103027,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 11928
+    },
+    {
+      "epoch": 1.3858844031367994,
+      "grad_norm": 0.5784077048301697,
+      "learning_rate": 0.0001,
+      "loss": 1.4109,
+      "step": 11929
+    },
+    {
+      "epoch": 1.3860005808887599,
+      "grad_norm": 0.5930120348930359,
+      "learning_rate": 0.0001,
+      "loss": 1.2691,
+      "step": 11930
+    },
+    {
+      "epoch": 1.3861167586407204,
+      "grad_norm": 0.599595308303833,
+      "learning_rate": 0.0001,
+      "loss": 1.3383,
+      "step": 11931
+    },
+    {
+      "epoch": 1.3862329363926809,
+      "grad_norm": 0.6132897734642029,
+      "learning_rate": 0.0001,
+      "loss": 1.5586,
+      "step": 11932
+    },
+    {
+      "epoch": 1.3863491141446413,
+      "grad_norm": 0.5871015191078186,
+      "learning_rate": 0.0001,
+      "loss": 1.3863,
+      "step": 11933
+    },
+    {
+      "epoch": 1.3864652918966018,
+      "grad_norm": 0.588950514793396,
+      "learning_rate": 0.0001,
+      "loss": 1.3364,
+      "step": 11934
+    },
+    {
+      "epoch": 1.3865814696485623,
+      "grad_norm": 0.5841085314750671,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 11935
+    },
+    {
+      "epoch": 1.3866976474005228,
+      "grad_norm": 0.5449914932250977,
+      "learning_rate": 0.0001,
+      "loss": 1.3709,
+      "step": 11936
+    },
+    {
+      "epoch": 1.3868138251524833,
+      "grad_norm": 0.6109606623649597,
+      "learning_rate": 0.0001,
+      "loss": 1.3732,
+      "step": 11937
+    },
+    {
+      "epoch": 1.3869300029044438,
+      "grad_norm": 0.6314137578010559,
+      "learning_rate": 0.0001,
+      "loss": 1.3493,
+      "step": 11938
+    },
+    {
+      "epoch": 1.3870461806564043,
+      "grad_norm": 0.6377010345458984,
+      "learning_rate": 0.0001,
+      "loss": 1.3789,
+      "step": 11939
+    },
+    {
+      "epoch": 1.3871623584083648,
+      "grad_norm": 0.6468605399131775,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 11940
+    },
+    {
+      "epoch": 1.3872785361603253,
+      "grad_norm": 0.6271966099739075,
+      "learning_rate": 0.0001,
+      "loss": 1.5676,
+      "step": 11941
+    },
+    {
+      "epoch": 1.3873947139122857,
+      "grad_norm": 0.5833338499069214,
+      "learning_rate": 0.0001,
+      "loss": 1.5451,
+      "step": 11942
+    },
+    {
+      "epoch": 1.3875108916642462,
+      "grad_norm": 0.5465646386146545,
+      "learning_rate": 0.0001,
+      "loss": 1.2402,
+      "step": 11943
+    },
+    {
+      "epoch": 1.3876270694162067,
+      "grad_norm": 0.5553290247917175,
+      "learning_rate": 0.0001,
+      "loss": 1.3301,
+      "step": 11944
+    },
+    {
+      "epoch": 1.3877432471681672,
+      "grad_norm": 0.6450720429420471,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 11945
+    },
+    {
+      "epoch": 1.3878594249201277,
+      "grad_norm": 0.6172477602958679,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 11946
+    },
+    {
+      "epoch": 1.3879756026720882,
+      "grad_norm": 0.6228959560394287,
+      "learning_rate": 0.0001,
+      "loss": 1.5618,
+      "step": 11947
+    },
+    {
+      "epoch": 1.3880917804240487,
+      "grad_norm": 0.6162193417549133,
+      "learning_rate": 0.0001,
+      "loss": 1.4889,
+      "step": 11948
+    },
+    {
+      "epoch": 1.3882079581760092,
+      "grad_norm": 0.6280121803283691,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 11949
+    },
+    {
+      "epoch": 1.3883241359279697,
+      "grad_norm": 0.5984324812889099,
+      "learning_rate": 0.0001,
+      "loss": 1.6882,
+      "step": 11950
+    },
+    {
+      "epoch": 1.3884403136799304,
+      "grad_norm": 0.5944661498069763,
+      "learning_rate": 0.0001,
+      "loss": 1.4478,
+      "step": 11951
+    },
+    {
+      "epoch": 1.3885564914318909,
+      "grad_norm": 0.6321014165878296,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 11952
+    },
+    {
+      "epoch": 1.3886726691838513,
+      "grad_norm": 0.6064655780792236,
+      "learning_rate": 0.0001,
+      "loss": 1.5876,
+      "step": 11953
+    },
+    {
+      "epoch": 1.3887888469358118,
+      "grad_norm": 0.5939200520515442,
+      "learning_rate": 0.0001,
+      "loss": 1.443,
+      "step": 11954
+    },
+    {
+      "epoch": 1.3889050246877723,
+      "grad_norm": 0.5834383368492126,
+      "learning_rate": 0.0001,
+      "loss": 1.3573,
+      "step": 11955
+    },
+    {
+      "epoch": 1.3890212024397328,
+      "grad_norm": 0.5999382138252258,
+      "learning_rate": 0.0001,
+      "loss": 1.3501,
+      "step": 11956
+    },
+    {
+      "epoch": 1.3891373801916933,
+      "grad_norm": 0.652275562286377,
+      "learning_rate": 0.0001,
+      "loss": 1.4214,
+      "step": 11957
+    },
+    {
+      "epoch": 1.3892535579436538,
+      "grad_norm": 0.6107578277587891,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 11958
+    },
+    {
+      "epoch": 1.3893697356956143,
+      "grad_norm": 0.6266455054283142,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 11959
+    },
+    {
+      "epoch": 1.3894859134475748,
+      "grad_norm": 0.5955777168273926,
+      "learning_rate": 0.0001,
+      "loss": 1.4787,
+      "step": 11960
+    },
+    {
+      "epoch": 1.3896020911995353,
+      "grad_norm": 0.6200416684150696,
+      "learning_rate": 0.0001,
+      "loss": 1.4147,
+      "step": 11961
+    },
+    {
+      "epoch": 1.3897182689514957,
+      "grad_norm": 0.6102170348167419,
+      "learning_rate": 0.0001,
+      "loss": 1.5664,
+      "step": 11962
+    },
+    {
+      "epoch": 1.3898344467034562,
+      "grad_norm": 0.5599600076675415,
+      "learning_rate": 0.0001,
+      "loss": 1.432,
+      "step": 11963
+    },
+    {
+      "epoch": 1.3899506244554167,
+      "grad_norm": 0.6110653281211853,
+      "learning_rate": 0.0001,
+      "loss": 1.5714,
+      "step": 11964
+    },
+    {
+      "epoch": 1.3900668022073772,
+      "grad_norm": 0.5802760720252991,
+      "learning_rate": 0.0001,
+      "loss": 1.488,
+      "step": 11965
+    },
+    {
+      "epoch": 1.3901829799593377,
+      "grad_norm": 0.5832214951515198,
+      "learning_rate": 0.0001,
+      "loss": 1.6663,
+      "step": 11966
+    },
+    {
+      "epoch": 1.3902991577112984,
+      "grad_norm": 0.5920937657356262,
+      "learning_rate": 0.0001,
+      "loss": 1.3629,
+      "step": 11967
+    },
+    {
+      "epoch": 1.390415335463259,
+      "grad_norm": 0.6126205921173096,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 11968
+    },
+    {
+      "epoch": 1.3905315132152194,
+      "grad_norm": 0.6226344108581543,
+      "learning_rate": 0.0001,
+      "loss": 1.6481,
+      "step": 11969
+    },
+    {
+      "epoch": 1.3906476909671799,
+      "grad_norm": 0.5618362426757812,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 11970
+    },
+    {
+      "epoch": 1.3907638687191404,
+      "grad_norm": 0.5985326766967773,
+      "learning_rate": 0.0001,
+      "loss": 1.5428,
+      "step": 11971
+    },
+    {
+      "epoch": 1.3908800464711009,
+      "grad_norm": 0.611813485622406,
+      "learning_rate": 0.0001,
+      "loss": 1.5763,
+      "step": 11972
+    },
+    {
+      "epoch": 1.3909962242230614,
+      "grad_norm": 0.5800209641456604,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 11973
+    },
+    {
+      "epoch": 1.3911124019750218,
+      "grad_norm": 0.6421018838882446,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 11974
+    },
+    {
+      "epoch": 1.3912285797269823,
+      "grad_norm": 0.5720347166061401,
+      "learning_rate": 0.0001,
+      "loss": 1.4541,
+      "step": 11975
+    },
+    {
+      "epoch": 1.3913447574789428,
+      "grad_norm": 0.5714155435562134,
+      "learning_rate": 0.0001,
+      "loss": 1.3498,
+      "step": 11976
+    },
+    {
+      "epoch": 1.3914609352309033,
+      "grad_norm": 0.5725839734077454,
+      "learning_rate": 0.0001,
+      "loss": 1.479,
+      "step": 11977
+    },
+    {
+      "epoch": 1.3915771129828638,
+      "grad_norm": 0.5719639658927917,
+      "learning_rate": 0.0001,
+      "loss": 1.5502,
+      "step": 11978
+    },
+    {
+      "epoch": 1.3916932907348243,
+      "grad_norm": 0.5681836605072021,
+      "learning_rate": 0.0001,
+      "loss": 1.5435,
+      "step": 11979
+    },
+    {
+      "epoch": 1.3918094684867848,
+      "grad_norm": 0.5832224488258362,
+      "learning_rate": 0.0001,
+      "loss": 1.5036,
+      "step": 11980
+    },
+    {
+      "epoch": 1.3919256462387453,
+      "grad_norm": 0.6468372941017151,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 11981
+    },
+    {
+      "epoch": 1.3920418239907058,
+      "grad_norm": 0.5739614367485046,
+      "learning_rate": 0.0001,
+      "loss": 1.5003,
+      "step": 11982
+    },
+    {
+      "epoch": 1.3921580017426662,
+      "grad_norm": 0.6735573410987854,
+      "learning_rate": 0.0001,
+      "loss": 1.6389,
+      "step": 11983
+    },
+    {
+      "epoch": 1.3922741794946267,
+      "grad_norm": 0.6430903077125549,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 11984
+    },
+    {
+      "epoch": 1.3923903572465872,
+      "grad_norm": 0.5974782705307007,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 11985
+    },
+    {
+      "epoch": 1.3925065349985477,
+      "grad_norm": 0.6093497276306152,
+      "learning_rate": 0.0001,
+      "loss": 1.4616,
+      "step": 11986
+    },
+    {
+      "epoch": 1.3926227127505082,
+      "grad_norm": 0.5952013731002808,
+      "learning_rate": 0.0001,
+      "loss": 1.481,
+      "step": 11987
+    },
+    {
+      "epoch": 1.3927388905024687,
+      "grad_norm": 0.577219545841217,
+      "learning_rate": 0.0001,
+      "loss": 1.5508,
+      "step": 11988
+    },
+    {
+      "epoch": 1.3928550682544292,
+      "grad_norm": 0.5776684284210205,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 11989
+    },
+    {
+      "epoch": 1.3929712460063897,
+      "grad_norm": 0.580201268196106,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 11990
+    },
+    {
+      "epoch": 1.3930874237583502,
+      "grad_norm": 0.5810071229934692,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 11991
+    },
+    {
+      "epoch": 1.3932036015103106,
+      "grad_norm": 0.5787037014961243,
+      "learning_rate": 0.0001,
+      "loss": 1.3579,
+      "step": 11992
+    },
+    {
+      "epoch": 1.3933197792622714,
+      "grad_norm": 0.5724977850914001,
+      "learning_rate": 0.0001,
+      "loss": 1.4215,
+      "step": 11993
+    },
+    {
+      "epoch": 1.3934359570142318,
+      "grad_norm": 0.5736649632453918,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 11994
+    },
+    {
+      "epoch": 1.3935521347661923,
+      "grad_norm": 0.6170047521591187,
+      "learning_rate": 0.0001,
+      "loss": 1.3581,
+      "step": 11995
+    },
+    {
+      "epoch": 1.3936683125181528,
+      "grad_norm": 0.5980851650238037,
+      "learning_rate": 0.0001,
+      "loss": 1.5077,
+      "step": 11996
+    },
+    {
+      "epoch": 1.3937844902701133,
+      "grad_norm": 0.5532777309417725,
+      "learning_rate": 0.0001,
+      "loss": 1.3881,
+      "step": 11997
+    },
+    {
+      "epoch": 1.3939006680220738,
+      "grad_norm": 0.6409569382667542,
+      "learning_rate": 0.0001,
+      "loss": 1.5328,
+      "step": 11998
+    },
+    {
+      "epoch": 1.3940168457740343,
+      "grad_norm": 0.564768373966217,
+      "learning_rate": 0.0001,
+      "loss": 1.3573,
+      "step": 11999
+    },
+    {
+      "epoch": 1.3941330235259948,
+      "grad_norm": 0.6319162845611572,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 12000
+    },
+    {
+      "epoch": 1.3942492012779553,
+      "grad_norm": 0.537635862827301,
+      "learning_rate": 0.0001,
+      "loss": 1.2859,
+      "step": 12001
+    },
+    {
+      "epoch": 1.3943653790299158,
+      "grad_norm": 0.6129380464553833,
+      "learning_rate": 0.0001,
+      "loss": 1.4286,
+      "step": 12002
+    },
+    {
+      "epoch": 1.3944815567818762,
+      "grad_norm": 0.5544732809066772,
+      "learning_rate": 0.0001,
+      "loss": 1.328,
+      "step": 12003
+    },
+    {
+      "epoch": 1.3945977345338367,
+      "grad_norm": 0.5666175484657288,
+      "learning_rate": 0.0001,
+      "loss": 1.448,
+      "step": 12004
+    },
+    {
+      "epoch": 1.3947139122857972,
+      "grad_norm": 0.5639254450798035,
+      "learning_rate": 0.0001,
+      "loss": 1.2761,
+      "step": 12005
+    },
+    {
+      "epoch": 1.3948300900377577,
+      "grad_norm": 0.6033602952957153,
+      "learning_rate": 0.0001,
+      "loss": 1.6798,
+      "step": 12006
+    },
+    {
+      "epoch": 1.3949462677897182,
+      "grad_norm": 0.6276678442955017,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 12007
+    },
+    {
+      "epoch": 1.3950624455416787,
+      "grad_norm": 0.5830793976783752,
+      "learning_rate": 0.0001,
+      "loss": 1.517,
+      "step": 12008
+    },
+    {
+      "epoch": 1.3951786232936394,
+      "grad_norm": 0.5950773358345032,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 12009
+    },
+    {
+      "epoch": 1.3952948010455999,
+      "grad_norm": 0.6116399168968201,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 12010
+    },
+    {
+      "epoch": 1.3954109787975604,
+      "grad_norm": 0.5737411975860596,
+      "learning_rate": 0.0001,
+      "loss": 1.4567,
+      "step": 12011
+    },
+    {
+      "epoch": 1.3955271565495209,
+      "grad_norm": 0.5670390129089355,
+      "learning_rate": 0.0001,
+      "loss": 1.3832,
+      "step": 12012
+    },
+    {
+      "epoch": 1.3956433343014814,
+      "grad_norm": 0.5643200278282166,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 12013
+    },
+    {
+      "epoch": 1.3957595120534418,
+      "grad_norm": 0.588569700717926,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 12014
+    },
+    {
+      "epoch": 1.3958756898054023,
+      "grad_norm": 0.5916202664375305,
+      "learning_rate": 0.0001,
+      "loss": 1.3856,
+      "step": 12015
+    },
+    {
+      "epoch": 1.3959918675573628,
+      "grad_norm": 0.57044517993927,
+      "learning_rate": 0.0001,
+      "loss": 1.4782,
+      "step": 12016
+    },
+    {
+      "epoch": 1.3961080453093233,
+      "grad_norm": 0.5924696326255798,
+      "learning_rate": 0.0001,
+      "loss": 1.4195,
+      "step": 12017
+    },
+    {
+      "epoch": 1.3962242230612838,
+      "grad_norm": 0.5659184455871582,
+      "learning_rate": 0.0001,
+      "loss": 1.3907,
+      "step": 12018
+    },
+    {
+      "epoch": 1.3963404008132443,
+      "grad_norm": 0.6118780970573425,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 12019
+    },
+    {
+      "epoch": 1.3964565785652048,
+      "grad_norm": 0.569577693939209,
+      "learning_rate": 0.0001,
+      "loss": 1.3672,
+      "step": 12020
+    },
+    {
+      "epoch": 1.3965727563171653,
+      "grad_norm": 0.5691976547241211,
+      "learning_rate": 0.0001,
+      "loss": 1.3072,
+      "step": 12021
+    },
+    {
+      "epoch": 1.3966889340691258,
+      "grad_norm": 0.6030403971672058,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 12022
+    },
+    {
+      "epoch": 1.3968051118210862,
+      "grad_norm": 0.560248851776123,
+      "learning_rate": 0.0001,
+      "loss": 1.2186,
+      "step": 12023
+    },
+    {
+      "epoch": 1.3969212895730467,
+      "grad_norm": 0.5866391658782959,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 12024
+    },
+    {
+      "epoch": 1.3970374673250072,
+      "grad_norm": 0.624195396900177,
+      "learning_rate": 0.0001,
+      "loss": 1.6459,
+      "step": 12025
+    },
+    {
+      "epoch": 1.3971536450769677,
+      "grad_norm": 0.6529926657676697,
+      "learning_rate": 0.0001,
+      "loss": 1.6331,
+      "step": 12026
+    },
+    {
+      "epoch": 1.3972698228289282,
+      "grad_norm": 0.6984851360321045,
+      "learning_rate": 0.0001,
+      "loss": 1.6492,
+      "step": 12027
+    },
+    {
+      "epoch": 1.3973860005808887,
+      "grad_norm": 0.6242703795433044,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 12028
+    },
+    {
+      "epoch": 1.3975021783328492,
+      "grad_norm": 0.5590314865112305,
+      "learning_rate": 0.0001,
+      "loss": 1.3682,
+      "step": 12029
+    },
+    {
+      "epoch": 1.3976183560848097,
+      "grad_norm": 0.6193909049034119,
+      "learning_rate": 0.0001,
+      "loss": 1.5752,
+      "step": 12030
+    },
+    {
+      "epoch": 1.3977345338367702,
+      "grad_norm": 0.5770513415336609,
+      "learning_rate": 0.0001,
+      "loss": 1.563,
+      "step": 12031
+    },
+    {
+      "epoch": 1.3978507115887306,
+      "grad_norm": 0.5817797780036926,
+      "learning_rate": 0.0001,
+      "loss": 1.4469,
+      "step": 12032
+    },
+    {
+      "epoch": 1.3979668893406911,
+      "grad_norm": 0.5368718504905701,
+      "learning_rate": 0.0001,
+      "loss": 1.5278,
+      "step": 12033
+    },
+    {
+      "epoch": 1.3980830670926516,
+      "grad_norm": 0.6158215403556824,
+      "learning_rate": 0.0001,
+      "loss": 1.5191,
+      "step": 12034
+    },
+    {
+      "epoch": 1.3981992448446123,
+      "grad_norm": 0.5757713317871094,
+      "learning_rate": 0.0001,
+      "loss": 1.3353,
+      "step": 12035
+    },
+    {
+      "epoch": 1.3983154225965728,
+      "grad_norm": 0.5898905992507935,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 12036
+    },
+    {
+      "epoch": 1.3984316003485333,
+      "grad_norm": 0.5727129578590393,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 12037
+    },
+    {
+      "epoch": 1.3985477781004938,
+      "grad_norm": 0.573549747467041,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 12038
+    },
+    {
+      "epoch": 1.3986639558524543,
+      "grad_norm": 0.5966438055038452,
+      "learning_rate": 0.0001,
+      "loss": 1.4093,
+      "step": 12039
+    },
+    {
+      "epoch": 1.3987801336044148,
+      "grad_norm": 0.5646287202835083,
+      "learning_rate": 0.0001,
+      "loss": 1.4349,
+      "step": 12040
+    },
+    {
+      "epoch": 1.3988963113563753,
+      "grad_norm": 0.6419237852096558,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 12041
+    },
+    {
+      "epoch": 1.3990124891083358,
+      "grad_norm": 0.62077397108078,
+      "learning_rate": 0.0001,
+      "loss": 1.5193,
+      "step": 12042
+    },
+    {
+      "epoch": 1.3991286668602962,
+      "grad_norm": 0.5966094136238098,
+      "learning_rate": 0.0001,
+      "loss": 1.6089,
+      "step": 12043
+    },
+    {
+      "epoch": 1.3992448446122567,
+      "grad_norm": 0.5946581363677979,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 12044
+    },
+    {
+      "epoch": 1.3993610223642172,
+      "grad_norm": 0.5981970429420471,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 12045
+    },
+    {
+      "epoch": 1.3994772001161777,
+      "grad_norm": 0.5579191446304321,
+      "learning_rate": 0.0001,
+      "loss": 1.3866,
+      "step": 12046
+    },
+    {
+      "epoch": 1.3995933778681382,
+      "grad_norm": 0.5763770937919617,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 12047
+    },
+    {
+      "epoch": 1.3997095556200987,
+      "grad_norm": 0.5983228087425232,
+      "learning_rate": 0.0001,
+      "loss": 1.3796,
+      "step": 12048
+    },
+    {
+      "epoch": 1.3998257333720592,
+      "grad_norm": 0.5548502206802368,
+      "learning_rate": 0.0001,
+      "loss": 1.2765,
+      "step": 12049
+    },
+    {
+      "epoch": 1.3999419111240199,
+      "grad_norm": 0.598829984664917,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 12050
+    },
+    {
+      "epoch": 1.4000580888759804,
+      "grad_norm": 0.5747253894805908,
+      "learning_rate": 0.0001,
+      "loss": 1.2931,
+      "step": 12051
+    },
+    {
+      "epoch": 1.4001742666279409,
+      "grad_norm": 0.5953359007835388,
+      "learning_rate": 0.0001,
+      "loss": 1.3444,
+      "step": 12052
+    },
+    {
+      "epoch": 1.4002904443799014,
+      "grad_norm": 0.5862138271331787,
+      "learning_rate": 0.0001,
+      "loss": 1.6761,
+      "step": 12053
+    },
+    {
+      "epoch": 1.4004066221318618,
+      "grad_norm": 0.5845853090286255,
+      "learning_rate": 0.0001,
+      "loss": 1.5413,
+      "step": 12054
+    },
+    {
+      "epoch": 1.4005227998838223,
+      "grad_norm": 0.5902648568153381,
+      "learning_rate": 0.0001,
+      "loss": 1.372,
+      "step": 12055
+    },
+    {
+      "epoch": 1.4006389776357828,
+      "grad_norm": 0.6484795212745667,
+      "learning_rate": 0.0001,
+      "loss": 1.6057,
+      "step": 12056
+    },
+    {
+      "epoch": 1.4007551553877433,
+      "grad_norm": 0.6351671814918518,
+      "learning_rate": 0.0001,
+      "loss": 1.6507,
+      "step": 12057
+    },
+    {
+      "epoch": 1.4008713331397038,
+      "grad_norm": 0.5990087389945984,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 12058
+    },
+    {
+      "epoch": 1.4009875108916643,
+      "grad_norm": 0.5999724864959717,
+      "learning_rate": 0.0001,
+      "loss": 1.5903,
+      "step": 12059
+    },
+    {
+      "epoch": 1.4011036886436248,
+      "grad_norm": 0.6274200081825256,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 12060
+    },
+    {
+      "epoch": 1.4012198663955853,
+      "grad_norm": 0.5672786831855774,
+      "learning_rate": 0.0001,
+      "loss": 1.4998,
+      "step": 12061
+    },
+    {
+      "epoch": 1.4013360441475458,
+      "grad_norm": 0.6059845089912415,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 12062
+    },
+    {
+      "epoch": 1.4014522218995062,
+      "grad_norm": 0.583419144153595,
+      "learning_rate": 0.0001,
+      "loss": 1.3577,
+      "step": 12063
+    },
+    {
+      "epoch": 1.4015683996514667,
+      "grad_norm": 0.5932354927062988,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 12064
+    },
+    {
+      "epoch": 1.4016845774034272,
+      "grad_norm": 0.6560299396514893,
+      "learning_rate": 0.0001,
+      "loss": 1.4958,
+      "step": 12065
+    },
+    {
+      "epoch": 1.4018007551553877,
+      "grad_norm": 0.5621389150619507,
+      "learning_rate": 0.0001,
+      "loss": 1.5461,
+      "step": 12066
+    },
+    {
+      "epoch": 1.4019169329073482,
+      "grad_norm": 0.5786882638931274,
+      "learning_rate": 0.0001,
+      "loss": 1.4813,
+      "step": 12067
+    },
+    {
+      "epoch": 1.4020331106593087,
+      "grad_norm": 0.5459783673286438,
+      "learning_rate": 0.0001,
+      "loss": 1.5,
+      "step": 12068
+    },
+    {
+      "epoch": 1.4021492884112692,
+      "grad_norm": 0.5643264651298523,
+      "learning_rate": 0.0001,
+      "loss": 1.5432,
+      "step": 12069
+    },
+    {
+      "epoch": 1.4022654661632297,
+      "grad_norm": 0.5764142870903015,
+      "learning_rate": 0.0001,
+      "loss": 1.3799,
+      "step": 12070
+    },
+    {
+      "epoch": 1.4023816439151902,
+      "grad_norm": 0.5647251009941101,
+      "learning_rate": 0.0001,
+      "loss": 1.4376,
+      "step": 12071
+    },
+    {
+      "epoch": 1.4024978216671506,
+      "grad_norm": 0.5865373611450195,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 12072
+    },
+    {
+      "epoch": 1.4026139994191111,
+      "grad_norm": 0.5826906561851501,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 12073
+    },
+    {
+      "epoch": 1.4027301771710716,
+      "grad_norm": 0.5692470669746399,
+      "learning_rate": 0.0001,
+      "loss": 1.3101,
+      "step": 12074
+    },
+    {
+      "epoch": 1.4028463549230321,
+      "grad_norm": 0.5442948937416077,
+      "learning_rate": 0.0001,
+      "loss": 1.5004,
+      "step": 12075
+    },
+    {
+      "epoch": 1.4029625326749926,
+      "grad_norm": 0.5658257007598877,
+      "learning_rate": 0.0001,
+      "loss": 1.3957,
+      "step": 12076
+    },
+    {
+      "epoch": 1.4030787104269533,
+      "grad_norm": 0.595670759677887,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 12077
+    },
+    {
+      "epoch": 1.4031948881789138,
+      "grad_norm": 0.6052472591400146,
+      "learning_rate": 0.0001,
+      "loss": 1.6607,
+      "step": 12078
+    },
+    {
+      "epoch": 1.4033110659308743,
+      "grad_norm": 0.571075975894928,
+      "learning_rate": 0.0001,
+      "loss": 1.3262,
+      "step": 12079
+    },
+    {
+      "epoch": 1.4034272436828348,
+      "grad_norm": 0.5579153299331665,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 12080
+    },
+    {
+      "epoch": 1.4035434214347953,
+      "grad_norm": 0.5987997055053711,
+      "learning_rate": 0.0001,
+      "loss": 1.442,
+      "step": 12081
+    },
+    {
+      "epoch": 1.4036595991867558,
+      "grad_norm": 0.5797942280769348,
+      "learning_rate": 0.0001,
+      "loss": 1.5384,
+      "step": 12082
+    },
+    {
+      "epoch": 1.4037757769387162,
+      "grad_norm": 0.5952906012535095,
+      "learning_rate": 0.0001,
+      "loss": 1.3313,
+      "step": 12083
+    },
+    {
+      "epoch": 1.4038919546906767,
+      "grad_norm": 0.6068854331970215,
+      "learning_rate": 0.0001,
+      "loss": 1.427,
+      "step": 12084
+    },
+    {
+      "epoch": 1.4040081324426372,
+      "grad_norm": 0.5405552983283997,
+      "learning_rate": 0.0001,
+      "loss": 1.3378,
+      "step": 12085
+    },
+    {
+      "epoch": 1.4041243101945977,
+      "grad_norm": 0.572024405002594,
+      "learning_rate": 0.0001,
+      "loss": 1.425,
+      "step": 12086
+    },
+    {
+      "epoch": 1.4042404879465582,
+      "grad_norm": 0.5662723183631897,
+      "learning_rate": 0.0001,
+      "loss": 1.3541,
+      "step": 12087
+    },
+    {
+      "epoch": 1.4043566656985187,
+      "grad_norm": 0.6154362559318542,
+      "learning_rate": 0.0001,
+      "loss": 1.5061,
+      "step": 12088
+    },
+    {
+      "epoch": 1.4044728434504792,
+      "grad_norm": 0.6339954733848572,
+      "learning_rate": 0.0001,
+      "loss": 1.4753,
+      "step": 12089
+    },
+    {
+      "epoch": 1.4045890212024397,
+      "grad_norm": 0.6176583170890808,
+      "learning_rate": 0.0001,
+      "loss": 1.4823,
+      "step": 12090
+    },
+    {
+      "epoch": 1.4047051989544002,
+      "grad_norm": 0.6026437878608704,
+      "learning_rate": 0.0001,
+      "loss": 1.5651,
+      "step": 12091
+    },
+    {
+      "epoch": 1.4048213767063609,
+      "grad_norm": 0.6249407529830933,
+      "learning_rate": 0.0001,
+      "loss": 1.4091,
+      "step": 12092
+    },
+    {
+      "epoch": 1.4049375544583214,
+      "grad_norm": 0.583416223526001,
+      "learning_rate": 0.0001,
+      "loss": 1.4451,
+      "step": 12093
+    },
+    {
+      "epoch": 1.4050537322102818,
+      "grad_norm": 0.6077203750610352,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 12094
+    },
+    {
+      "epoch": 1.4051699099622423,
+      "grad_norm": 0.6278978586196899,
+      "learning_rate": 0.0001,
+      "loss": 1.6374,
+      "step": 12095
+    },
+    {
+      "epoch": 1.4052860877142028,
+      "grad_norm": 0.5796852707862854,
+      "learning_rate": 0.0001,
+      "loss": 1.4089,
+      "step": 12096
+    },
+    {
+      "epoch": 1.4054022654661633,
+      "grad_norm": 0.6815804839134216,
+      "learning_rate": 0.0001,
+      "loss": 1.596,
+      "step": 12097
+    },
+    {
+      "epoch": 1.4055184432181238,
+      "grad_norm": 0.5999861359596252,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 12098
+    },
+    {
+      "epoch": 1.4056346209700843,
+      "grad_norm": 0.5877710580825806,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 12099
+    },
+    {
+      "epoch": 1.4057507987220448,
+      "grad_norm": 0.5974360108375549,
+      "learning_rate": 0.0001,
+      "loss": 1.4517,
+      "step": 12100
+    },
+    {
+      "epoch": 1.4058669764740053,
+      "grad_norm": 0.6353722810745239,
+      "learning_rate": 0.0001,
+      "loss": 1.5599,
+      "step": 12101
+    },
+    {
+      "epoch": 1.4059831542259658,
+      "grad_norm": 0.5967050194740295,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 12102
+    },
+    {
+      "epoch": 1.4060993319779262,
+      "grad_norm": 0.5885502099990845,
+      "learning_rate": 0.0001,
+      "loss": 1.4882,
+      "step": 12103
+    },
+    {
+      "epoch": 1.4062155097298867,
+      "grad_norm": 0.5896927118301392,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 12104
+    },
+    {
+      "epoch": 1.4063316874818472,
+      "grad_norm": 0.536248505115509,
+      "learning_rate": 0.0001,
+      "loss": 1.3211,
+      "step": 12105
+    },
+    {
+      "epoch": 1.4064478652338077,
+      "grad_norm": 0.5472630858421326,
+      "learning_rate": 0.0001,
+      "loss": 1.43,
+      "step": 12106
+    },
+    {
+      "epoch": 1.4065640429857682,
+      "grad_norm": 0.5885692238807678,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 12107
+    },
+    {
+      "epoch": 1.4066802207377287,
+      "grad_norm": 0.5371417999267578,
+      "learning_rate": 0.0001,
+      "loss": 1.187,
+      "step": 12108
+    },
+    {
+      "epoch": 1.4067963984896892,
+      "grad_norm": 0.5889664888381958,
+      "learning_rate": 0.0001,
+      "loss": 1.3359,
+      "step": 12109
+    },
+    {
+      "epoch": 1.4069125762416497,
+      "grad_norm": 0.6135367751121521,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 12110
+    },
+    {
+      "epoch": 1.4070287539936102,
+      "grad_norm": 0.5663803815841675,
+      "learning_rate": 0.0001,
+      "loss": 1.2486,
+      "step": 12111
+    },
+    {
+      "epoch": 1.4071449317455706,
+      "grad_norm": 0.6089535355567932,
+      "learning_rate": 0.0001,
+      "loss": 1.6265,
+      "step": 12112
+    },
+    {
+      "epoch": 1.4072611094975311,
+      "grad_norm": 0.6175603270530701,
+      "learning_rate": 0.0001,
+      "loss": 1.6734,
+      "step": 12113
+    },
+    {
+      "epoch": 1.4073772872494916,
+      "grad_norm": 0.5911760330200195,
+      "learning_rate": 0.0001,
+      "loss": 1.3986,
+      "step": 12114
+    },
+    {
+      "epoch": 1.4074934650014521,
+      "grad_norm": 0.6093399524688721,
+      "learning_rate": 0.0001,
+      "loss": 1.5308,
+      "step": 12115
+    },
+    {
+      "epoch": 1.4076096427534126,
+      "grad_norm": 0.5803685188293457,
+      "learning_rate": 0.0001,
+      "loss": 1.3183,
+      "step": 12116
+    },
+    {
+      "epoch": 1.407725820505373,
+      "grad_norm": 0.5896522998809814,
+      "learning_rate": 0.0001,
+      "loss": 1.6055,
+      "step": 12117
+    },
+    {
+      "epoch": 1.4078419982573336,
+      "grad_norm": 0.573806881904602,
+      "learning_rate": 0.0001,
+      "loss": 1.3718,
+      "step": 12118
+    },
+    {
+      "epoch": 1.4079581760092943,
+      "grad_norm": 0.5594179034233093,
+      "learning_rate": 0.0001,
+      "loss": 1.3664,
+      "step": 12119
+    },
+    {
+      "epoch": 1.4080743537612548,
+      "grad_norm": 0.6025107502937317,
+      "learning_rate": 0.0001,
+      "loss": 1.3603,
+      "step": 12120
+    },
+    {
+      "epoch": 1.4081905315132153,
+      "grad_norm": 0.5635510683059692,
+      "learning_rate": 0.0001,
+      "loss": 1.5932,
+      "step": 12121
+    },
+    {
+      "epoch": 1.4083067092651758,
+      "grad_norm": 0.5392552614212036,
+      "learning_rate": 0.0001,
+      "loss": 1.3761,
+      "step": 12122
+    },
+    {
+      "epoch": 1.4084228870171362,
+      "grad_norm": 0.5588886141777039,
+      "learning_rate": 0.0001,
+      "loss": 1.4119,
+      "step": 12123
+    },
+    {
+      "epoch": 1.4085390647690967,
+      "grad_norm": 0.5739279389381409,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 12124
+    },
+    {
+      "epoch": 1.4086552425210572,
+      "grad_norm": 0.5879178047180176,
+      "learning_rate": 0.0001,
+      "loss": 1.6335,
+      "step": 12125
+    },
+    {
+      "epoch": 1.4087714202730177,
+      "grad_norm": 0.5897621512413025,
+      "learning_rate": 0.0001,
+      "loss": 1.5896,
+      "step": 12126
+    },
+    {
+      "epoch": 1.4088875980249782,
+      "grad_norm": 0.6275389194488525,
+      "learning_rate": 0.0001,
+      "loss": 1.5319,
+      "step": 12127
+    },
+    {
+      "epoch": 1.4090037757769387,
+      "grad_norm": 0.6143684983253479,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 12128
+    },
+    {
+      "epoch": 1.4091199535288992,
+      "grad_norm": 0.5457308292388916,
+      "learning_rate": 0.0001,
+      "loss": 1.305,
+      "step": 12129
+    },
+    {
+      "epoch": 1.4092361312808597,
+      "grad_norm": 0.6137149930000305,
+      "learning_rate": 0.0001,
+      "loss": 1.5955,
+      "step": 12130
+    },
+    {
+      "epoch": 1.4093523090328202,
+      "grad_norm": 0.5918616056442261,
+      "learning_rate": 0.0001,
+      "loss": 1.3573,
+      "step": 12131
+    },
+    {
+      "epoch": 1.4094684867847806,
+      "grad_norm": 0.5807372331619263,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 12132
+    },
+    {
+      "epoch": 1.4095846645367411,
+      "grad_norm": 0.6110495328903198,
+      "learning_rate": 0.0001,
+      "loss": 1.4377,
+      "step": 12133
+    },
+    {
+      "epoch": 1.4097008422887018,
+      "grad_norm": 0.5787477493286133,
+      "learning_rate": 0.0001,
+      "loss": 1.3938,
+      "step": 12134
+    },
+    {
+      "epoch": 1.4098170200406623,
+      "grad_norm": 0.5760836005210876,
+      "learning_rate": 0.0001,
+      "loss": 1.429,
+      "step": 12135
+    },
+    {
+      "epoch": 1.4099331977926228,
+      "grad_norm": 0.6453091502189636,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 12136
+    },
+    {
+      "epoch": 1.4100493755445833,
+      "grad_norm": 0.6404998898506165,
+      "learning_rate": 0.0001,
+      "loss": 1.5553,
+      "step": 12137
+    },
+    {
+      "epoch": 1.4101655532965438,
+      "grad_norm": 0.6601923108100891,
+      "learning_rate": 0.0001,
+      "loss": 1.5512,
+      "step": 12138
+    },
+    {
+      "epoch": 1.4102817310485043,
+      "grad_norm": 0.6087546944618225,
+      "learning_rate": 0.0001,
+      "loss": 1.4711,
+      "step": 12139
+    },
+    {
+      "epoch": 1.4103979088004648,
+      "grad_norm": 0.5570737719535828,
+      "learning_rate": 0.0001,
+      "loss": 1.2991,
+      "step": 12140
+    },
+    {
+      "epoch": 1.4105140865524253,
+      "grad_norm": 0.5413463115692139,
+      "learning_rate": 0.0001,
+      "loss": 1.185,
+      "step": 12141
+    },
+    {
+      "epoch": 1.4106302643043858,
+      "grad_norm": 0.5919083952903748,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 12142
+    },
+    {
+      "epoch": 1.4107464420563463,
+      "grad_norm": 0.593340277671814,
+      "learning_rate": 0.0001,
+      "loss": 1.7068,
+      "step": 12143
+    },
+    {
+      "epoch": 1.4108626198083067,
+      "grad_norm": 0.6711177229881287,
+      "learning_rate": 0.0001,
+      "loss": 1.4589,
+      "step": 12144
+    },
+    {
+      "epoch": 1.4109787975602672,
+      "grad_norm": 0.6454046368598938,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 12145
+    },
+    {
+      "epoch": 1.4110949753122277,
+      "grad_norm": 0.576941967010498,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 12146
+    },
+    {
+      "epoch": 1.4112111530641882,
+      "grad_norm": 0.5717135667800903,
+      "learning_rate": 0.0001,
+      "loss": 1.4607,
+      "step": 12147
+    },
+    {
+      "epoch": 1.4113273308161487,
+      "grad_norm": 0.5811867117881775,
+      "learning_rate": 0.0001,
+      "loss": 1.4087,
+      "step": 12148
+    },
+    {
+      "epoch": 1.4114435085681092,
+      "grad_norm": 0.5930108428001404,
+      "learning_rate": 0.0001,
+      "loss": 1.425,
+      "step": 12149
+    },
+    {
+      "epoch": 1.4115596863200697,
+      "grad_norm": 0.6307324767112732,
+      "learning_rate": 0.0001,
+      "loss": 1.4217,
+      "step": 12150
+    },
+    {
+      "epoch": 1.4116758640720302,
+      "grad_norm": 0.5676311254501343,
+      "learning_rate": 0.0001,
+      "loss": 1.3643,
+      "step": 12151
+    },
+    {
+      "epoch": 1.4117920418239907,
+      "grad_norm": 0.6175429821014404,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 12152
+    },
+    {
+      "epoch": 1.4119082195759511,
+      "grad_norm": 0.6128252744674683,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 12153
+    },
+    {
+      "epoch": 1.4120243973279116,
+      "grad_norm": 0.5954226851463318,
+      "learning_rate": 0.0001,
+      "loss": 1.4271,
+      "step": 12154
+    },
+    {
+      "epoch": 1.4121405750798721,
+      "grad_norm": 0.6151418089866638,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 12155
+    },
+    {
+      "epoch": 1.4122567528318326,
+      "grad_norm": 0.6414467096328735,
+      "learning_rate": 0.0001,
+      "loss": 1.4582,
+      "step": 12156
+    },
+    {
+      "epoch": 1.412372930583793,
+      "grad_norm": 0.6506933569908142,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 12157
+    },
+    {
+      "epoch": 1.4124891083357536,
+      "grad_norm": 0.603602409362793,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 12158
+    },
+    {
+      "epoch": 1.412605286087714,
+      "grad_norm": 0.5889068245887756,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 12159
+    },
+    {
+      "epoch": 1.4127214638396748,
+      "grad_norm": 0.5658511519432068,
+      "learning_rate": 0.0001,
+      "loss": 1.4418,
+      "step": 12160
+    },
+    {
+      "epoch": 1.4128376415916353,
+      "grad_norm": 0.5495668053627014,
+      "learning_rate": 0.0001,
+      "loss": 1.3484,
+      "step": 12161
+    },
+    {
+      "epoch": 1.4129538193435958,
+      "grad_norm": 0.5731701254844666,
+      "learning_rate": 0.0001,
+      "loss": 1.3556,
+      "step": 12162
+    },
+    {
+      "epoch": 1.4130699970955563,
+      "grad_norm": 0.6092509031295776,
+      "learning_rate": 0.0001,
+      "loss": 1.5438,
+      "step": 12163
+    },
+    {
+      "epoch": 1.4131861748475167,
+      "grad_norm": 0.6485771536827087,
+      "learning_rate": 0.0001,
+      "loss": 1.5345,
+      "step": 12164
+    },
+    {
+      "epoch": 1.4133023525994772,
+      "grad_norm": 0.5795009136199951,
+      "learning_rate": 0.0001,
+      "loss": 1.492,
+      "step": 12165
+    },
+    {
+      "epoch": 1.4134185303514377,
+      "grad_norm": 0.5661119222640991,
+      "learning_rate": 0.0001,
+      "loss": 1.3841,
+      "step": 12166
+    },
+    {
+      "epoch": 1.4135347081033982,
+      "grad_norm": 0.5987005829811096,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 12167
+    },
+    {
+      "epoch": 1.4136508858553587,
+      "grad_norm": 0.5951616168022156,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 12168
+    },
+    {
+      "epoch": 1.4137670636073192,
+      "grad_norm": 0.6180946826934814,
+      "learning_rate": 0.0001,
+      "loss": 1.5858,
+      "step": 12169
+    },
+    {
+      "epoch": 1.4138832413592797,
+      "grad_norm": 0.5951772928237915,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 12170
+    },
+    {
+      "epoch": 1.4139994191112402,
+      "grad_norm": 0.5945565700531006,
+      "learning_rate": 0.0001,
+      "loss": 1.3671,
+      "step": 12171
+    },
+    {
+      "epoch": 1.4141155968632007,
+      "grad_norm": 0.6298597455024719,
+      "learning_rate": 0.0001,
+      "loss": 1.5741,
+      "step": 12172
+    },
+    {
+      "epoch": 1.4142317746151611,
+      "grad_norm": 0.5758437514305115,
+      "learning_rate": 0.0001,
+      "loss": 1.478,
+      "step": 12173
+    },
+    {
+      "epoch": 1.4143479523671216,
+      "grad_norm": 0.6084062457084656,
+      "learning_rate": 0.0001,
+      "loss": 1.635,
+      "step": 12174
+    },
+    {
+      "epoch": 1.4144641301190821,
+      "grad_norm": 0.5881122946739197,
+      "learning_rate": 0.0001,
+      "loss": 1.3891,
+      "step": 12175
+    },
+    {
+      "epoch": 1.4145803078710428,
+      "grad_norm": 0.6065901517868042,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 12176
+    },
+    {
+      "epoch": 1.4146964856230033,
+      "grad_norm": 0.57171231508255,
+      "learning_rate": 0.0001,
+      "loss": 1.353,
+      "step": 12177
+    },
+    {
+      "epoch": 1.4148126633749638,
+      "grad_norm": 0.606834888458252,
+      "learning_rate": 0.0001,
+      "loss": 1.4338,
+      "step": 12178
+    },
+    {
+      "epoch": 1.4149288411269243,
+      "grad_norm": 0.6262538433074951,
+      "learning_rate": 0.0001,
+      "loss": 1.4993,
+      "step": 12179
+    },
+    {
+      "epoch": 1.4150450188788848,
+      "grad_norm": 0.6646376252174377,
+      "learning_rate": 0.0001,
+      "loss": 1.6679,
+      "step": 12180
+    },
+    {
+      "epoch": 1.4151611966308453,
+      "grad_norm": 0.6121298670768738,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 12181
+    },
+    {
+      "epoch": 1.4152773743828058,
+      "grad_norm": 0.5688852071762085,
+      "learning_rate": 0.0001,
+      "loss": 1.4792,
+      "step": 12182
+    },
+    {
+      "epoch": 1.4153935521347663,
+      "grad_norm": 0.5843722820281982,
+      "learning_rate": 0.0001,
+      "loss": 1.3999,
+      "step": 12183
+    },
+    {
+      "epoch": 1.4155097298867267,
+      "grad_norm": 0.6013531684875488,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 12184
+    },
+    {
+      "epoch": 1.4156259076386872,
+      "grad_norm": 0.5897423028945923,
+      "learning_rate": 0.0001,
+      "loss": 1.4682,
+      "step": 12185
+    },
+    {
+      "epoch": 1.4157420853906477,
+      "grad_norm": 0.6187875270843506,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 12186
+    },
+    {
+      "epoch": 1.4158582631426082,
+      "grad_norm": 0.5957651734352112,
+      "learning_rate": 0.0001,
+      "loss": 1.4374,
+      "step": 12187
+    },
+    {
+      "epoch": 1.4159744408945687,
+      "grad_norm": 0.6102905869483948,
+      "learning_rate": 0.0001,
+      "loss": 1.5182,
+      "step": 12188
+    },
+    {
+      "epoch": 1.4160906186465292,
+      "grad_norm": 0.5826073288917542,
+      "learning_rate": 0.0001,
+      "loss": 1.3984,
+      "step": 12189
+    },
+    {
+      "epoch": 1.4162067963984897,
+      "grad_norm": 0.6257545948028564,
+      "learning_rate": 0.0001,
+      "loss": 1.4765,
+      "step": 12190
+    },
+    {
+      "epoch": 1.4163229741504502,
+      "grad_norm": 0.6139261722564697,
+      "learning_rate": 0.0001,
+      "loss": 1.4693,
+      "step": 12191
+    },
+    {
+      "epoch": 1.4164391519024107,
+      "grad_norm": 0.6307207942008972,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 12192
+    },
+    {
+      "epoch": 1.4165553296543711,
+      "grad_norm": 0.603281557559967,
+      "learning_rate": 0.0001,
+      "loss": 1.4784,
+      "step": 12193
+    },
+    {
+      "epoch": 1.4166715074063316,
+      "grad_norm": 0.5983949303627014,
+      "learning_rate": 0.0001,
+      "loss": 1.4863,
+      "step": 12194
+    },
+    {
+      "epoch": 1.4167876851582921,
+      "grad_norm": 0.5481703281402588,
+      "learning_rate": 0.0001,
+      "loss": 1.4079,
+      "step": 12195
+    },
+    {
+      "epoch": 1.4169038629102526,
+      "grad_norm": 0.6520917415618896,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 12196
+    },
+    {
+      "epoch": 1.417020040662213,
+      "grad_norm": 0.553615927696228,
+      "learning_rate": 0.0001,
+      "loss": 1.3623,
+      "step": 12197
+    },
+    {
+      "epoch": 1.4171362184141736,
+      "grad_norm": 0.5994259119033813,
+      "learning_rate": 0.0001,
+      "loss": 1.6403,
+      "step": 12198
+    },
+    {
+      "epoch": 1.417252396166134,
+      "grad_norm": 0.5477364659309387,
+      "learning_rate": 0.0001,
+      "loss": 1.3191,
+      "step": 12199
+    },
+    {
+      "epoch": 1.4173685739180946,
+      "grad_norm": 0.6125954389572144,
+      "learning_rate": 0.0001,
+      "loss": 1.4055,
+      "step": 12200
+    },
+    {
+      "epoch": 1.417484751670055,
+      "grad_norm": 0.642378032207489,
+      "learning_rate": 0.0001,
+      "loss": 1.5665,
+      "step": 12201
+    },
+    {
+      "epoch": 1.4176009294220158,
+      "grad_norm": 0.5678790807723999,
+      "learning_rate": 0.0001,
+      "loss": 1.3472,
+      "step": 12202
+    },
+    {
+      "epoch": 1.4177171071739763,
+      "grad_norm": 0.6221288442611694,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 12203
+    },
+    {
+      "epoch": 1.4178332849259367,
+      "grad_norm": 0.6019090414047241,
+      "learning_rate": 0.0001,
+      "loss": 1.6035,
+      "step": 12204
+    },
+    {
+      "epoch": 1.4179494626778972,
+      "grad_norm": 0.6357331275939941,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 12205
+    },
+    {
+      "epoch": 1.4180656404298577,
+      "grad_norm": 0.5553552508354187,
+      "learning_rate": 0.0001,
+      "loss": 1.2786,
+      "step": 12206
+    },
+    {
+      "epoch": 1.4181818181818182,
+      "grad_norm": 0.5938397645950317,
+      "learning_rate": 0.0001,
+      "loss": 1.4726,
+      "step": 12207
+    },
+    {
+      "epoch": 1.4182979959337787,
+      "grad_norm": 0.5866955518722534,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 12208
+    },
+    {
+      "epoch": 1.4184141736857392,
+      "grad_norm": 0.5551325678825378,
+      "learning_rate": 0.0001,
+      "loss": 1.4403,
+      "step": 12209
+    },
+    {
+      "epoch": 1.4185303514376997,
+      "grad_norm": 0.5848428606987,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 12210
+    },
+    {
+      "epoch": 1.4186465291896602,
+      "grad_norm": 0.5819724798202515,
+      "learning_rate": 0.0001,
+      "loss": 1.288,
+      "step": 12211
+    },
+    {
+      "epoch": 1.4187627069416207,
+      "grad_norm": 0.5821203589439392,
+      "learning_rate": 0.0001,
+      "loss": 1.3893,
+      "step": 12212
+    },
+    {
+      "epoch": 1.4188788846935811,
+      "grad_norm": 0.6024807691574097,
+      "learning_rate": 0.0001,
+      "loss": 1.6227,
+      "step": 12213
+    },
+    {
+      "epoch": 1.4189950624455416,
+      "grad_norm": 0.5902191996574402,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 12214
+    },
+    {
+      "epoch": 1.4191112401975021,
+      "grad_norm": 0.5712905526161194,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 12215
+    },
+    {
+      "epoch": 1.4192274179494626,
+      "grad_norm": 0.5864148139953613,
+      "learning_rate": 0.0001,
+      "loss": 1.5394,
+      "step": 12216
+    },
+    {
+      "epoch": 1.419343595701423,
+      "grad_norm": 0.5870895385742188,
+      "learning_rate": 0.0001,
+      "loss": 1.4937,
+      "step": 12217
+    },
+    {
+      "epoch": 1.4194597734533838,
+      "grad_norm": 0.6374808549880981,
+      "learning_rate": 0.0001,
+      "loss": 1.6601,
+      "step": 12218
+    },
+    {
+      "epoch": 1.4195759512053443,
+      "grad_norm": 0.5821941494941711,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 12219
+    },
+    {
+      "epoch": 1.4196921289573048,
+      "grad_norm": 0.6282082796096802,
+      "learning_rate": 0.0001,
+      "loss": 1.6674,
+      "step": 12220
+    },
+    {
+      "epoch": 1.4198083067092653,
+      "grad_norm": 0.5595186352729797,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 12221
+    },
+    {
+      "epoch": 1.4199244844612258,
+      "grad_norm": 0.5783095359802246,
+      "learning_rate": 0.0001,
+      "loss": 1.2876,
+      "step": 12222
+    },
+    {
+      "epoch": 1.4200406622131863,
+      "grad_norm": 0.6268340349197388,
+      "learning_rate": 0.0001,
+      "loss": 1.2669,
+      "step": 12223
+    },
+    {
+      "epoch": 1.4201568399651467,
+      "grad_norm": 0.6090834140777588,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 12224
+    },
+    {
+      "epoch": 1.4202730177171072,
+      "grad_norm": 0.616369903087616,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 12225
+    },
+    {
+      "epoch": 1.4203891954690677,
+      "grad_norm": 0.5462055802345276,
+      "learning_rate": 0.0001,
+      "loss": 1.3814,
+      "step": 12226
+    },
+    {
+      "epoch": 1.4205053732210282,
+      "grad_norm": 0.6000543236732483,
+      "learning_rate": 0.0001,
+      "loss": 1.4903,
+      "step": 12227
+    },
+    {
+      "epoch": 1.4206215509729887,
+      "grad_norm": 0.5703785419464111,
+      "learning_rate": 0.0001,
+      "loss": 1.3543,
+      "step": 12228
+    },
+    {
+      "epoch": 1.4207377287249492,
+      "grad_norm": 0.6079279780387878,
+      "learning_rate": 0.0001,
+      "loss": 1.6149,
+      "step": 12229
+    },
+    {
+      "epoch": 1.4208539064769097,
+      "grad_norm": 0.5839101076126099,
+      "learning_rate": 0.0001,
+      "loss": 1.3544,
+      "step": 12230
+    },
+    {
+      "epoch": 1.4209700842288702,
+      "grad_norm": 0.5797370672225952,
+      "learning_rate": 0.0001,
+      "loss": 1.3705,
+      "step": 12231
+    },
+    {
+      "epoch": 1.4210862619808307,
+      "grad_norm": 0.6282606720924377,
+      "learning_rate": 0.0001,
+      "loss": 1.7211,
+      "step": 12232
+    },
+    {
+      "epoch": 1.4212024397327911,
+      "grad_norm": 0.6144933700561523,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 12233
+    },
+    {
+      "epoch": 1.4213186174847516,
+      "grad_norm": 0.5617024898529053,
+      "learning_rate": 0.0001,
+      "loss": 1.4082,
+      "step": 12234
+    },
+    {
+      "epoch": 1.4214347952367121,
+      "grad_norm": 0.6399552226066589,
+      "learning_rate": 0.0001,
+      "loss": 1.4947,
+      "step": 12235
+    },
+    {
+      "epoch": 1.4215509729886726,
+      "grad_norm": 0.5742717385292053,
+      "learning_rate": 0.0001,
+      "loss": 1.477,
+      "step": 12236
+    },
+    {
+      "epoch": 1.421667150740633,
+      "grad_norm": 0.5643975734710693,
+      "learning_rate": 0.0001,
+      "loss": 1.6196,
+      "step": 12237
+    },
+    {
+      "epoch": 1.4217833284925936,
+      "grad_norm": 0.593525230884552,
+      "learning_rate": 0.0001,
+      "loss": 1.5359,
+      "step": 12238
+    },
+    {
+      "epoch": 1.421899506244554,
+      "grad_norm": 0.5971342325210571,
+      "learning_rate": 0.0001,
+      "loss": 1.488,
+      "step": 12239
+    },
+    {
+      "epoch": 1.4220156839965146,
+      "grad_norm": 0.6006376147270203,
+      "learning_rate": 0.0001,
+      "loss": 1.4928,
+      "step": 12240
+    },
+    {
+      "epoch": 1.422131861748475,
+      "grad_norm": 0.6138136982917786,
+      "learning_rate": 0.0001,
+      "loss": 1.4406,
+      "step": 12241
+    },
+    {
+      "epoch": 1.4222480395004355,
+      "grad_norm": 0.5824870467185974,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 12242
+    },
+    {
+      "epoch": 1.422364217252396,
+      "grad_norm": 0.6046587824821472,
+      "learning_rate": 0.0001,
+      "loss": 1.5849,
+      "step": 12243
+    },
+    {
+      "epoch": 1.4224803950043567,
+      "grad_norm": 0.5978199243545532,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 12244
+    },
+    {
+      "epoch": 1.4225965727563172,
+      "grad_norm": 0.5881246328353882,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 12245
+    },
+    {
+      "epoch": 1.4227127505082777,
+      "grad_norm": 0.5994898676872253,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 12246
+    },
+    {
+      "epoch": 1.4228289282602382,
+      "grad_norm": 0.6213389039039612,
+      "learning_rate": 0.0001,
+      "loss": 1.571,
+      "step": 12247
+    },
+    {
+      "epoch": 1.4229451060121987,
+      "grad_norm": 0.5992786884307861,
+      "learning_rate": 0.0001,
+      "loss": 1.3701,
+      "step": 12248
+    },
+    {
+      "epoch": 1.4230612837641592,
+      "grad_norm": 0.5737839341163635,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 12249
+    },
+    {
+      "epoch": 1.4231774615161197,
+      "grad_norm": 0.6159092783927917,
+      "learning_rate": 0.0001,
+      "loss": 1.6564,
+      "step": 12250
+    },
+    {
+      "epoch": 1.4232936392680802,
+      "grad_norm": 0.596993625164032,
+      "learning_rate": 0.0001,
+      "loss": 1.3847,
+      "step": 12251
+    },
+    {
+      "epoch": 1.4234098170200407,
+      "grad_norm": 0.5880178213119507,
+      "learning_rate": 0.0001,
+      "loss": 1.5226,
+      "step": 12252
+    },
+    {
+      "epoch": 1.4235259947720011,
+      "grad_norm": 0.6582285165786743,
+      "learning_rate": 0.0001,
+      "loss": 1.5161,
+      "step": 12253
+    },
+    {
+      "epoch": 1.4236421725239616,
+      "grad_norm": 0.608213484287262,
+      "learning_rate": 0.0001,
+      "loss": 1.5923,
+      "step": 12254
+    },
+    {
+      "epoch": 1.4237583502759221,
+      "grad_norm": 0.5370960235595703,
+      "learning_rate": 0.0001,
+      "loss": 1.4218,
+      "step": 12255
+    },
+    {
+      "epoch": 1.4238745280278826,
+      "grad_norm": 0.6302644610404968,
+      "learning_rate": 0.0001,
+      "loss": 1.6416,
+      "step": 12256
+    },
+    {
+      "epoch": 1.423990705779843,
+      "grad_norm": 0.5852667093276978,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 12257
+    },
+    {
+      "epoch": 1.4241068835318036,
+      "grad_norm": 0.6241486072540283,
+      "learning_rate": 0.0001,
+      "loss": 1.5519,
+      "step": 12258
+    },
+    {
+      "epoch": 1.424223061283764,
+      "grad_norm": 0.5911672711372375,
+      "learning_rate": 0.0001,
+      "loss": 1.597,
+      "step": 12259
+    },
+    {
+      "epoch": 1.4243392390357248,
+      "grad_norm": 0.5758203268051147,
+      "learning_rate": 0.0001,
+      "loss": 1.4032,
+      "step": 12260
+    },
+    {
+      "epoch": 1.4244554167876853,
+      "grad_norm": 0.6107878088951111,
+      "learning_rate": 0.0001,
+      "loss": 1.5565,
+      "step": 12261
+    },
+    {
+      "epoch": 1.4245715945396458,
+      "grad_norm": 0.5618066787719727,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 12262
+    },
+    {
+      "epoch": 1.4246877722916063,
+      "grad_norm": 0.6158115863800049,
+      "learning_rate": 0.0001,
+      "loss": 1.8096,
+      "step": 12263
+    },
+    {
+      "epoch": 1.4248039500435667,
+      "grad_norm": 0.6369965672492981,
+      "learning_rate": 0.0001,
+      "loss": 1.6449,
+      "step": 12264
+    },
+    {
+      "epoch": 1.4249201277955272,
+      "grad_norm": 0.6416753530502319,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 12265
+    },
+    {
+      "epoch": 1.4250363055474877,
+      "grad_norm": 0.5738315582275391,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 12266
+    },
+    {
+      "epoch": 1.4251524832994482,
+      "grad_norm": 0.559740424156189,
+      "learning_rate": 0.0001,
+      "loss": 1.5697,
+      "step": 12267
+    },
+    {
+      "epoch": 1.4252686610514087,
+      "grad_norm": 0.5917907953262329,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 12268
+    },
+    {
+      "epoch": 1.4253848388033692,
+      "grad_norm": 0.5424224734306335,
+      "learning_rate": 0.0001,
+      "loss": 1.3014,
+      "step": 12269
+    },
+    {
+      "epoch": 1.4255010165553297,
+      "grad_norm": 0.5904528498649597,
+      "learning_rate": 0.0001,
+      "loss": 1.4219,
+      "step": 12270
+    },
+    {
+      "epoch": 1.4256171943072902,
+      "grad_norm": 0.6030588746070862,
+      "learning_rate": 0.0001,
+      "loss": 1.454,
+      "step": 12271
+    },
+    {
+      "epoch": 1.4257333720592507,
+      "grad_norm": 0.5945565104484558,
+      "learning_rate": 0.0001,
+      "loss": 1.3429,
+      "step": 12272
+    },
+    {
+      "epoch": 1.4258495498112111,
+      "grad_norm": 0.6340063810348511,
+      "learning_rate": 0.0001,
+      "loss": 1.5204,
+      "step": 12273
+    },
+    {
+      "epoch": 1.4259657275631716,
+      "grad_norm": 0.5972619652748108,
+      "learning_rate": 0.0001,
+      "loss": 1.2667,
+      "step": 12274
+    },
+    {
+      "epoch": 1.4260819053151321,
+      "grad_norm": 0.6026214361190796,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 12275
+    },
+    {
+      "epoch": 1.4261980830670926,
+      "grad_norm": 0.6557515859603882,
+      "learning_rate": 0.0001,
+      "loss": 1.6499,
+      "step": 12276
+    },
+    {
+      "epoch": 1.426314260819053,
+      "grad_norm": 0.6378829479217529,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 12277
+    },
+    {
+      "epoch": 1.4264304385710136,
+      "grad_norm": 0.655852735042572,
+      "learning_rate": 0.0001,
+      "loss": 1.6301,
+      "step": 12278
+    },
+    {
+      "epoch": 1.426546616322974,
+      "grad_norm": 0.5908859968185425,
+      "learning_rate": 0.0001,
+      "loss": 1.3729,
+      "step": 12279
+    },
+    {
+      "epoch": 1.4266627940749346,
+      "grad_norm": 0.6130912899971008,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 12280
+    },
+    {
+      "epoch": 1.426778971826895,
+      "grad_norm": 0.5807083249092102,
+      "learning_rate": 0.0001,
+      "loss": 1.4061,
+      "step": 12281
+    },
+    {
+      "epoch": 1.4268951495788555,
+      "grad_norm": 0.6125562191009521,
+      "learning_rate": 0.0001,
+      "loss": 1.3565,
+      "step": 12282
+    },
+    {
+      "epoch": 1.427011327330816,
+      "grad_norm": 0.5933031439781189,
+      "learning_rate": 0.0001,
+      "loss": 1.6421,
+      "step": 12283
+    },
+    {
+      "epoch": 1.4271275050827765,
+      "grad_norm": 0.5931923389434814,
+      "learning_rate": 0.0001,
+      "loss": 1.4202,
+      "step": 12284
+    },
+    {
+      "epoch": 1.427243682834737,
+      "grad_norm": 0.5963932275772095,
+      "learning_rate": 0.0001,
+      "loss": 1.3515,
+      "step": 12285
+    },
+    {
+      "epoch": 1.4273598605866977,
+      "grad_norm": 0.5983773469924927,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 12286
+    },
+    {
+      "epoch": 1.4274760383386582,
+      "grad_norm": 0.5747543573379517,
+      "learning_rate": 0.0001,
+      "loss": 1.398,
+      "step": 12287
+    },
+    {
+      "epoch": 1.4275922160906187,
+      "grad_norm": 0.6180018782615662,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 12288
+    },
+    {
+      "epoch": 1.4277083938425792,
+      "grad_norm": 0.6087707281112671,
+      "learning_rate": 0.0001,
+      "loss": 1.5047,
+      "step": 12289
+    },
+    {
+      "epoch": 1.4278245715945397,
+      "grad_norm": 0.5981317758560181,
+      "learning_rate": 0.0001,
+      "loss": 1.4278,
+      "step": 12290
+    },
+    {
+      "epoch": 1.4279407493465002,
+      "grad_norm": 0.5648661255836487,
+      "learning_rate": 0.0001,
+      "loss": 1.2978,
+      "step": 12291
+    },
+    {
+      "epoch": 1.4280569270984607,
+      "grad_norm": 0.6909130215644836,
+      "learning_rate": 0.0001,
+      "loss": 1.7132,
+      "step": 12292
+    },
+    {
+      "epoch": 1.4281731048504211,
+      "grad_norm": 0.603410005569458,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 12293
+    },
+    {
+      "epoch": 1.4282892826023816,
+      "grad_norm": 0.5939321517944336,
+      "learning_rate": 0.0001,
+      "loss": 1.6074,
+      "step": 12294
+    },
+    {
+      "epoch": 1.4284054603543421,
+      "grad_norm": 0.5982928276062012,
+      "learning_rate": 0.0001,
+      "loss": 1.4063,
+      "step": 12295
+    },
+    {
+      "epoch": 1.4285216381063026,
+      "grad_norm": 0.6024434566497803,
+      "learning_rate": 0.0001,
+      "loss": 1.3077,
+      "step": 12296
+    },
+    {
+      "epoch": 1.428637815858263,
+      "grad_norm": 0.656810998916626,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 12297
+    },
+    {
+      "epoch": 1.4287539936102236,
+      "grad_norm": 0.6280006170272827,
+      "learning_rate": 0.0001,
+      "loss": 1.7002,
+      "step": 12298
+    },
+    {
+      "epoch": 1.428870171362184,
+      "grad_norm": 0.6182981729507446,
+      "learning_rate": 0.0001,
+      "loss": 1.4683,
+      "step": 12299
+    },
+    {
+      "epoch": 1.4289863491141446,
+      "grad_norm": 0.6073260307312012,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 12300
+    },
+    {
+      "epoch": 1.429102526866105,
+      "grad_norm": 0.6266109943389893,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 12301
+    },
+    {
+      "epoch": 1.4292187046180658,
+      "grad_norm": 0.5566325187683105,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 12302
+    },
+    {
+      "epoch": 1.4293348823700263,
+      "grad_norm": 0.6451058387756348,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 12303
+    },
+    {
+      "epoch": 1.4294510601219867,
+      "grad_norm": 0.6131970882415771,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 12304
+    },
+    {
+      "epoch": 1.4295672378739472,
+      "grad_norm": 0.5836344361305237,
+      "learning_rate": 0.0001,
+      "loss": 1.4836,
+      "step": 12305
+    },
+    {
+      "epoch": 1.4296834156259077,
+      "grad_norm": 0.6282364726066589,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 12306
+    },
+    {
+      "epoch": 1.4297995933778682,
+      "grad_norm": 0.5855079889297485,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 12307
+    },
+    {
+      "epoch": 1.4299157711298287,
+      "grad_norm": 0.606343150138855,
+      "learning_rate": 0.0001,
+      "loss": 1.5755,
+      "step": 12308
+    },
+    {
+      "epoch": 1.4300319488817892,
+      "grad_norm": 0.5874128341674805,
+      "learning_rate": 0.0001,
+      "loss": 1.5141,
+      "step": 12309
+    },
+    {
+      "epoch": 1.4301481266337497,
+      "grad_norm": 0.6407783031463623,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 12310
+    },
+    {
+      "epoch": 1.4302643043857102,
+      "grad_norm": 0.6048932671546936,
+      "learning_rate": 0.0001,
+      "loss": 1.4187,
+      "step": 12311
+    },
+    {
+      "epoch": 1.4303804821376707,
+      "grad_norm": 0.6171375513076782,
+      "learning_rate": 0.0001,
+      "loss": 1.4019,
+      "step": 12312
+    },
+    {
+      "epoch": 1.4304966598896312,
+      "grad_norm": 0.5970510244369507,
+      "learning_rate": 0.0001,
+      "loss": 1.3356,
+      "step": 12313
+    },
+    {
+      "epoch": 1.4306128376415916,
+      "grad_norm": 0.5908089280128479,
+      "learning_rate": 0.0001,
+      "loss": 1.4893,
+      "step": 12314
+    },
+    {
+      "epoch": 1.4307290153935521,
+      "grad_norm": 0.5906482338905334,
+      "learning_rate": 0.0001,
+      "loss": 1.4824,
+      "step": 12315
+    },
+    {
+      "epoch": 1.4308451931455126,
+      "grad_norm": 0.5727223753929138,
+      "learning_rate": 0.0001,
+      "loss": 1.2765,
+      "step": 12316
+    },
+    {
+      "epoch": 1.430961370897473,
+      "grad_norm": 0.5993191599845886,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 12317
+    },
+    {
+      "epoch": 1.4310775486494336,
+      "grad_norm": 0.6337776184082031,
+      "learning_rate": 0.0001,
+      "loss": 1.5402,
+      "step": 12318
+    },
+    {
+      "epoch": 1.431193726401394,
+      "grad_norm": 0.6048466563224792,
+      "learning_rate": 0.0001,
+      "loss": 1.3284,
+      "step": 12319
+    },
+    {
+      "epoch": 1.4313099041533546,
+      "grad_norm": 0.5903589129447937,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 12320
+    },
+    {
+      "epoch": 1.431426081905315,
+      "grad_norm": 0.5887618064880371,
+      "learning_rate": 0.0001,
+      "loss": 1.3265,
+      "step": 12321
+    },
+    {
+      "epoch": 1.4315422596572756,
+      "grad_norm": 0.6316114068031311,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 12322
+    },
+    {
+      "epoch": 1.431658437409236,
+      "grad_norm": 0.6685931086540222,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 12323
+    },
+    {
+      "epoch": 1.4317746151611965,
+      "grad_norm": 0.6843975186347961,
+      "learning_rate": 0.0001,
+      "loss": 1.6115,
+      "step": 12324
+    },
+    {
+      "epoch": 1.431890792913157,
+      "grad_norm": 0.5919124484062195,
+      "learning_rate": 0.0001,
+      "loss": 1.3703,
+      "step": 12325
+    },
+    {
+      "epoch": 1.4320069706651175,
+      "grad_norm": 0.5988925695419312,
+      "learning_rate": 0.0001,
+      "loss": 1.5675,
+      "step": 12326
+    },
+    {
+      "epoch": 1.432123148417078,
+      "grad_norm": 0.574201762676239,
+      "learning_rate": 0.0001,
+      "loss": 1.4022,
+      "step": 12327
+    },
+    {
+      "epoch": 1.4322393261690387,
+      "grad_norm": 0.6083642840385437,
+      "learning_rate": 0.0001,
+      "loss": 1.3776,
+      "step": 12328
+    },
+    {
+      "epoch": 1.4323555039209992,
+      "grad_norm": 0.5715479850769043,
+      "learning_rate": 0.0001,
+      "loss": 1.4453,
+      "step": 12329
+    },
+    {
+      "epoch": 1.4324716816729597,
+      "grad_norm": 0.6147144436836243,
+      "learning_rate": 0.0001,
+      "loss": 1.5643,
+      "step": 12330
+    },
+    {
+      "epoch": 1.4325878594249202,
+      "grad_norm": 0.6399287581443787,
+      "learning_rate": 0.0001,
+      "loss": 1.2921,
+      "step": 12331
+    },
+    {
+      "epoch": 1.4327040371768807,
+      "grad_norm": 0.6057835221290588,
+      "learning_rate": 0.0001,
+      "loss": 1.5514,
+      "step": 12332
+    },
+    {
+      "epoch": 1.4328202149288412,
+      "grad_norm": 0.6630198955535889,
+      "learning_rate": 0.0001,
+      "loss": 1.5945,
+      "step": 12333
+    },
+    {
+      "epoch": 1.4329363926808016,
+      "grad_norm": 0.6036765575408936,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 12334
+    },
+    {
+      "epoch": 1.4330525704327621,
+      "grad_norm": 0.5870527029037476,
+      "learning_rate": 0.0001,
+      "loss": 1.3427,
+      "step": 12335
+    },
+    {
+      "epoch": 1.4331687481847226,
+      "grad_norm": 0.6059634685516357,
+      "learning_rate": 0.0001,
+      "loss": 1.349,
+      "step": 12336
+    },
+    {
+      "epoch": 1.433284925936683,
+      "grad_norm": 0.5747280120849609,
+      "learning_rate": 0.0001,
+      "loss": 1.4929,
+      "step": 12337
+    },
+    {
+      "epoch": 1.4334011036886436,
+      "grad_norm": 0.6289669275283813,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 12338
+    },
+    {
+      "epoch": 1.433517281440604,
+      "grad_norm": 0.5724921226501465,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 12339
+    },
+    {
+      "epoch": 1.4336334591925646,
+      "grad_norm": 0.6268770098686218,
+      "learning_rate": 0.0001,
+      "loss": 1.4344,
+      "step": 12340
+    },
+    {
+      "epoch": 1.433749636944525,
+      "grad_norm": 0.6141247749328613,
+      "learning_rate": 0.0001,
+      "loss": 1.5246,
+      "step": 12341
+    },
+    {
+      "epoch": 1.4338658146964856,
+      "grad_norm": 0.702262282371521,
+      "learning_rate": 0.0001,
+      "loss": 1.6,
+      "step": 12342
+    },
+    {
+      "epoch": 1.433981992448446,
+      "grad_norm": 0.595066249370575,
+      "learning_rate": 0.0001,
+      "loss": 1.5602,
+      "step": 12343
+    },
+    {
+      "epoch": 1.4340981702004068,
+      "grad_norm": 0.6045823693275452,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 12344
+    },
+    {
+      "epoch": 1.4342143479523672,
+      "grad_norm": 0.5839776992797852,
+      "learning_rate": 0.0001,
+      "loss": 1.4221,
+      "step": 12345
+    },
+    {
+      "epoch": 1.4343305257043277,
+      "grad_norm": 0.6514957547187805,
+      "learning_rate": 0.0001,
+      "loss": 1.5948,
+      "step": 12346
+    },
+    {
+      "epoch": 1.4344467034562882,
+      "grad_norm": 0.5899053812026978,
+      "learning_rate": 0.0001,
+      "loss": 1.3677,
+      "step": 12347
+    },
+    {
+      "epoch": 1.4345628812082487,
+      "grad_norm": 0.5787526369094849,
+      "learning_rate": 0.0001,
+      "loss": 1.4326,
+      "step": 12348
+    },
+    {
+      "epoch": 1.4346790589602092,
+      "grad_norm": 0.5898358821868896,
+      "learning_rate": 0.0001,
+      "loss": 1.4003,
+      "step": 12349
+    },
+    {
+      "epoch": 1.4347952367121697,
+      "grad_norm": 0.6458524465560913,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 12350
+    },
+    {
+      "epoch": 1.4349114144641302,
+      "grad_norm": 0.646244466304779,
+      "learning_rate": 0.0001,
+      "loss": 1.5125,
+      "step": 12351
+    },
+    {
+      "epoch": 1.4350275922160907,
+      "grad_norm": 0.5955740213394165,
+      "learning_rate": 0.0001,
+      "loss": 1.3983,
+      "step": 12352
+    },
+    {
+      "epoch": 1.4351437699680512,
+      "grad_norm": 0.5988348722457886,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 12353
+    },
+    {
+      "epoch": 1.4352599477200116,
+      "grad_norm": 0.5847843885421753,
+      "learning_rate": 0.0001,
+      "loss": 1.295,
+      "step": 12354
+    },
+    {
+      "epoch": 1.4353761254719721,
+      "grad_norm": 0.6146335005760193,
+      "learning_rate": 0.0001,
+      "loss": 1.5285,
+      "step": 12355
+    },
+    {
+      "epoch": 1.4354923032239326,
+      "grad_norm": 0.604469895362854,
+      "learning_rate": 0.0001,
+      "loss": 1.4477,
+      "step": 12356
+    },
+    {
+      "epoch": 1.435608480975893,
+      "grad_norm": 0.6842860579490662,
+      "learning_rate": 0.0001,
+      "loss": 1.5657,
+      "step": 12357
+    },
+    {
+      "epoch": 1.4357246587278536,
+      "grad_norm": 0.5877547860145569,
+      "learning_rate": 0.0001,
+      "loss": 1.4932,
+      "step": 12358
+    },
+    {
+      "epoch": 1.435840836479814,
+      "grad_norm": 0.687014102935791,
+      "learning_rate": 0.0001,
+      "loss": 1.5146,
+      "step": 12359
+    },
+    {
+      "epoch": 1.4359570142317746,
+      "grad_norm": 0.5968105792999268,
+      "learning_rate": 0.0001,
+      "loss": 1.4974,
+      "step": 12360
+    },
+    {
+      "epoch": 1.436073191983735,
+      "grad_norm": 0.5516940355300903,
+      "learning_rate": 0.0001,
+      "loss": 1.3209,
+      "step": 12361
+    },
+    {
+      "epoch": 1.4361893697356956,
+      "grad_norm": 0.6110051274299622,
+      "learning_rate": 0.0001,
+      "loss": 1.5701,
+      "step": 12362
+    },
+    {
+      "epoch": 1.436305547487656,
+      "grad_norm": 0.5829412937164307,
+      "learning_rate": 0.0001,
+      "loss": 1.4339,
+      "step": 12363
+    },
+    {
+      "epoch": 1.4364217252396165,
+      "grad_norm": 0.6736522912979126,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 12364
+    },
+    {
+      "epoch": 1.436537902991577,
+      "grad_norm": 0.5804245471954346,
+      "learning_rate": 0.0001,
+      "loss": 1.3728,
+      "step": 12365
+    },
+    {
+      "epoch": 1.4366540807435375,
+      "grad_norm": 0.6237603425979614,
+      "learning_rate": 0.0001,
+      "loss": 1.5145,
+      "step": 12366
+    },
+    {
+      "epoch": 1.436770258495498,
+      "grad_norm": 0.6224261522293091,
+      "learning_rate": 0.0001,
+      "loss": 1.4202,
+      "step": 12367
+    },
+    {
+      "epoch": 1.4368864362474585,
+      "grad_norm": 0.599617600440979,
+      "learning_rate": 0.0001,
+      "loss": 1.3436,
+      "step": 12368
+    },
+    {
+      "epoch": 1.437002613999419,
+      "grad_norm": 0.5778672695159912,
+      "learning_rate": 0.0001,
+      "loss": 1.3741,
+      "step": 12369
+    },
+    {
+      "epoch": 1.4371187917513797,
+      "grad_norm": 0.6169531941413879,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 12370
+    },
+    {
+      "epoch": 1.4372349695033402,
+      "grad_norm": 0.6243462562561035,
+      "learning_rate": 0.0001,
+      "loss": 1.4946,
+      "step": 12371
+    },
+    {
+      "epoch": 1.4373511472553007,
+      "grad_norm": 0.6082959175109863,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 12372
+    },
+    {
+      "epoch": 1.4374673250072612,
+      "grad_norm": 0.5401190519332886,
+      "learning_rate": 0.0001,
+      "loss": 1.303,
+      "step": 12373
+    },
+    {
+      "epoch": 1.4375835027592216,
+      "grad_norm": 0.5593122243881226,
+      "learning_rate": 0.0001,
+      "loss": 1.3716,
+      "step": 12374
+    },
+    {
+      "epoch": 1.4376996805111821,
+      "grad_norm": 0.6530990600585938,
+      "learning_rate": 0.0001,
+      "loss": 1.4847,
+      "step": 12375
+    },
+    {
+      "epoch": 1.4378158582631426,
+      "grad_norm": 0.5983834266662598,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 12376
+    },
+    {
+      "epoch": 1.437932036015103,
+      "grad_norm": 0.6394164562225342,
+      "learning_rate": 0.0001,
+      "loss": 1.5291,
+      "step": 12377
+    },
+    {
+      "epoch": 1.4380482137670636,
+      "grad_norm": 0.5918828845024109,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 12378
+    },
+    {
+      "epoch": 1.438164391519024,
+      "grad_norm": 0.5928050875663757,
+      "learning_rate": 0.0001,
+      "loss": 1.6254,
+      "step": 12379
+    },
+    {
+      "epoch": 1.4382805692709846,
+      "grad_norm": 0.6321570873260498,
+      "learning_rate": 0.0001,
+      "loss": 1.499,
+      "step": 12380
+    },
+    {
+      "epoch": 1.438396747022945,
+      "grad_norm": 0.5853205323219299,
+      "learning_rate": 0.0001,
+      "loss": 1.4488,
+      "step": 12381
+    },
+    {
+      "epoch": 1.4385129247749056,
+      "grad_norm": 0.5675798058509827,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 12382
+    },
+    {
+      "epoch": 1.438629102526866,
+      "grad_norm": 0.6351810097694397,
+      "learning_rate": 0.0001,
+      "loss": 1.5845,
+      "step": 12383
+    },
+    {
+      "epoch": 1.4387452802788265,
+      "grad_norm": 0.575139582157135,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 12384
+    },
+    {
+      "epoch": 1.438861458030787,
+      "grad_norm": 0.625085711479187,
+      "learning_rate": 0.0001,
+      "loss": 1.5833,
+      "step": 12385
+    },
+    {
+      "epoch": 1.4389776357827477,
+      "grad_norm": 0.5660273432731628,
+      "learning_rate": 0.0001,
+      "loss": 1.5369,
+      "step": 12386
+    },
+    {
+      "epoch": 1.4390938135347082,
+      "grad_norm": 0.642395555973053,
+      "learning_rate": 0.0001,
+      "loss": 1.6152,
+      "step": 12387
+    },
+    {
+      "epoch": 1.4392099912866687,
+      "grad_norm": 0.5889819860458374,
+      "learning_rate": 0.0001,
+      "loss": 1.4482,
+      "step": 12388
+    },
+    {
+      "epoch": 1.4393261690386292,
+      "grad_norm": 0.5822515487670898,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 12389
+    },
+    {
+      "epoch": 1.4394423467905897,
+      "grad_norm": 0.5766950249671936,
+      "learning_rate": 0.0001,
+      "loss": 1.4415,
+      "step": 12390
+    },
+    {
+      "epoch": 1.4395585245425502,
+      "grad_norm": 0.6112682819366455,
+      "learning_rate": 0.0001,
+      "loss": 1.3334,
+      "step": 12391
+    },
+    {
+      "epoch": 1.4396747022945107,
+      "grad_norm": 0.6508562564849854,
+      "learning_rate": 0.0001,
+      "loss": 1.4605,
+      "step": 12392
+    },
+    {
+      "epoch": 1.4397908800464712,
+      "grad_norm": 0.6107341647148132,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 12393
+    },
+    {
+      "epoch": 1.4399070577984316,
+      "grad_norm": 0.5927090048789978,
+      "learning_rate": 0.0001,
+      "loss": 1.3687,
+      "step": 12394
+    },
+    {
+      "epoch": 1.4400232355503921,
+      "grad_norm": 0.6269469261169434,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 12395
+    },
+    {
+      "epoch": 1.4401394133023526,
+      "grad_norm": 0.6517143249511719,
+      "learning_rate": 0.0001,
+      "loss": 1.4674,
+      "step": 12396
+    },
+    {
+      "epoch": 1.4402555910543131,
+      "grad_norm": 0.5869902968406677,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 12397
+    },
+    {
+      "epoch": 1.4403717688062736,
+      "grad_norm": 0.6312974691390991,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 12398
+    },
+    {
+      "epoch": 1.440487946558234,
+      "grad_norm": 0.5878767371177673,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 12399
+    },
+    {
+      "epoch": 1.4406041243101946,
+      "grad_norm": 0.6097482442855835,
+      "learning_rate": 0.0001,
+      "loss": 1.5549,
+      "step": 12400
+    },
+    {
+      "epoch": 1.440720302062155,
+      "grad_norm": 0.5648863315582275,
+      "learning_rate": 0.0001,
+      "loss": 1.2942,
+      "step": 12401
+    },
+    {
+      "epoch": 1.4408364798141156,
+      "grad_norm": 0.5527321100234985,
+      "learning_rate": 0.0001,
+      "loss": 1.2734,
+      "step": 12402
+    },
+    {
+      "epoch": 1.440952657566076,
+      "grad_norm": 0.587592601776123,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 12403
+    },
+    {
+      "epoch": 1.4410688353180365,
+      "grad_norm": 0.6088840961456299,
+      "learning_rate": 0.0001,
+      "loss": 1.6002,
+      "step": 12404
+    },
+    {
+      "epoch": 1.441185013069997,
+      "grad_norm": 0.5852679014205933,
+      "learning_rate": 0.0001,
+      "loss": 1.3525,
+      "step": 12405
+    },
+    {
+      "epoch": 1.4413011908219575,
+      "grad_norm": 0.65325528383255,
+      "learning_rate": 0.0001,
+      "loss": 1.3384,
+      "step": 12406
+    },
+    {
+      "epoch": 1.441417368573918,
+      "grad_norm": 0.6041934490203857,
+      "learning_rate": 0.0001,
+      "loss": 1.4435,
+      "step": 12407
+    },
+    {
+      "epoch": 1.4415335463258785,
+      "grad_norm": 0.6076403856277466,
+      "learning_rate": 0.0001,
+      "loss": 1.3845,
+      "step": 12408
+    },
+    {
+      "epoch": 1.441649724077839,
+      "grad_norm": 0.6228477954864502,
+      "learning_rate": 0.0001,
+      "loss": 1.6537,
+      "step": 12409
+    },
+    {
+      "epoch": 1.4417659018297995,
+      "grad_norm": 0.6112052202224731,
+      "learning_rate": 0.0001,
+      "loss": 1.5046,
+      "step": 12410
+    },
+    {
+      "epoch": 1.44188207958176,
+      "grad_norm": 0.5732336640357971,
+      "learning_rate": 0.0001,
+      "loss": 1.2829,
+      "step": 12411
+    },
+    {
+      "epoch": 1.4419982573337207,
+      "grad_norm": 0.6284494400024414,
+      "learning_rate": 0.0001,
+      "loss": 1.6029,
+      "step": 12412
+    },
+    {
+      "epoch": 1.4421144350856812,
+      "grad_norm": 0.5829569101333618,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 12413
+    },
+    {
+      "epoch": 1.4422306128376416,
+      "grad_norm": 0.6057117581367493,
+      "learning_rate": 0.0001,
+      "loss": 1.4082,
+      "step": 12414
+    },
+    {
+      "epoch": 1.4423467905896021,
+      "grad_norm": 0.6320570111274719,
+      "learning_rate": 0.0001,
+      "loss": 1.5682,
+      "step": 12415
+    },
+    {
+      "epoch": 1.4424629683415626,
+      "grad_norm": 0.6196115612983704,
+      "learning_rate": 0.0001,
+      "loss": 1.5874,
+      "step": 12416
+    },
+    {
+      "epoch": 1.4425791460935231,
+      "grad_norm": 0.5623551607131958,
+      "learning_rate": 0.0001,
+      "loss": 1.3886,
+      "step": 12417
+    },
+    {
+      "epoch": 1.4426953238454836,
+      "grad_norm": 0.6607520580291748,
+      "learning_rate": 0.0001,
+      "loss": 1.6463,
+      "step": 12418
+    },
+    {
+      "epoch": 1.442811501597444,
+      "grad_norm": 0.617299497127533,
+      "learning_rate": 0.0001,
+      "loss": 1.3807,
+      "step": 12419
+    },
+    {
+      "epoch": 1.4429276793494046,
+      "grad_norm": 0.6651842594146729,
+      "learning_rate": 0.0001,
+      "loss": 1.7093,
+      "step": 12420
+    },
+    {
+      "epoch": 1.443043857101365,
+      "grad_norm": 0.5932329893112183,
+      "learning_rate": 0.0001,
+      "loss": 1.4479,
+      "step": 12421
+    },
+    {
+      "epoch": 1.4431600348533256,
+      "grad_norm": 0.6579514145851135,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 12422
+    },
+    {
+      "epoch": 1.443276212605286,
+      "grad_norm": 0.664814829826355,
+      "learning_rate": 0.0001,
+      "loss": 1.4231,
+      "step": 12423
+    },
+    {
+      "epoch": 1.4433923903572465,
+      "grad_norm": 0.6053681373596191,
+      "learning_rate": 0.0001,
+      "loss": 1.4038,
+      "step": 12424
+    },
+    {
+      "epoch": 1.443508568109207,
+      "grad_norm": 0.5771030187606812,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 12425
+    },
+    {
+      "epoch": 1.4436247458611675,
+      "grad_norm": 0.594408392906189,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 12426
+    },
+    {
+      "epoch": 1.4437409236131282,
+      "grad_norm": 0.6161181330680847,
+      "learning_rate": 0.0001,
+      "loss": 1.3431,
+      "step": 12427
+    },
+    {
+      "epoch": 1.4438571013650887,
+      "grad_norm": 0.6196521520614624,
+      "learning_rate": 0.0001,
+      "loss": 1.5768,
+      "step": 12428
+    },
+    {
+      "epoch": 1.4439732791170492,
+      "grad_norm": 0.6354605555534363,
+      "learning_rate": 0.0001,
+      "loss": 1.4162,
+      "step": 12429
+    },
+    {
+      "epoch": 1.4440894568690097,
+      "grad_norm": 0.6167144179344177,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 12430
+    },
+    {
+      "epoch": 1.4442056346209702,
+      "grad_norm": 0.6271317601203918,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 12431
+    },
+    {
+      "epoch": 1.4443218123729307,
+      "grad_norm": 0.6042360067367554,
+      "learning_rate": 0.0001,
+      "loss": 1.547,
+      "step": 12432
+    },
+    {
+      "epoch": 1.4444379901248912,
+      "grad_norm": 0.5653975605964661,
+      "learning_rate": 0.0001,
+      "loss": 1.3804,
+      "step": 12433
+    },
+    {
+      "epoch": 1.4445541678768516,
+      "grad_norm": 0.5969522595405579,
+      "learning_rate": 0.0001,
+      "loss": 1.5427,
+      "step": 12434
+    },
+    {
+      "epoch": 1.4446703456288121,
+      "grad_norm": 0.5756595730781555,
+      "learning_rate": 0.0001,
+      "loss": 1.4859,
+      "step": 12435
+    },
+    {
+      "epoch": 1.4447865233807726,
+      "grad_norm": 0.5508007407188416,
+      "learning_rate": 0.0001,
+      "loss": 1.3658,
+      "step": 12436
+    },
+    {
+      "epoch": 1.4449027011327331,
+      "grad_norm": 0.578050971031189,
+      "learning_rate": 0.0001,
+      "loss": 1.4602,
+      "step": 12437
+    },
+    {
+      "epoch": 1.4450188788846936,
+      "grad_norm": 0.5895441174507141,
+      "learning_rate": 0.0001,
+      "loss": 1.3945,
+      "step": 12438
+    },
+    {
+      "epoch": 1.445135056636654,
+      "grad_norm": 0.5890359282493591,
+      "learning_rate": 0.0001,
+      "loss": 1.3473,
+      "step": 12439
+    },
+    {
+      "epoch": 1.4452512343886146,
+      "grad_norm": 0.5464291572570801,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 12440
+    },
+    {
+      "epoch": 1.445367412140575,
+      "grad_norm": 0.6562072038650513,
+      "learning_rate": 0.0001,
+      "loss": 1.5293,
+      "step": 12441
+    },
+    {
+      "epoch": 1.4454835898925356,
+      "grad_norm": 0.5997947454452515,
+      "learning_rate": 0.0001,
+      "loss": 1.5211,
+      "step": 12442
+    },
+    {
+      "epoch": 1.445599767644496,
+      "grad_norm": 0.5633268356323242,
+      "learning_rate": 0.0001,
+      "loss": 1.2866,
+      "step": 12443
+    },
+    {
+      "epoch": 1.4457159453964565,
+      "grad_norm": 0.5751778483390808,
+      "learning_rate": 0.0001,
+      "loss": 1.3926,
+      "step": 12444
+    },
+    {
+      "epoch": 1.445832123148417,
+      "grad_norm": 0.6264200210571289,
+      "learning_rate": 0.0001,
+      "loss": 1.4634,
+      "step": 12445
+    },
+    {
+      "epoch": 1.4459483009003775,
+      "grad_norm": 0.5811033844947815,
+      "learning_rate": 0.0001,
+      "loss": 1.3373,
+      "step": 12446
+    },
+    {
+      "epoch": 1.446064478652338,
+      "grad_norm": 0.5755273103713989,
+      "learning_rate": 0.0001,
+      "loss": 1.4198,
+      "step": 12447
+    },
+    {
+      "epoch": 1.4461806564042985,
+      "grad_norm": 0.6127162575721741,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 12448
+    },
+    {
+      "epoch": 1.446296834156259,
+      "grad_norm": 0.631125271320343,
+      "learning_rate": 0.0001,
+      "loss": 1.5738,
+      "step": 12449
+    },
+    {
+      "epoch": 1.4464130119082195,
+      "grad_norm": 0.6190559267997742,
+      "learning_rate": 0.0001,
+      "loss": 1.6324,
+      "step": 12450
+    },
+    {
+      "epoch": 1.44652918966018,
+      "grad_norm": 0.6076474189758301,
+      "learning_rate": 0.0001,
+      "loss": 1.4934,
+      "step": 12451
+    },
+    {
+      "epoch": 1.4466453674121404,
+      "grad_norm": 0.6262378692626953,
+      "learning_rate": 0.0001,
+      "loss": 1.5541,
+      "step": 12452
+    },
+    {
+      "epoch": 1.446761545164101,
+      "grad_norm": 0.6727213859558105,
+      "learning_rate": 0.0001,
+      "loss": 1.6884,
+      "step": 12453
+    },
+    {
+      "epoch": 1.4468777229160616,
+      "grad_norm": 0.5651970505714417,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 12454
+    },
+    {
+      "epoch": 1.4469939006680221,
+      "grad_norm": 0.5852252244949341,
+      "learning_rate": 0.0001,
+      "loss": 1.5358,
+      "step": 12455
+    },
+    {
+      "epoch": 1.4471100784199826,
+      "grad_norm": 0.6040740609169006,
+      "learning_rate": 0.0001,
+      "loss": 1.3637,
+      "step": 12456
+    },
+    {
+      "epoch": 1.4472262561719431,
+      "grad_norm": 0.6351321935653687,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 12457
+    },
+    {
+      "epoch": 1.4473424339239036,
+      "grad_norm": 0.5973447561264038,
+      "learning_rate": 0.0001,
+      "loss": 1.5772,
+      "step": 12458
+    },
+    {
+      "epoch": 1.447458611675864,
+      "grad_norm": 0.5978641510009766,
+      "learning_rate": 0.0001,
+      "loss": 1.5503,
+      "step": 12459
+    },
+    {
+      "epoch": 1.4475747894278246,
+      "grad_norm": 0.6435196995735168,
+      "learning_rate": 0.0001,
+      "loss": 1.6186,
+      "step": 12460
+    },
+    {
+      "epoch": 1.447690967179785,
+      "grad_norm": 0.5686025619506836,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 12461
+    },
+    {
+      "epoch": 1.4478071449317456,
+      "grad_norm": 0.5564260482788086,
+      "learning_rate": 0.0001,
+      "loss": 1.2317,
+      "step": 12462
+    },
+    {
+      "epoch": 1.447923322683706,
+      "grad_norm": 0.6230119466781616,
+      "learning_rate": 0.0001,
+      "loss": 1.49,
+      "step": 12463
+    },
+    {
+      "epoch": 1.4480395004356665,
+      "grad_norm": 0.5975537896156311,
+      "learning_rate": 0.0001,
+      "loss": 1.4125,
+      "step": 12464
+    },
+    {
+      "epoch": 1.448155678187627,
+      "grad_norm": 0.5980433821678162,
+      "learning_rate": 0.0001,
+      "loss": 1.444,
+      "step": 12465
+    },
+    {
+      "epoch": 1.4482718559395875,
+      "grad_norm": 0.5679630041122437,
+      "learning_rate": 0.0001,
+      "loss": 1.2907,
+      "step": 12466
+    },
+    {
+      "epoch": 1.448388033691548,
+      "grad_norm": 0.643408477306366,
+      "learning_rate": 0.0001,
+      "loss": 1.5006,
+      "step": 12467
+    },
+    {
+      "epoch": 1.4485042114435085,
+      "grad_norm": 0.6029306054115295,
+      "learning_rate": 0.0001,
+      "loss": 1.4144,
+      "step": 12468
+    },
+    {
+      "epoch": 1.4486203891954692,
+      "grad_norm": 0.624788224697113,
+      "learning_rate": 0.0001,
+      "loss": 1.4684,
+      "step": 12469
+    },
+    {
+      "epoch": 1.4487365669474297,
+      "grad_norm": 0.6190298199653625,
+      "learning_rate": 0.0001,
+      "loss": 1.2963,
+      "step": 12470
+    },
+    {
+      "epoch": 1.4488527446993902,
+      "grad_norm": 0.6062245965003967,
+      "learning_rate": 0.0001,
+      "loss": 1.5758,
+      "step": 12471
+    },
+    {
+      "epoch": 1.4489689224513507,
+      "grad_norm": 0.6660163402557373,
+      "learning_rate": 0.0001,
+      "loss": 1.5742,
+      "step": 12472
+    },
+    {
+      "epoch": 1.4490851002033112,
+      "grad_norm": 0.5759906768798828,
+      "learning_rate": 0.0001,
+      "loss": 1.3949,
+      "step": 12473
+    },
+    {
+      "epoch": 1.4492012779552716,
+      "grad_norm": 0.6217337846755981,
+      "learning_rate": 0.0001,
+      "loss": 1.483,
+      "step": 12474
+    },
+    {
+      "epoch": 1.4493174557072321,
+      "grad_norm": 0.5870078802108765,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 12475
+    },
+    {
+      "epoch": 1.4494336334591926,
+      "grad_norm": 0.5740808844566345,
+      "learning_rate": 0.0001,
+      "loss": 1.2145,
+      "step": 12476
+    },
+    {
+      "epoch": 1.4495498112111531,
+      "grad_norm": 0.603854775428772,
+      "learning_rate": 0.0001,
+      "loss": 1.5352,
+      "step": 12477
+    },
+    {
+      "epoch": 1.4496659889631136,
+      "grad_norm": 0.5887278914451599,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 12478
+    },
+    {
+      "epoch": 1.449782166715074,
+      "grad_norm": 0.5863503217697144,
+      "learning_rate": 0.0001,
+      "loss": 1.4259,
+      "step": 12479
+    },
+    {
+      "epoch": 1.4498983444670346,
+      "grad_norm": 0.5903075337409973,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 12480
+    },
+    {
+      "epoch": 1.450014522218995,
+      "grad_norm": 0.6201311945915222,
+      "learning_rate": 0.0001,
+      "loss": 1.4628,
+      "step": 12481
+    },
+    {
+      "epoch": 1.4501306999709556,
+      "grad_norm": 0.6051203608512878,
+      "learning_rate": 0.0001,
+      "loss": 1.4167,
+      "step": 12482
+    },
+    {
+      "epoch": 1.450246877722916,
+      "grad_norm": 0.6963009834289551,
+      "learning_rate": 0.0001,
+      "loss": 1.5327,
+      "step": 12483
+    },
+    {
+      "epoch": 1.4503630554748765,
+      "grad_norm": 0.6066775918006897,
+      "learning_rate": 0.0001,
+      "loss": 1.401,
+      "step": 12484
+    },
+    {
+      "epoch": 1.450479233226837,
+      "grad_norm": 0.5624590516090393,
+      "learning_rate": 0.0001,
+      "loss": 1.2795,
+      "step": 12485
+    },
+    {
+      "epoch": 1.4505954109787975,
+      "grad_norm": 0.5763733983039856,
+      "learning_rate": 0.0001,
+      "loss": 1.3443,
+      "step": 12486
+    },
+    {
+      "epoch": 1.450711588730758,
+      "grad_norm": 0.6045727729797363,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 12487
+    },
+    {
+      "epoch": 1.4508277664827185,
+      "grad_norm": 0.6144154667854309,
+      "learning_rate": 0.0001,
+      "loss": 1.4593,
+      "step": 12488
+    },
+    {
+      "epoch": 1.450943944234679,
+      "grad_norm": 0.6358152031898499,
+      "learning_rate": 0.0001,
+      "loss": 1.5782,
+      "step": 12489
+    },
+    {
+      "epoch": 1.4510601219866395,
+      "grad_norm": 0.5821871161460876,
+      "learning_rate": 0.0001,
+      "loss": 1.4008,
+      "step": 12490
+    },
+    {
+      "epoch": 1.4511762997386,
+      "grad_norm": 0.6219722032546997,
+      "learning_rate": 0.0001,
+      "loss": 1.4496,
+      "step": 12491
+    },
+    {
+      "epoch": 1.4512924774905605,
+      "grad_norm": 0.5885584354400635,
+      "learning_rate": 0.0001,
+      "loss": 1.3345,
+      "step": 12492
+    },
+    {
+      "epoch": 1.451408655242521,
+      "grad_norm": 0.6406071782112122,
+      "learning_rate": 0.0001,
+      "loss": 1.5645,
+      "step": 12493
+    },
+    {
+      "epoch": 1.4515248329944814,
+      "grad_norm": 0.5714595317840576,
+      "learning_rate": 0.0001,
+      "loss": 1.2703,
+      "step": 12494
+    },
+    {
+      "epoch": 1.451641010746442,
+      "grad_norm": 0.6086097955703735,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 12495
+    },
+    {
+      "epoch": 1.4517571884984026,
+      "grad_norm": 0.6160340309143066,
+      "learning_rate": 0.0001,
+      "loss": 1.7256,
+      "step": 12496
+    },
+    {
+      "epoch": 1.4518733662503631,
+      "grad_norm": 0.6223379969596863,
+      "learning_rate": 0.0001,
+      "loss": 1.5614,
+      "step": 12497
+    },
+    {
+      "epoch": 1.4519895440023236,
+      "grad_norm": 0.6125460267066956,
+      "learning_rate": 0.0001,
+      "loss": 1.3264,
+      "step": 12498
+    },
+    {
+      "epoch": 1.452105721754284,
+      "grad_norm": 0.6102118492126465,
+      "learning_rate": 0.0001,
+      "loss": 1.3831,
+      "step": 12499
+    },
+    {
+      "epoch": 1.4522218995062446,
+      "grad_norm": 0.6099421381950378,
+      "learning_rate": 0.0001,
+      "loss": 1.3743,
+      "step": 12500
+    },
+    {
+      "epoch": 1.452338077258205,
+      "grad_norm": 0.559723436832428,
+      "learning_rate": 0.0001,
+      "loss": 1.3618,
+      "step": 12501
+    },
+    {
+      "epoch": 1.4524542550101656,
+      "grad_norm": 0.6058681011199951,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 12502
+    },
+    {
+      "epoch": 1.452570432762126,
+      "grad_norm": 0.6532595753669739,
+      "learning_rate": 0.0001,
+      "loss": 1.6282,
+      "step": 12503
+    },
+    {
+      "epoch": 1.4526866105140865,
+      "grad_norm": 0.6153339743614197,
+      "learning_rate": 0.0001,
+      "loss": 1.5724,
+      "step": 12504
+    },
+    {
+      "epoch": 1.452802788266047,
+      "grad_norm": 0.5688385963439941,
+      "learning_rate": 0.0001,
+      "loss": 1.2545,
+      "step": 12505
+    },
+    {
+      "epoch": 1.4529189660180075,
+      "grad_norm": 0.6670095920562744,
+      "learning_rate": 0.0001,
+      "loss": 1.6536,
+      "step": 12506
+    },
+    {
+      "epoch": 1.453035143769968,
+      "grad_norm": 0.6240342259407043,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 12507
+    },
+    {
+      "epoch": 1.4531513215219285,
+      "grad_norm": 0.6448224186897278,
+      "learning_rate": 0.0001,
+      "loss": 1.629,
+      "step": 12508
+    },
+    {
+      "epoch": 1.453267499273889,
+      "grad_norm": 0.6262985467910767,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 12509
+    },
+    {
+      "epoch": 1.4533836770258495,
+      "grad_norm": 0.580199122428894,
+      "learning_rate": 0.0001,
+      "loss": 1.3952,
+      "step": 12510
+    },
+    {
+      "epoch": 1.4534998547778102,
+      "grad_norm": 0.5817636847496033,
+      "learning_rate": 0.0001,
+      "loss": 1.3002,
+      "step": 12511
+    },
+    {
+      "epoch": 1.4536160325297707,
+      "grad_norm": 0.6186819672584534,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 12512
+    },
+    {
+      "epoch": 1.4537322102817312,
+      "grad_norm": 0.6286445260047913,
+      "learning_rate": 0.0001,
+      "loss": 1.4373,
+      "step": 12513
+    },
+    {
+      "epoch": 1.4538483880336917,
+      "grad_norm": 0.6268229484558105,
+      "learning_rate": 0.0001,
+      "loss": 1.5821,
+      "step": 12514
+    },
+    {
+      "epoch": 1.4539645657856521,
+      "grad_norm": 0.5959900617599487,
+      "learning_rate": 0.0001,
+      "loss": 1.4463,
+      "step": 12515
+    },
+    {
+      "epoch": 1.4540807435376126,
+      "grad_norm": 0.6190244555473328,
+      "learning_rate": 0.0001,
+      "loss": 1.3746,
+      "step": 12516
+    },
+    {
+      "epoch": 1.4541969212895731,
+      "grad_norm": 0.6077380180358887,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 12517
+    },
+    {
+      "epoch": 1.4543130990415336,
+      "grad_norm": 0.6312915086746216,
+      "learning_rate": 0.0001,
+      "loss": 1.5431,
+      "step": 12518
+    },
+    {
+      "epoch": 1.454429276793494,
+      "grad_norm": 0.642156720161438,
+      "learning_rate": 0.0001,
+      "loss": 1.6197,
+      "step": 12519
+    },
+    {
+      "epoch": 1.4545454545454546,
+      "grad_norm": 0.6618093252182007,
+      "learning_rate": 0.0001,
+      "loss": 1.4224,
+      "step": 12520
+    },
+    {
+      "epoch": 1.454661632297415,
+      "grad_norm": 0.61574786901474,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 12521
+    },
+    {
+      "epoch": 1.4547778100493756,
+      "grad_norm": 0.5520590543746948,
+      "learning_rate": 0.0001,
+      "loss": 1.1745,
+      "step": 12522
+    },
+    {
+      "epoch": 1.454893987801336,
+      "grad_norm": 0.6121432185173035,
+      "learning_rate": 0.0001,
+      "loss": 1.4132,
+      "step": 12523
+    },
+    {
+      "epoch": 1.4550101655532965,
+      "grad_norm": 0.6063690185546875,
+      "learning_rate": 0.0001,
+      "loss": 1.3249,
+      "step": 12524
+    },
+    {
+      "epoch": 1.455126343305257,
+      "grad_norm": 0.6351370811462402,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 12525
+    },
+    {
+      "epoch": 1.4552425210572175,
+      "grad_norm": 0.5611215829849243,
+      "learning_rate": 0.0001,
+      "loss": 1.5057,
+      "step": 12526
+    },
+    {
+      "epoch": 1.455358698809178,
+      "grad_norm": 0.6196964383125305,
+      "learning_rate": 0.0001,
+      "loss": 1.4851,
+      "step": 12527
+    },
+    {
+      "epoch": 1.4554748765611385,
+      "grad_norm": 0.5891339778900146,
+      "learning_rate": 0.0001,
+      "loss": 1.3492,
+      "step": 12528
+    },
+    {
+      "epoch": 1.455591054313099,
+      "grad_norm": 0.634013295173645,
+      "learning_rate": 0.0001,
+      "loss": 1.5088,
+      "step": 12529
+    },
+    {
+      "epoch": 1.4557072320650595,
+      "grad_norm": 0.5923905968666077,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 12530
+    },
+    {
+      "epoch": 1.45582340981702,
+      "grad_norm": 0.6321912407875061,
+      "learning_rate": 0.0001,
+      "loss": 1.5716,
+      "step": 12531
+    },
+    {
+      "epoch": 1.4559395875689805,
+      "grad_norm": 0.5776755213737488,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 12532
+    },
+    {
+      "epoch": 1.456055765320941,
+      "grad_norm": 0.5691312551498413,
+      "learning_rate": 0.0001,
+      "loss": 1.4853,
+      "step": 12533
+    },
+    {
+      "epoch": 1.4561719430729014,
+      "grad_norm": 0.6355810165405273,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 12534
+    },
+    {
+      "epoch": 1.456288120824862,
+      "grad_norm": 0.6114268898963928,
+      "learning_rate": 0.0001,
+      "loss": 1.5544,
+      "step": 12535
+    },
+    {
+      "epoch": 1.4564042985768224,
+      "grad_norm": 0.5496631860733032,
+      "learning_rate": 0.0001,
+      "loss": 1.3485,
+      "step": 12536
+    },
+    {
+      "epoch": 1.4565204763287831,
+      "grad_norm": 0.6105073094367981,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 12537
+    },
+    {
+      "epoch": 1.4566366540807436,
+      "grad_norm": 0.5859121680259705,
+      "learning_rate": 0.0001,
+      "loss": 1.4094,
+      "step": 12538
+    },
+    {
+      "epoch": 1.456752831832704,
+      "grad_norm": 0.6114695072174072,
+      "learning_rate": 0.0001,
+      "loss": 1.4062,
+      "step": 12539
+    },
+    {
+      "epoch": 1.4568690095846646,
+      "grad_norm": 0.6075177788734436,
+      "learning_rate": 0.0001,
+      "loss": 1.3643,
+      "step": 12540
+    },
+    {
+      "epoch": 1.456985187336625,
+      "grad_norm": 0.5910870432853699,
+      "learning_rate": 0.0001,
+      "loss": 1.4241,
+      "step": 12541
+    },
+    {
+      "epoch": 1.4571013650885856,
+      "grad_norm": 0.595967710018158,
+      "learning_rate": 0.0001,
+      "loss": 1.4613,
+      "step": 12542
+    },
+    {
+      "epoch": 1.457217542840546,
+      "grad_norm": 0.6669548749923706,
+      "learning_rate": 0.0001,
+      "loss": 1.6616,
+      "step": 12543
+    },
+    {
+      "epoch": 1.4573337205925065,
+      "grad_norm": 0.6057718396186829,
+      "learning_rate": 0.0001,
+      "loss": 1.5482,
+      "step": 12544
+    },
+    {
+      "epoch": 1.457449898344467,
+      "grad_norm": 0.5876947641372681,
+      "learning_rate": 0.0001,
+      "loss": 1.4503,
+      "step": 12545
+    },
+    {
+      "epoch": 1.4575660760964275,
+      "grad_norm": 0.6106549501419067,
+      "learning_rate": 0.0001,
+      "loss": 1.4327,
+      "step": 12546
+    },
+    {
+      "epoch": 1.457682253848388,
+      "grad_norm": 0.6266193985939026,
+      "learning_rate": 0.0001,
+      "loss": 1.2981,
+      "step": 12547
+    },
+    {
+      "epoch": 1.4577984316003485,
+      "grad_norm": 0.6059079170227051,
+      "learning_rate": 0.0001,
+      "loss": 1.5158,
+      "step": 12548
+    },
+    {
+      "epoch": 1.457914609352309,
+      "grad_norm": 0.5994994640350342,
+      "learning_rate": 0.0001,
+      "loss": 1.6222,
+      "step": 12549
+    },
+    {
+      "epoch": 1.4580307871042695,
+      "grad_norm": 0.6154032945632935,
+      "learning_rate": 0.0001,
+      "loss": 1.5556,
+      "step": 12550
+    },
+    {
+      "epoch": 1.45814696485623,
+      "grad_norm": 0.6104035377502441,
+      "learning_rate": 0.0001,
+      "loss": 1.5981,
+      "step": 12551
+    },
+    {
+      "epoch": 1.4582631426081905,
+      "grad_norm": 0.6142469644546509,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 12552
+    },
+    {
+      "epoch": 1.4583793203601512,
+      "grad_norm": 0.574055552482605,
+      "learning_rate": 0.0001,
+      "loss": 1.309,
+      "step": 12553
+    },
+    {
+      "epoch": 1.4584954981121117,
+      "grad_norm": 0.5898140668869019,
+      "learning_rate": 0.0001,
+      "loss": 1.3786,
+      "step": 12554
+    },
+    {
+      "epoch": 1.4586116758640721,
+      "grad_norm": 0.5797202587127686,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 12555
+    },
+    {
+      "epoch": 1.4587278536160326,
+      "grad_norm": 0.554664134979248,
+      "learning_rate": 0.0001,
+      "loss": 1.3166,
+      "step": 12556
+    },
+    {
+      "epoch": 1.4588440313679931,
+      "grad_norm": 0.6085032820701599,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 12557
+    },
+    {
+      "epoch": 1.4589602091199536,
+      "grad_norm": 0.625577986240387,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 12558
+    },
+    {
+      "epoch": 1.459076386871914,
+      "grad_norm": 0.6222672462463379,
+      "learning_rate": 0.0001,
+      "loss": 1.5968,
+      "step": 12559
+    },
+    {
+      "epoch": 1.4591925646238746,
+      "grad_norm": 0.61932772397995,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 12560
+    },
+    {
+      "epoch": 1.459308742375835,
+      "grad_norm": 0.590408444404602,
+      "learning_rate": 0.0001,
+      "loss": 1.4954,
+      "step": 12561
+    },
+    {
+      "epoch": 1.4594249201277956,
+      "grad_norm": 0.6052151918411255,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 12562
+    },
+    {
+      "epoch": 1.459541097879756,
+      "grad_norm": 0.5636199712753296,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 12563
+    },
+    {
+      "epoch": 1.4596572756317165,
+      "grad_norm": 0.639377236366272,
+      "learning_rate": 0.0001,
+      "loss": 1.6294,
+      "step": 12564
+    },
+    {
+      "epoch": 1.459773453383677,
+      "grad_norm": 0.5761369466781616,
+      "learning_rate": 0.0001,
+      "loss": 1.5082,
+      "step": 12565
+    },
+    {
+      "epoch": 1.4598896311356375,
+      "grad_norm": 0.59371018409729,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 12566
+    },
+    {
+      "epoch": 1.460005808887598,
+      "grad_norm": 0.6135494112968445,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 12567
+    },
+    {
+      "epoch": 1.4601219866395585,
+      "grad_norm": 0.6341174840927124,
+      "learning_rate": 0.0001,
+      "loss": 1.2997,
+      "step": 12568
+    },
+    {
+      "epoch": 1.460238164391519,
+      "grad_norm": 0.6097371578216553,
+      "learning_rate": 0.0001,
+      "loss": 1.579,
+      "step": 12569
+    },
+    {
+      "epoch": 1.4603543421434795,
+      "grad_norm": 0.5795385837554932,
+      "learning_rate": 0.0001,
+      "loss": 1.3574,
+      "step": 12570
+    },
+    {
+      "epoch": 1.46047051989544,
+      "grad_norm": 0.618484377861023,
+      "learning_rate": 0.0001,
+      "loss": 1.4424,
+      "step": 12571
+    },
+    {
+      "epoch": 1.4605866976474005,
+      "grad_norm": 0.5799589157104492,
+      "learning_rate": 0.0001,
+      "loss": 1.4669,
+      "step": 12572
+    },
+    {
+      "epoch": 1.460702875399361,
+      "grad_norm": 0.6043416261672974,
+      "learning_rate": 0.0001,
+      "loss": 1.39,
+      "step": 12573
+    },
+    {
+      "epoch": 1.4608190531513214,
+      "grad_norm": 0.5822587013244629,
+      "learning_rate": 0.0001,
+      "loss": 1.3416,
+      "step": 12574
+    },
+    {
+      "epoch": 1.460935230903282,
+      "grad_norm": 0.6009185910224915,
+      "learning_rate": 0.0001,
+      "loss": 1.3142,
+      "step": 12575
+    },
+    {
+      "epoch": 1.4610514086552424,
+      "grad_norm": 0.5679329633712769,
+      "learning_rate": 0.0001,
+      "loss": 1.3666,
+      "step": 12576
+    },
+    {
+      "epoch": 1.461167586407203,
+      "grad_norm": 0.6215576529502869,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 12577
+    },
+    {
+      "epoch": 1.4612837641591634,
+      "grad_norm": 0.5975043177604675,
+      "learning_rate": 0.0001,
+      "loss": 1.39,
+      "step": 12578
+    },
+    {
+      "epoch": 1.461399941911124,
+      "grad_norm": 0.5994958281517029,
+      "learning_rate": 0.0001,
+      "loss": 1.3617,
+      "step": 12579
+    },
+    {
+      "epoch": 1.4615161196630846,
+      "grad_norm": 0.6470338106155396,
+      "learning_rate": 0.0001,
+      "loss": 1.6114,
+      "step": 12580
+    },
+    {
+      "epoch": 1.461632297415045,
+      "grad_norm": 0.7016136050224304,
+      "learning_rate": 0.0001,
+      "loss": 1.5484,
+      "step": 12581
+    },
+    {
+      "epoch": 1.4617484751670056,
+      "grad_norm": 0.5727640986442566,
+      "learning_rate": 0.0001,
+      "loss": 1.3699,
+      "step": 12582
+    },
+    {
+      "epoch": 1.461864652918966,
+      "grad_norm": 0.5754709839820862,
+      "learning_rate": 0.0001,
+      "loss": 1.4004,
+      "step": 12583
+    },
+    {
+      "epoch": 1.4619808306709265,
+      "grad_norm": 0.5758077502250671,
+      "learning_rate": 0.0001,
+      "loss": 1.4384,
+      "step": 12584
+    },
+    {
+      "epoch": 1.462097008422887,
+      "grad_norm": 0.5902067422866821,
+      "learning_rate": 0.0001,
+      "loss": 1.3025,
+      "step": 12585
+    },
+    {
+      "epoch": 1.4622131861748475,
+      "grad_norm": 0.5974416136741638,
+      "learning_rate": 0.0001,
+      "loss": 1.6851,
+      "step": 12586
+    },
+    {
+      "epoch": 1.462329363926808,
+      "grad_norm": 0.5915681719779968,
+      "learning_rate": 0.0001,
+      "loss": 1.4601,
+      "step": 12587
+    },
+    {
+      "epoch": 1.4624455416787685,
+      "grad_norm": 0.5832463502883911,
+      "learning_rate": 0.0001,
+      "loss": 1.4135,
+      "step": 12588
+    },
+    {
+      "epoch": 1.462561719430729,
+      "grad_norm": 0.6045186519622803,
+      "learning_rate": 0.0001,
+      "loss": 1.5844,
+      "step": 12589
+    },
+    {
+      "epoch": 1.4626778971826895,
+      "grad_norm": 0.634339451789856,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 12590
+    },
+    {
+      "epoch": 1.46279407493465,
+      "grad_norm": 0.5863074660301208,
+      "learning_rate": 0.0001,
+      "loss": 1.4268,
+      "step": 12591
+    },
+    {
+      "epoch": 1.4629102526866105,
+      "grad_norm": 0.577804684638977,
+      "learning_rate": 0.0001,
+      "loss": 1.4489,
+      "step": 12592
+    },
+    {
+      "epoch": 1.463026430438571,
+      "grad_norm": 0.6123642921447754,
+      "learning_rate": 0.0001,
+      "loss": 1.5616,
+      "step": 12593
+    },
+    {
+      "epoch": 1.4631426081905314,
+      "grad_norm": 0.6405830979347229,
+      "learning_rate": 0.0001,
+      "loss": 1.658,
+      "step": 12594
+    },
+    {
+      "epoch": 1.4632587859424921,
+      "grad_norm": 0.6347610354423523,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 12595
+    },
+    {
+      "epoch": 1.4633749636944526,
+      "grad_norm": 0.5713246464729309,
+      "learning_rate": 0.0001,
+      "loss": 1.4615,
+      "step": 12596
+    },
+    {
+      "epoch": 1.4634911414464131,
+      "grad_norm": 0.5953773260116577,
+      "learning_rate": 0.0001,
+      "loss": 1.4558,
+      "step": 12597
+    },
+    {
+      "epoch": 1.4636073191983736,
+      "grad_norm": 0.5882254838943481,
+      "learning_rate": 0.0001,
+      "loss": 1.5515,
+      "step": 12598
+    },
+    {
+      "epoch": 1.463723496950334,
+      "grad_norm": 0.5993982553482056,
+      "learning_rate": 0.0001,
+      "loss": 1.3766,
+      "step": 12599
+    },
+    {
+      "epoch": 1.4638396747022946,
+      "grad_norm": 0.6374693512916565,
+      "learning_rate": 0.0001,
+      "loss": 1.5259,
+      "step": 12600
+    },
+    {
+      "epoch": 1.463955852454255,
+      "grad_norm": 0.6188412308692932,
+      "learning_rate": 0.0001,
+      "loss": 1.2187,
+      "step": 12601
+    },
+    {
+      "epoch": 1.4640720302062156,
+      "grad_norm": 0.5834885835647583,
+      "learning_rate": 0.0001,
+      "loss": 1.4844,
+      "step": 12602
+    },
+    {
+      "epoch": 1.464188207958176,
+      "grad_norm": 0.6189852356910706,
+      "learning_rate": 0.0001,
+      "loss": 1.5104,
+      "step": 12603
+    },
+    {
+      "epoch": 1.4643043857101365,
+      "grad_norm": 0.6446365714073181,
+      "learning_rate": 0.0001,
+      "loss": 1.6013,
+      "step": 12604
+    },
+    {
+      "epoch": 1.464420563462097,
+      "grad_norm": 0.6015826463699341,
+      "learning_rate": 0.0001,
+      "loss": 1.4581,
+      "step": 12605
+    },
+    {
+      "epoch": 1.4645367412140575,
+      "grad_norm": 0.6048887372016907,
+      "learning_rate": 0.0001,
+      "loss": 1.391,
+      "step": 12606
+    },
+    {
+      "epoch": 1.464652918966018,
+      "grad_norm": 0.5923680663108826,
+      "learning_rate": 0.0001,
+      "loss": 1.3832,
+      "step": 12607
+    },
+    {
+      "epoch": 1.4647690967179785,
+      "grad_norm": 0.6021586656570435,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 12608
+    },
+    {
+      "epoch": 1.464885274469939,
+      "grad_norm": 0.5922961831092834,
+      "learning_rate": 0.0001,
+      "loss": 1.5496,
+      "step": 12609
+    },
+    {
+      "epoch": 1.4650014522218995,
+      "grad_norm": 0.5911400318145752,
+      "learning_rate": 0.0001,
+      "loss": 1.3201,
+      "step": 12610
+    },
+    {
+      "epoch": 1.46511762997386,
+      "grad_norm": 0.6118112206459045,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 12611
+    },
+    {
+      "epoch": 1.4652338077258205,
+      "grad_norm": 0.5985094308853149,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 12612
+    },
+    {
+      "epoch": 1.465349985477781,
+      "grad_norm": 0.6050367951393127,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 12613
+    },
+    {
+      "epoch": 1.4654661632297414,
+      "grad_norm": 0.5625033378601074,
+      "learning_rate": 0.0001,
+      "loss": 1.1263,
+      "step": 12614
+    },
+    {
+      "epoch": 1.465582340981702,
+      "grad_norm": 0.6526144742965698,
+      "learning_rate": 0.0001,
+      "loss": 1.4681,
+      "step": 12615
+    },
+    {
+      "epoch": 1.4656985187336624,
+      "grad_norm": 0.5911290645599365,
+      "learning_rate": 0.0001,
+      "loss": 1.4809,
+      "step": 12616
+    },
+    {
+      "epoch": 1.465814696485623,
+      "grad_norm": 0.610272228717804,
+      "learning_rate": 0.0001,
+      "loss": 1.3182,
+      "step": 12617
+    },
+    {
+      "epoch": 1.4659308742375834,
+      "grad_norm": 0.6235426068305969,
+      "learning_rate": 0.0001,
+      "loss": 1.6892,
+      "step": 12618
+    },
+    {
+      "epoch": 1.4660470519895439,
+      "grad_norm": 0.5913504362106323,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 12619
+    },
+    {
+      "epoch": 1.4661632297415044,
+      "grad_norm": 0.6157997846603394,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 12620
+    },
+    {
+      "epoch": 1.466279407493465,
+      "grad_norm": 0.6300349235534668,
+      "learning_rate": 0.0001,
+      "loss": 1.507,
+      "step": 12621
+    },
+    {
+      "epoch": 1.4663955852454256,
+      "grad_norm": 0.5744130611419678,
+      "learning_rate": 0.0001,
+      "loss": 1.2242,
+      "step": 12622
+    },
+    {
+      "epoch": 1.466511762997386,
+      "grad_norm": 0.6219088435173035,
+      "learning_rate": 0.0001,
+      "loss": 1.5588,
+      "step": 12623
+    },
+    {
+      "epoch": 1.4666279407493465,
+      "grad_norm": 0.5827035307884216,
+      "learning_rate": 0.0001,
+      "loss": 1.4531,
+      "step": 12624
+    },
+    {
+      "epoch": 1.466744118501307,
+      "grad_norm": 0.6625955104827881,
+      "learning_rate": 0.0001,
+      "loss": 1.6221,
+      "step": 12625
+    },
+    {
+      "epoch": 1.4668602962532675,
+      "grad_norm": 0.6402918100357056,
+      "learning_rate": 0.0001,
+      "loss": 1.6257,
+      "step": 12626
+    },
+    {
+      "epoch": 1.466976474005228,
+      "grad_norm": 0.5460782647132874,
+      "learning_rate": 0.0001,
+      "loss": 1.3716,
+      "step": 12627
+    },
+    {
+      "epoch": 1.4670926517571885,
+      "grad_norm": 0.5440313220024109,
+      "learning_rate": 0.0001,
+      "loss": 1.3949,
+      "step": 12628
+    },
+    {
+      "epoch": 1.467208829509149,
+      "grad_norm": 0.6203206181526184,
+      "learning_rate": 0.0001,
+      "loss": 1.6124,
+      "step": 12629
+    },
+    {
+      "epoch": 1.4673250072611095,
+      "grad_norm": 0.59822678565979,
+      "learning_rate": 0.0001,
+      "loss": 1.4645,
+      "step": 12630
+    },
+    {
+      "epoch": 1.46744118501307,
+      "grad_norm": 0.596192479133606,
+      "learning_rate": 0.0001,
+      "loss": 1.4839,
+      "step": 12631
+    },
+    {
+      "epoch": 1.4675573627650305,
+      "grad_norm": 0.567420482635498,
+      "learning_rate": 0.0001,
+      "loss": 1.495,
+      "step": 12632
+    },
+    {
+      "epoch": 1.467673540516991,
+      "grad_norm": 0.6413534879684448,
+      "learning_rate": 0.0001,
+      "loss": 1.3717,
+      "step": 12633
+    },
+    {
+      "epoch": 1.4677897182689514,
+      "grad_norm": 0.5863707065582275,
+      "learning_rate": 0.0001,
+      "loss": 1.411,
+      "step": 12634
+    },
+    {
+      "epoch": 1.467905896020912,
+      "grad_norm": 0.5698891282081604,
+      "learning_rate": 0.0001,
+      "loss": 1.3382,
+      "step": 12635
+    },
+    {
+      "epoch": 1.4680220737728724,
+      "grad_norm": 0.6080556511878967,
+      "learning_rate": 0.0001,
+      "loss": 1.4324,
+      "step": 12636
+    },
+    {
+      "epoch": 1.4681382515248331,
+      "grad_norm": 0.6110317707061768,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 12637
+    },
+    {
+      "epoch": 1.4682544292767936,
+      "grad_norm": 0.6099866032600403,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 12638
+    },
+    {
+      "epoch": 1.468370607028754,
+      "grad_norm": 0.5823593139648438,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 12639
+    },
+    {
+      "epoch": 1.4684867847807146,
+      "grad_norm": 0.6026428937911987,
+      "learning_rate": 0.0001,
+      "loss": 1.6118,
+      "step": 12640
+    },
+    {
+      "epoch": 1.468602962532675,
+      "grad_norm": 0.5801883935928345,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 12641
+    },
+    {
+      "epoch": 1.4687191402846356,
+      "grad_norm": 0.5360670685768127,
+      "learning_rate": 0.0001,
+      "loss": 1.3604,
+      "step": 12642
+    },
+    {
+      "epoch": 1.468835318036596,
+      "grad_norm": 0.5586822032928467,
+      "learning_rate": 0.0001,
+      "loss": 1.3045,
+      "step": 12643
+    },
+    {
+      "epoch": 1.4689514957885565,
+      "grad_norm": 0.5789737701416016,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 12644
+    },
+    {
+      "epoch": 1.469067673540517,
+      "grad_norm": 0.5735986828804016,
+      "learning_rate": 0.0001,
+      "loss": 1.3718,
+      "step": 12645
+    },
+    {
+      "epoch": 1.4691838512924775,
+      "grad_norm": 0.7054757475852966,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 12646
+    },
+    {
+      "epoch": 1.469300029044438,
+      "grad_norm": 0.6231496930122375,
+      "learning_rate": 0.0001,
+      "loss": 1.4146,
+      "step": 12647
+    },
+    {
+      "epoch": 1.4694162067963985,
+      "grad_norm": 0.5634230375289917,
+      "learning_rate": 0.0001,
+      "loss": 1.3909,
+      "step": 12648
+    },
+    {
+      "epoch": 1.469532384548359,
+      "grad_norm": 0.6670734286308289,
+      "learning_rate": 0.0001,
+      "loss": 1.5684,
+      "step": 12649
+    },
+    {
+      "epoch": 1.4696485623003195,
+      "grad_norm": 0.592263400554657,
+      "learning_rate": 0.0001,
+      "loss": 1.399,
+      "step": 12650
+    },
+    {
+      "epoch": 1.46976474005228,
+      "grad_norm": 0.6188017725944519,
+      "learning_rate": 0.0001,
+      "loss": 1.5407,
+      "step": 12651
+    },
+    {
+      "epoch": 1.4698809178042405,
+      "grad_norm": 0.5997496247291565,
+      "learning_rate": 0.0001,
+      "loss": 1.4838,
+      "step": 12652
+    },
+    {
+      "epoch": 1.469997095556201,
+      "grad_norm": 0.6102793216705322,
+      "learning_rate": 0.0001,
+      "loss": 1.4263,
+      "step": 12653
+    },
+    {
+      "epoch": 1.4701132733081614,
+      "grad_norm": 0.605833888053894,
+      "learning_rate": 0.0001,
+      "loss": 1.5273,
+      "step": 12654
+    },
+    {
+      "epoch": 1.470229451060122,
+      "grad_norm": 0.606594443321228,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 12655
+    },
+    {
+      "epoch": 1.4703456288120824,
+      "grad_norm": 0.5853731036186218,
+      "learning_rate": 0.0001,
+      "loss": 1.3908,
+      "step": 12656
+    },
+    {
+      "epoch": 1.470461806564043,
+      "grad_norm": 0.59587162733078,
+      "learning_rate": 0.0001,
+      "loss": 1.343,
+      "step": 12657
+    },
+    {
+      "epoch": 1.4705779843160034,
+      "grad_norm": 0.6557053327560425,
+      "learning_rate": 0.0001,
+      "loss": 1.7309,
+      "step": 12658
+    },
+    {
+      "epoch": 1.4706941620679639,
+      "grad_norm": 0.6285101771354675,
+      "learning_rate": 0.0001,
+      "loss": 1.5067,
+      "step": 12659
+    },
+    {
+      "epoch": 1.4708103398199244,
+      "grad_norm": 0.5789119005203247,
+      "learning_rate": 0.0001,
+      "loss": 1.2987,
+      "step": 12660
+    },
+    {
+      "epoch": 1.4709265175718849,
+      "grad_norm": 0.5933101177215576,
+      "learning_rate": 0.0001,
+      "loss": 1.3469,
+      "step": 12661
+    },
+    {
+      "epoch": 1.4710426953238454,
+      "grad_norm": 0.5343286395072937,
+      "learning_rate": 0.0001,
+      "loss": 1.3598,
+      "step": 12662
+    },
+    {
+      "epoch": 1.471158873075806,
+      "grad_norm": 0.6221956610679626,
+      "learning_rate": 0.0001,
+      "loss": 1.4563,
+      "step": 12663
+    },
+    {
+      "epoch": 1.4712750508277666,
+      "grad_norm": 0.6174246668815613,
+      "learning_rate": 0.0001,
+      "loss": 1.4157,
+      "step": 12664
+    },
+    {
+      "epoch": 1.471391228579727,
+      "grad_norm": 0.6641348600387573,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 12665
+    },
+    {
+      "epoch": 1.4715074063316875,
+      "grad_norm": 0.5612043738365173,
+      "learning_rate": 0.0001,
+      "loss": 1.2723,
+      "step": 12666
+    },
+    {
+      "epoch": 1.471623584083648,
+      "grad_norm": 0.6334659457206726,
+      "learning_rate": 0.0001,
+      "loss": 1.5654,
+      "step": 12667
+    },
+    {
+      "epoch": 1.4717397618356085,
+      "grad_norm": 0.5860728025436401,
+      "learning_rate": 0.0001,
+      "loss": 1.505,
+      "step": 12668
+    },
+    {
+      "epoch": 1.471855939587569,
+      "grad_norm": 0.6046940684318542,
+      "learning_rate": 0.0001,
+      "loss": 1.5854,
+      "step": 12669
+    },
+    {
+      "epoch": 1.4719721173395295,
+      "grad_norm": 0.5735712051391602,
+      "learning_rate": 0.0001,
+      "loss": 1.3659,
+      "step": 12670
+    },
+    {
+      "epoch": 1.47208829509149,
+      "grad_norm": 0.6405760049819946,
+      "learning_rate": 0.0001,
+      "loss": 1.4848,
+      "step": 12671
+    },
+    {
+      "epoch": 1.4722044728434505,
+      "grad_norm": 0.6492209434509277,
+      "learning_rate": 0.0001,
+      "loss": 1.5792,
+      "step": 12672
+    },
+    {
+      "epoch": 1.472320650595411,
+      "grad_norm": 0.5880120396614075,
+      "learning_rate": 0.0001,
+      "loss": 1.2946,
+      "step": 12673
+    },
+    {
+      "epoch": 1.4724368283473714,
+      "grad_norm": 0.5945394039154053,
+      "learning_rate": 0.0001,
+      "loss": 1.3463,
+      "step": 12674
+    },
+    {
+      "epoch": 1.472553006099332,
+      "grad_norm": 0.6104127168655396,
+      "learning_rate": 0.0001,
+      "loss": 1.4624,
+      "step": 12675
+    },
+    {
+      "epoch": 1.4726691838512924,
+      "grad_norm": 0.6284909248352051,
+      "learning_rate": 0.0001,
+      "loss": 1.6087,
+      "step": 12676
+    },
+    {
+      "epoch": 1.472785361603253,
+      "grad_norm": 0.6103723645210266,
+      "learning_rate": 0.0001,
+      "loss": 1.5014,
+      "step": 12677
+    },
+    {
+      "epoch": 1.4729015393552134,
+      "grad_norm": 0.6179032921791077,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 12678
+    },
+    {
+      "epoch": 1.473017717107174,
+      "grad_norm": 0.5750223398208618,
+      "learning_rate": 0.0001,
+      "loss": 1.4715,
+      "step": 12679
+    },
+    {
+      "epoch": 1.4731338948591346,
+      "grad_norm": 0.5653707981109619,
+      "learning_rate": 0.0001,
+      "loss": 1.2779,
+      "step": 12680
+    },
+    {
+      "epoch": 1.473250072611095,
+      "grad_norm": 0.6165471076965332,
+      "learning_rate": 0.0001,
+      "loss": 1.3156,
+      "step": 12681
+    },
+    {
+      "epoch": 1.4733662503630556,
+      "grad_norm": 0.5788655281066895,
+      "learning_rate": 0.0001,
+      "loss": 1.3653,
+      "step": 12682
+    },
+    {
+      "epoch": 1.473482428115016,
+      "grad_norm": 0.5805605053901672,
+      "learning_rate": 0.0001,
+      "loss": 1.4103,
+      "step": 12683
+    },
+    {
+      "epoch": 1.4735986058669766,
+      "grad_norm": 0.6233782172203064,
+      "learning_rate": 0.0001,
+      "loss": 1.427,
+      "step": 12684
+    },
+    {
+      "epoch": 1.473714783618937,
+      "grad_norm": 0.6231675744056702,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 12685
+    },
+    {
+      "epoch": 1.4738309613708975,
+      "grad_norm": 0.5923117399215698,
+      "learning_rate": 0.0001,
+      "loss": 1.3208,
+      "step": 12686
+    },
+    {
+      "epoch": 1.473947139122858,
+      "grad_norm": 0.6191889047622681,
+      "learning_rate": 0.0001,
+      "loss": 1.4431,
+      "step": 12687
+    },
+    {
+      "epoch": 1.4740633168748185,
+      "grad_norm": 0.5808857083320618,
+      "learning_rate": 0.0001,
+      "loss": 1.4314,
+      "step": 12688
+    },
+    {
+      "epoch": 1.474179494626779,
+      "grad_norm": 0.6818951964378357,
+      "learning_rate": 0.0001,
+      "loss": 1.5522,
+      "step": 12689
+    },
+    {
+      "epoch": 1.4742956723787395,
+      "grad_norm": 0.6566728353500366,
+      "learning_rate": 0.0001,
+      "loss": 1.5566,
+      "step": 12690
+    },
+    {
+      "epoch": 1.4744118501307,
+      "grad_norm": 0.6200664639472961,
+      "learning_rate": 0.0001,
+      "loss": 1.5531,
+      "step": 12691
+    },
+    {
+      "epoch": 1.4745280278826605,
+      "grad_norm": 0.5786384344100952,
+      "learning_rate": 0.0001,
+      "loss": 1.4235,
+      "step": 12692
+    },
+    {
+      "epoch": 1.474644205634621,
+      "grad_norm": 0.5918937921524048,
+      "learning_rate": 0.0001,
+      "loss": 1.3155,
+      "step": 12693
+    },
+    {
+      "epoch": 1.4747603833865814,
+      "grad_norm": 0.6062121987342834,
+      "learning_rate": 0.0001,
+      "loss": 1.4284,
+      "step": 12694
+    },
+    {
+      "epoch": 1.474876561138542,
+      "grad_norm": 0.614924967288971,
+      "learning_rate": 0.0001,
+      "loss": 1.6247,
+      "step": 12695
+    },
+    {
+      "epoch": 1.4749927388905024,
+      "grad_norm": 0.6098072528839111,
+      "learning_rate": 0.0001,
+      "loss": 1.3263,
+      "step": 12696
+    },
+    {
+      "epoch": 1.475108916642463,
+      "grad_norm": 0.6092824935913086,
+      "learning_rate": 0.0001,
+      "loss": 1.4069,
+      "step": 12697
+    },
+    {
+      "epoch": 1.4752250943944234,
+      "grad_norm": 0.6376292109489441,
+      "learning_rate": 0.0001,
+      "loss": 1.4499,
+      "step": 12698
+    },
+    {
+      "epoch": 1.4753412721463839,
+      "grad_norm": 0.6538935899734497,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 12699
+    },
+    {
+      "epoch": 1.4754574498983444,
+      "grad_norm": 0.6000575423240662,
+      "learning_rate": 0.0001,
+      "loss": 1.3261,
+      "step": 12700
+    },
+    {
+      "epoch": 1.4755736276503049,
+      "grad_norm": 0.6823968887329102,
+      "learning_rate": 0.0001,
+      "loss": 1.2802,
+      "step": 12701
+    },
+    {
+      "epoch": 1.4756898054022654,
+      "grad_norm": 0.6618595123291016,
+      "learning_rate": 0.0001,
+      "loss": 1.5822,
+      "step": 12702
+    },
+    {
+      "epoch": 1.4758059831542258,
+      "grad_norm": 0.6516097187995911,
+      "learning_rate": 0.0001,
+      "loss": 1.5829,
+      "step": 12703
+    },
+    {
+      "epoch": 1.4759221609061863,
+      "grad_norm": 0.6231207251548767,
+      "learning_rate": 0.0001,
+      "loss": 1.687,
+      "step": 12704
+    },
+    {
+      "epoch": 1.476038338658147,
+      "grad_norm": 0.6442578434944153,
+      "learning_rate": 0.0001,
+      "loss": 1.4111,
+      "step": 12705
+    },
+    {
+      "epoch": 1.4761545164101075,
+      "grad_norm": 0.6241633296012878,
+      "learning_rate": 0.0001,
+      "loss": 1.6373,
+      "step": 12706
+    },
+    {
+      "epoch": 1.476270694162068,
+      "grad_norm": 0.5882102847099304,
+      "learning_rate": 0.0001,
+      "loss": 1.3378,
+      "step": 12707
+    },
+    {
+      "epoch": 1.4763868719140285,
+      "grad_norm": 0.6080202460289001,
+      "learning_rate": 0.0001,
+      "loss": 1.5929,
+      "step": 12708
+    },
+    {
+      "epoch": 1.476503049665989,
+      "grad_norm": 0.6191835999488831,
+      "learning_rate": 0.0001,
+      "loss": 1.4236,
+      "step": 12709
+    },
+    {
+      "epoch": 1.4766192274179495,
+      "grad_norm": 0.5939556360244751,
+      "learning_rate": 0.0001,
+      "loss": 1.3395,
+      "step": 12710
+    },
+    {
+      "epoch": 1.47673540516991,
+      "grad_norm": 0.6249536275863647,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 12711
+    },
+    {
+      "epoch": 1.4768515829218705,
+      "grad_norm": 0.5902261137962341,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 12712
+    },
+    {
+      "epoch": 1.476967760673831,
+      "grad_norm": 0.6220241785049438,
+      "learning_rate": 0.0001,
+      "loss": 1.5443,
+      "step": 12713
+    },
+    {
+      "epoch": 1.4770839384257914,
+      "grad_norm": 0.5737898349761963,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 12714
+    },
+    {
+      "epoch": 1.477200116177752,
+      "grad_norm": 0.6006242036819458,
+      "learning_rate": 0.0001,
+      "loss": 1.5542,
+      "step": 12715
+    },
+    {
+      "epoch": 1.4773162939297124,
+      "grad_norm": 0.6238159537315369,
+      "learning_rate": 0.0001,
+      "loss": 1.3714,
+      "step": 12716
+    },
+    {
+      "epoch": 1.477432471681673,
+      "grad_norm": 0.5738762617111206,
+      "learning_rate": 0.0001,
+      "loss": 1.3738,
+      "step": 12717
+    },
+    {
+      "epoch": 1.4775486494336334,
+      "grad_norm": 0.573409914970398,
+      "learning_rate": 0.0001,
+      "loss": 1.2755,
+      "step": 12718
+    },
+    {
+      "epoch": 1.4776648271855939,
+      "grad_norm": 0.5664563775062561,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 12719
+    },
+    {
+      "epoch": 1.4777810049375544,
+      "grad_norm": 0.6218703389167786,
+      "learning_rate": 0.0001,
+      "loss": 1.6621,
+      "step": 12720
+    },
+    {
+      "epoch": 1.477897182689515,
+      "grad_norm": 0.5889036655426025,
+      "learning_rate": 0.0001,
+      "loss": 1.6083,
+      "step": 12721
+    },
+    {
+      "epoch": 1.4780133604414756,
+      "grad_norm": 0.6417556405067444,
+      "learning_rate": 0.0001,
+      "loss": 1.6125,
+      "step": 12722
+    },
+    {
+      "epoch": 1.478129538193436,
+      "grad_norm": 0.5941647291183472,
+      "learning_rate": 0.0001,
+      "loss": 1.5546,
+      "step": 12723
+    },
+    {
+      "epoch": 1.4782457159453966,
+      "grad_norm": 0.5559972524642944,
+      "learning_rate": 0.0001,
+      "loss": 1.4296,
+      "step": 12724
+    },
+    {
+      "epoch": 1.478361893697357,
+      "grad_norm": 0.6716883182525635,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 12725
+    },
+    {
+      "epoch": 1.4784780714493175,
+      "grad_norm": 0.582317054271698,
+      "learning_rate": 0.0001,
+      "loss": 1.556,
+      "step": 12726
+    },
+    {
+      "epoch": 1.478594249201278,
+      "grad_norm": 0.5578384399414062,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 12727
+    },
+    {
+      "epoch": 1.4787104269532385,
+      "grad_norm": 0.6159401535987854,
+      "learning_rate": 0.0001,
+      "loss": 1.4841,
+      "step": 12728
+    },
+    {
+      "epoch": 1.478826604705199,
+      "grad_norm": 0.5790227055549622,
+      "learning_rate": 0.0001,
+      "loss": 1.2968,
+      "step": 12729
+    },
+    {
+      "epoch": 1.4789427824571595,
+      "grad_norm": 0.5429813265800476,
+      "learning_rate": 0.0001,
+      "loss": 1.2731,
+      "step": 12730
+    },
+    {
+      "epoch": 1.47905896020912,
+      "grad_norm": 0.6098136901855469,
+      "learning_rate": 0.0001,
+      "loss": 1.5287,
+      "step": 12731
+    },
+    {
+      "epoch": 1.4791751379610805,
+      "grad_norm": 0.6206817030906677,
+      "learning_rate": 0.0001,
+      "loss": 1.3591,
+      "step": 12732
+    },
+    {
+      "epoch": 1.479291315713041,
+      "grad_norm": 0.6183215975761414,
+      "learning_rate": 0.0001,
+      "loss": 1.5038,
+      "step": 12733
+    },
+    {
+      "epoch": 1.4794074934650014,
+      "grad_norm": 0.6770209670066833,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 12734
+    },
+    {
+      "epoch": 1.479523671216962,
+      "grad_norm": 0.6122325658798218,
+      "learning_rate": 0.0001,
+      "loss": 1.3449,
+      "step": 12735
+    },
+    {
+      "epoch": 1.4796398489689224,
+      "grad_norm": 0.6098016500473022,
+      "learning_rate": 0.0001,
+      "loss": 1.4898,
+      "step": 12736
+    },
+    {
+      "epoch": 1.479756026720883,
+      "grad_norm": 0.6338397860527039,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 12737
+    },
+    {
+      "epoch": 1.4798722044728434,
+      "grad_norm": 0.6147565841674805,
+      "learning_rate": 0.0001,
+      "loss": 1.5078,
+      "step": 12738
+    },
+    {
+      "epoch": 1.4799883822248039,
+      "grad_norm": 0.622284471988678,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 12739
+    },
+    {
+      "epoch": 1.4801045599767644,
+      "grad_norm": 0.5711101293563843,
+      "learning_rate": 0.0001,
+      "loss": 1.4056,
+      "step": 12740
+    },
+    {
+      "epoch": 1.4802207377287249,
+      "grad_norm": 0.5917078852653503,
+      "learning_rate": 0.0001,
+      "loss": 1.4603,
+      "step": 12741
+    },
+    {
+      "epoch": 1.4803369154806854,
+      "grad_norm": 0.5660672783851624,
+      "learning_rate": 0.0001,
+      "loss": 1.3994,
+      "step": 12742
+    },
+    {
+      "epoch": 1.4804530932326458,
+      "grad_norm": 0.586409330368042,
+      "learning_rate": 0.0001,
+      "loss": 1.3446,
+      "step": 12743
+    },
+    {
+      "epoch": 1.4805692709846063,
+      "grad_norm": 0.6024650931358337,
+      "learning_rate": 0.0001,
+      "loss": 1.4112,
+      "step": 12744
+    },
+    {
+      "epoch": 1.4806854487365668,
+      "grad_norm": 0.5979096293449402,
+      "learning_rate": 0.0001,
+      "loss": 1.4108,
+      "step": 12745
+    },
+    {
+      "epoch": 1.4808016264885273,
+      "grad_norm": 0.6005404591560364,
+      "learning_rate": 0.0001,
+      "loss": 1.4058,
+      "step": 12746
+    },
+    {
+      "epoch": 1.480917804240488,
+      "grad_norm": 0.6152083873748779,
+      "learning_rate": 0.0001,
+      "loss": 1.4269,
+      "step": 12747
+    },
+    {
+      "epoch": 1.4810339819924485,
+      "grad_norm": 0.6299720406532288,
+      "learning_rate": 0.0001,
+      "loss": 1.306,
+      "step": 12748
+    },
+    {
+      "epoch": 1.481150159744409,
+      "grad_norm": 0.6793599724769592,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 12749
+    },
+    {
+      "epoch": 1.4812663374963695,
+      "grad_norm": 0.5687341690063477,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 12750
+    },
+    {
+      "epoch": 1.48138251524833,
+      "grad_norm": 0.5625494718551636,
+      "learning_rate": 0.0001,
+      "loss": 1.3357,
+      "step": 12751
+    },
+    {
+      "epoch": 1.4814986930002905,
+      "grad_norm": 0.6246379017829895,
+      "learning_rate": 0.0001,
+      "loss": 1.4443,
+      "step": 12752
+    },
+    {
+      "epoch": 1.481614870752251,
+      "grad_norm": 0.6253631114959717,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 12753
+    },
+    {
+      "epoch": 1.4817310485042114,
+      "grad_norm": 0.603517472743988,
+      "learning_rate": 0.0001,
+      "loss": 1.3907,
+      "step": 12754
+    },
+    {
+      "epoch": 1.481847226256172,
+      "grad_norm": 0.5724433660507202,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 12755
+    },
+    {
+      "epoch": 1.4819634040081324,
+      "grad_norm": 0.6194106340408325,
+      "learning_rate": 0.0001,
+      "loss": 1.4979,
+      "step": 12756
+    },
+    {
+      "epoch": 1.482079581760093,
+      "grad_norm": 0.6359567642211914,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 12757
+    },
+    {
+      "epoch": 1.4821957595120534,
+      "grad_norm": 0.5960992574691772,
+      "learning_rate": 0.0001,
+      "loss": 1.4689,
+      "step": 12758
+    },
+    {
+      "epoch": 1.482311937264014,
+      "grad_norm": 0.5790872573852539,
+      "learning_rate": 0.0001,
+      "loss": 1.3061,
+      "step": 12759
+    },
+    {
+      "epoch": 1.4824281150159744,
+      "grad_norm": 0.5998115539550781,
+      "learning_rate": 0.0001,
+      "loss": 1.5584,
+      "step": 12760
+    },
+    {
+      "epoch": 1.4825442927679349,
+      "grad_norm": 0.5446498394012451,
+      "learning_rate": 0.0001,
+      "loss": 1.3484,
+      "step": 12761
+    },
+    {
+      "epoch": 1.4826604705198954,
+      "grad_norm": 0.6086790561676025,
+      "learning_rate": 0.0001,
+      "loss": 1.4267,
+      "step": 12762
+    },
+    {
+      "epoch": 1.482776648271856,
+      "grad_norm": 0.5855661630630493,
+      "learning_rate": 0.0001,
+      "loss": 1.1741,
+      "step": 12763
+    },
+    {
+      "epoch": 1.4828928260238166,
+      "grad_norm": 0.6026800274848938,
+      "learning_rate": 0.0001,
+      "loss": 1.5557,
+      "step": 12764
+    },
+    {
+      "epoch": 1.483009003775777,
+      "grad_norm": 0.5826662182807922,
+      "learning_rate": 0.0001,
+      "loss": 1.5116,
+      "step": 12765
+    },
+    {
+      "epoch": 1.4831251815277375,
+      "grad_norm": 0.6261777281761169,
+      "learning_rate": 0.0001,
+      "loss": 1.3935,
+      "step": 12766
+    },
+    {
+      "epoch": 1.483241359279698,
+      "grad_norm": 0.6069555282592773,
+      "learning_rate": 0.0001,
+      "loss": 1.3201,
+      "step": 12767
+    },
+    {
+      "epoch": 1.4833575370316585,
+      "grad_norm": 0.6280861496925354,
+      "learning_rate": 0.0001,
+      "loss": 1.5374,
+      "step": 12768
+    },
+    {
+      "epoch": 1.483473714783619,
+      "grad_norm": 0.6123801469802856,
+      "learning_rate": 0.0001,
+      "loss": 1.2891,
+      "step": 12769
+    },
+    {
+      "epoch": 1.4835898925355795,
+      "grad_norm": 0.5813351273536682,
+      "learning_rate": 0.0001,
+      "loss": 1.3189,
+      "step": 12770
+    },
+    {
+      "epoch": 1.48370607028754,
+      "grad_norm": 0.6434810161590576,
+      "learning_rate": 0.0001,
+      "loss": 1.5881,
+      "step": 12771
+    },
+    {
+      "epoch": 1.4838222480395005,
+      "grad_norm": 0.6274320483207703,
+      "learning_rate": 0.0001,
+      "loss": 1.5605,
+      "step": 12772
+    },
+    {
+      "epoch": 1.483938425791461,
+      "grad_norm": 0.5794488787651062,
+      "learning_rate": 0.0001,
+      "loss": 1.5702,
+      "step": 12773
+    },
+    {
+      "epoch": 1.4840546035434214,
+      "grad_norm": 0.5843713879585266,
+      "learning_rate": 0.0001,
+      "loss": 1.4887,
+      "step": 12774
+    },
+    {
+      "epoch": 1.484170781295382,
+      "grad_norm": 0.5706945657730103,
+      "learning_rate": 0.0001,
+      "loss": 1.4438,
+      "step": 12775
+    },
+    {
+      "epoch": 1.4842869590473424,
+      "grad_norm": 0.6139296293258667,
+      "learning_rate": 0.0001,
+      "loss": 1.5284,
+      "step": 12776
+    },
+    {
+      "epoch": 1.484403136799303,
+      "grad_norm": 0.658415675163269,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 12777
+    },
+    {
+      "epoch": 1.4845193145512634,
+      "grad_norm": 0.6205752491950989,
+      "learning_rate": 0.0001,
+      "loss": 1.4368,
+      "step": 12778
+    },
+    {
+      "epoch": 1.484635492303224,
+      "grad_norm": 0.5956551432609558,
+      "learning_rate": 0.0001,
+      "loss": 1.4247,
+      "step": 12779
+    },
+    {
+      "epoch": 1.4847516700551844,
+      "grad_norm": 0.5731070637702942,
+      "learning_rate": 0.0001,
+      "loss": 1.4603,
+      "step": 12780
+    },
+    {
+      "epoch": 1.4848678478071449,
+      "grad_norm": 0.5833913683891296,
+      "learning_rate": 0.0001,
+      "loss": 1.4769,
+      "step": 12781
+    },
+    {
+      "epoch": 1.4849840255591054,
+      "grad_norm": 0.5474773645401001,
+      "learning_rate": 0.0001,
+      "loss": 1.509,
+      "step": 12782
+    },
+    {
+      "epoch": 1.4851002033110658,
+      "grad_norm": 0.5622663497924805,
+      "learning_rate": 0.0001,
+      "loss": 1.4723,
+      "step": 12783
+    },
+    {
+      "epoch": 1.4852163810630263,
+      "grad_norm": 0.5745351314544678,
+      "learning_rate": 0.0001,
+      "loss": 1.4861,
+      "step": 12784
+    },
+    {
+      "epoch": 1.4853325588149868,
+      "grad_norm": 0.5686841607093811,
+      "learning_rate": 0.0001,
+      "loss": 1.4275,
+      "step": 12785
+    },
+    {
+      "epoch": 1.4854487365669473,
+      "grad_norm": 0.5676813721656799,
+      "learning_rate": 0.0001,
+      "loss": 1.3021,
+      "step": 12786
+    },
+    {
+      "epoch": 1.4855649143189078,
+      "grad_norm": 0.5953998565673828,
+      "learning_rate": 0.0001,
+      "loss": 1.3295,
+      "step": 12787
+    },
+    {
+      "epoch": 1.4856810920708683,
+      "grad_norm": 0.5832408666610718,
+      "learning_rate": 0.0001,
+      "loss": 1.4133,
+      "step": 12788
+    },
+    {
+      "epoch": 1.485797269822829,
+      "grad_norm": 0.6295830011367798,
+      "learning_rate": 0.0001,
+      "loss": 1.4368,
+      "step": 12789
+    },
+    {
+      "epoch": 1.4859134475747895,
+      "grad_norm": 0.595488965511322,
+      "learning_rate": 0.0001,
+      "loss": 1.5149,
+      "step": 12790
+    },
+    {
+      "epoch": 1.48602962532675,
+      "grad_norm": 0.5632891654968262,
+      "learning_rate": 0.0001,
+      "loss": 1.3456,
+      "step": 12791
+    },
+    {
+      "epoch": 1.4861458030787105,
+      "grad_norm": 0.5827213525772095,
+      "learning_rate": 0.0001,
+      "loss": 1.4786,
+      "step": 12792
+    },
+    {
+      "epoch": 1.486261980830671,
+      "grad_norm": 0.58870929479599,
+      "learning_rate": 0.0001,
+      "loss": 1.3798,
+      "step": 12793
+    },
+    {
+      "epoch": 1.4863781585826314,
+      "grad_norm": 0.566349983215332,
+      "learning_rate": 0.0001,
+      "loss": 1.2853,
+      "step": 12794
+    },
+    {
+      "epoch": 1.486494336334592,
+      "grad_norm": 0.5674836039543152,
+      "learning_rate": 0.0001,
+      "loss": 1.4405,
+      "step": 12795
+    },
+    {
+      "epoch": 1.4866105140865524,
+      "grad_norm": 0.6061223745346069,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 12796
+    },
+    {
+      "epoch": 1.486726691838513,
+      "grad_norm": 0.6239713430404663,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 12797
+    },
+    {
+      "epoch": 1.4868428695904734,
+      "grad_norm": 0.5661698579788208,
+      "learning_rate": 0.0001,
+      "loss": 1.3916,
+      "step": 12798
+    },
+    {
+      "epoch": 1.486959047342434,
+      "grad_norm": 0.6092931628227234,
+      "learning_rate": 0.0001,
+      "loss": 1.506,
+      "step": 12799
+    },
+    {
+      "epoch": 1.4870752250943944,
+      "grad_norm": 0.6277779936790466,
+      "learning_rate": 0.0001,
+      "loss": 1.6359,
+      "step": 12800
+    },
+    {
+      "epoch": 1.4871914028463549,
+      "grad_norm": 0.5889933705329895,
+      "learning_rate": 0.0001,
+      "loss": 1.3339,
+      "step": 12801
+    },
+    {
+      "epoch": 1.4873075805983154,
+      "grad_norm": 0.5621128082275391,
+      "learning_rate": 0.0001,
+      "loss": 1.2968,
+      "step": 12802
+    },
+    {
+      "epoch": 1.4874237583502758,
+      "grad_norm": 0.5749889016151428,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 12803
+    },
+    {
+      "epoch": 1.4875399361022366,
+      "grad_norm": 0.6799257397651672,
+      "learning_rate": 0.0001,
+      "loss": 1.616,
+      "step": 12804
+    },
+    {
+      "epoch": 1.487656113854197,
+      "grad_norm": 0.575088381767273,
+      "learning_rate": 0.0001,
+      "loss": 1.5002,
+      "step": 12805
+    },
+    {
+      "epoch": 1.4877722916061575,
+      "grad_norm": 0.6447527408599854,
+      "learning_rate": 0.0001,
+      "loss": 1.5963,
+      "step": 12806
+    },
+    {
+      "epoch": 1.487888469358118,
+      "grad_norm": 0.6129550337791443,
+      "learning_rate": 0.0001,
+      "loss": 1.4895,
+      "step": 12807
+    },
+    {
+      "epoch": 1.4880046471100785,
+      "grad_norm": 0.6152228713035583,
+      "learning_rate": 0.0001,
+      "loss": 1.388,
+      "step": 12808
+    },
+    {
+      "epoch": 1.488120824862039,
+      "grad_norm": 0.6193944215774536,
+      "learning_rate": 0.0001,
+      "loss": 1.5167,
+      "step": 12809
+    },
+    {
+      "epoch": 1.4882370026139995,
+      "grad_norm": 0.6093383431434631,
+      "learning_rate": 0.0001,
+      "loss": 1.6097,
+      "step": 12810
+    },
+    {
+      "epoch": 1.48835318036596,
+      "grad_norm": 0.589266300201416,
+      "learning_rate": 0.0001,
+      "loss": 1.3637,
+      "step": 12811
+    },
+    {
+      "epoch": 1.4884693581179205,
+      "grad_norm": 0.6045963764190674,
+      "learning_rate": 0.0001,
+      "loss": 1.4242,
+      "step": 12812
+    },
+    {
+      "epoch": 1.488585535869881,
+      "grad_norm": 0.5933640003204346,
+      "learning_rate": 0.0001,
+      "loss": 1.3325,
+      "step": 12813
+    },
+    {
+      "epoch": 1.4887017136218414,
+      "grad_norm": 0.5831065773963928,
+      "learning_rate": 0.0001,
+      "loss": 1.4035,
+      "step": 12814
+    },
+    {
+      "epoch": 1.488817891373802,
+      "grad_norm": 0.6712548136711121,
+      "learning_rate": 0.0001,
+      "loss": 1.4683,
+      "step": 12815
+    },
+    {
+      "epoch": 1.4889340691257624,
+      "grad_norm": 0.5910773277282715,
+      "learning_rate": 0.0001,
+      "loss": 1.4458,
+      "step": 12816
+    },
+    {
+      "epoch": 1.489050246877723,
+      "grad_norm": 0.6439294219017029,
+      "learning_rate": 0.0001,
+      "loss": 1.6244,
+      "step": 12817
+    },
+    {
+      "epoch": 1.4891664246296834,
+      "grad_norm": 0.5771844387054443,
+      "learning_rate": 0.0001,
+      "loss": 1.4584,
+      "step": 12818
+    },
+    {
+      "epoch": 1.489282602381644,
+      "grad_norm": 0.6531968712806702,
+      "learning_rate": 0.0001,
+      "loss": 1.533,
+      "step": 12819
+    },
+    {
+      "epoch": 1.4893987801336044,
+      "grad_norm": 0.6017487049102783,
+      "learning_rate": 0.0001,
+      "loss": 1.4118,
+      "step": 12820
+    },
+    {
+      "epoch": 1.4895149578855649,
+      "grad_norm": 0.6122699975967407,
+      "learning_rate": 0.0001,
+      "loss": 1.4481,
+      "step": 12821
+    },
+    {
+      "epoch": 1.4896311356375254,
+      "grad_norm": 0.6408692598342896,
+      "learning_rate": 0.0001,
+      "loss": 1.5795,
+      "step": 12822
+    },
+    {
+      "epoch": 1.4897473133894858,
+      "grad_norm": 0.6655322909355164,
+      "learning_rate": 0.0001,
+      "loss": 1.5516,
+      "step": 12823
+    },
+    {
+      "epoch": 1.4898634911414463,
+      "grad_norm": 0.5901119709014893,
+      "learning_rate": 0.0001,
+      "loss": 1.3471,
+      "step": 12824
+    },
+    {
+      "epoch": 1.4899796688934068,
+      "grad_norm": 0.574163019657135,
+      "learning_rate": 0.0001,
+      "loss": 1.3641,
+      "step": 12825
+    },
+    {
+      "epoch": 1.4900958466453673,
+      "grad_norm": 0.641035795211792,
+      "learning_rate": 0.0001,
+      "loss": 1.4546,
+      "step": 12826
+    },
+    {
+      "epoch": 1.4902120243973278,
+      "grad_norm": 0.6070435643196106,
+      "learning_rate": 0.0001,
+      "loss": 1.5669,
+      "step": 12827
+    },
+    {
+      "epoch": 1.4903282021492883,
+      "grad_norm": 0.6125596761703491,
+      "learning_rate": 0.0001,
+      "loss": 1.5613,
+      "step": 12828
+    },
+    {
+      "epoch": 1.4904443799012488,
+      "grad_norm": 0.5792481899261475,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 12829
+    },
+    {
+      "epoch": 1.4905605576532093,
+      "grad_norm": 0.6352988481521606,
+      "learning_rate": 0.0001,
+      "loss": 1.6555,
+      "step": 12830
+    },
+    {
+      "epoch": 1.49067673540517,
+      "grad_norm": 0.6231880187988281,
+      "learning_rate": 0.0001,
+      "loss": 1.6766,
+      "step": 12831
+    },
+    {
+      "epoch": 1.4907929131571305,
+      "grad_norm": 0.5636422634124756,
+      "learning_rate": 0.0001,
+      "loss": 1.1924,
+      "step": 12832
+    },
+    {
+      "epoch": 1.490909090909091,
+      "grad_norm": 0.6103519797325134,
+      "learning_rate": 0.0001,
+      "loss": 1.5181,
+      "step": 12833
+    },
+    {
+      "epoch": 1.4910252686610515,
+      "grad_norm": 0.5947346687316895,
+      "learning_rate": 0.0001,
+      "loss": 1.3394,
+      "step": 12834
+    },
+    {
+      "epoch": 1.491141446413012,
+      "grad_norm": 0.6331939101219177,
+      "learning_rate": 0.0001,
+      "loss": 1.4172,
+      "step": 12835
+    },
+    {
+      "epoch": 1.4912576241649724,
+      "grad_norm": 0.5931614637374878,
+      "learning_rate": 0.0001,
+      "loss": 1.2726,
+      "step": 12836
+    },
+    {
+      "epoch": 1.491373801916933,
+      "grad_norm": 0.6468703150749207,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 12837
+    },
+    {
+      "epoch": 1.4914899796688934,
+      "grad_norm": 0.6426784992218018,
+      "learning_rate": 0.0001,
+      "loss": 1.5128,
+      "step": 12838
+    },
+    {
+      "epoch": 1.491606157420854,
+      "grad_norm": 0.6114495992660522,
+      "learning_rate": 0.0001,
+      "loss": 1.3851,
+      "step": 12839
+    },
+    {
+      "epoch": 1.4917223351728144,
+      "grad_norm": 0.5668593645095825,
+      "learning_rate": 0.0001,
+      "loss": 1.3923,
+      "step": 12840
+    },
+    {
+      "epoch": 1.4918385129247749,
+      "grad_norm": 0.6152085661888123,
+      "learning_rate": 0.0001,
+      "loss": 1.4781,
+      "step": 12841
+    },
+    {
+      "epoch": 1.4919546906767354,
+      "grad_norm": 0.6062020063400269,
+      "learning_rate": 0.0001,
+      "loss": 1.4925,
+      "step": 12842
+    },
+    {
+      "epoch": 1.4920708684286959,
+      "grad_norm": 0.589048445224762,
+      "learning_rate": 0.0001,
+      "loss": 1.4028,
+      "step": 12843
+    },
+    {
+      "epoch": 1.4921870461806563,
+      "grad_norm": 0.6015335917472839,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 12844
+    },
+    {
+      "epoch": 1.4923032239326168,
+      "grad_norm": 0.581270158290863,
+      "learning_rate": 0.0001,
+      "loss": 1.4128,
+      "step": 12845
+    },
+    {
+      "epoch": 1.4924194016845775,
+      "grad_norm": 0.5742862820625305,
+      "learning_rate": 0.0001,
+      "loss": 1.4333,
+      "step": 12846
+    },
+    {
+      "epoch": 1.492535579436538,
+      "grad_norm": 0.6063739061355591,
+      "learning_rate": 0.0001,
+      "loss": 1.4433,
+      "step": 12847
+    },
+    {
+      "epoch": 1.4926517571884985,
+      "grad_norm": 0.5964763760566711,
+      "learning_rate": 0.0001,
+      "loss": 1.3218,
+      "step": 12848
+    },
+    {
+      "epoch": 1.492767934940459,
+      "grad_norm": 0.6101694107055664,
+      "learning_rate": 0.0001,
+      "loss": 1.5799,
+      "step": 12849
+    },
+    {
+      "epoch": 1.4928841126924195,
+      "grad_norm": 0.6307281851768494,
+      "learning_rate": 0.0001,
+      "loss": 1.5571,
+      "step": 12850
+    },
+    {
+      "epoch": 1.49300029044438,
+      "grad_norm": 0.5971539616584778,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 12851
+    },
+    {
+      "epoch": 1.4931164681963405,
+      "grad_norm": 0.6096937656402588,
+      "learning_rate": 0.0001,
+      "loss": 1.4911,
+      "step": 12852
+    },
+    {
+      "epoch": 1.493232645948301,
+      "grad_norm": 0.6482102274894714,
+      "learning_rate": 0.0001,
+      "loss": 1.4308,
+      "step": 12853
+    },
+    {
+      "epoch": 1.4933488237002615,
+      "grad_norm": 0.6256344318389893,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 12854
+    },
+    {
+      "epoch": 1.493465001452222,
+      "grad_norm": 0.5984079837799072,
+      "learning_rate": 0.0001,
+      "loss": 1.4259,
+      "step": 12855
+    },
+    {
+      "epoch": 1.4935811792041824,
+      "grad_norm": 0.5901398658752441,
+      "learning_rate": 0.0001,
+      "loss": 1.4518,
+      "step": 12856
+    },
+    {
+      "epoch": 1.493697356956143,
+      "grad_norm": 0.6232024431228638,
+      "learning_rate": 0.0001,
+      "loss": 1.5517,
+      "step": 12857
+    },
+    {
+      "epoch": 1.4938135347081034,
+      "grad_norm": 0.5728425979614258,
+      "learning_rate": 0.0001,
+      "loss": 1.2854,
+      "step": 12858
+    },
+    {
+      "epoch": 1.493929712460064,
+      "grad_norm": 0.617725670337677,
+      "learning_rate": 0.0001,
+      "loss": 1.447,
+      "step": 12859
+    },
+    {
+      "epoch": 1.4940458902120244,
+      "grad_norm": 0.6785517334938049,
+      "learning_rate": 0.0001,
+      "loss": 1.4123,
+      "step": 12860
+    },
+    {
+      "epoch": 1.4941620679639849,
+      "grad_norm": 0.6345421075820923,
+      "learning_rate": 0.0001,
+      "loss": 1.5892,
+      "step": 12861
+    },
+    {
+      "epoch": 1.4942782457159454,
+      "grad_norm": 0.5922654271125793,
+      "learning_rate": 0.0001,
+      "loss": 1.4769,
+      "step": 12862
+    },
+    {
+      "epoch": 1.4943944234679059,
+      "grad_norm": 0.6028105616569519,
+      "learning_rate": 0.0001,
+      "loss": 1.518,
+      "step": 12863
+    },
+    {
+      "epoch": 1.4945106012198663,
+      "grad_norm": 0.6089810132980347,
+      "learning_rate": 0.0001,
+      "loss": 1.333,
+      "step": 12864
+    },
+    {
+      "epoch": 1.4946267789718268,
+      "grad_norm": 0.6128267645835876,
+      "learning_rate": 0.0001,
+      "loss": 1.5329,
+      "step": 12865
+    },
+    {
+      "epoch": 1.4947429567237873,
+      "grad_norm": 0.7104302644729614,
+      "learning_rate": 0.0001,
+      "loss": 1.7867,
+      "step": 12866
+    },
+    {
+      "epoch": 1.4948591344757478,
+      "grad_norm": 0.6253650784492493,
+      "learning_rate": 0.0001,
+      "loss": 1.567,
+      "step": 12867
+    },
+    {
+      "epoch": 1.4949753122277083,
+      "grad_norm": 0.6499916911125183,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 12868
+    },
+    {
+      "epoch": 1.4950914899796688,
+      "grad_norm": 0.6280500888824463,
+      "learning_rate": 0.0001,
+      "loss": 1.421,
+      "step": 12869
+    },
+    {
+      "epoch": 1.4952076677316293,
+      "grad_norm": 0.6782636046409607,
+      "learning_rate": 0.0001,
+      "loss": 1.6919,
+      "step": 12870
+    },
+    {
+      "epoch": 1.4953238454835898,
+      "grad_norm": 0.6510661840438843,
+      "learning_rate": 0.0001,
+      "loss": 1.6278,
+      "step": 12871
+    },
+    {
+      "epoch": 1.4954400232355505,
+      "grad_norm": 0.6078386306762695,
+      "learning_rate": 0.0001,
+      "loss": 1.2536,
+      "step": 12872
+    },
+    {
+      "epoch": 1.495556200987511,
+      "grad_norm": 0.5885645747184753,
+      "learning_rate": 0.0001,
+      "loss": 1.3384,
+      "step": 12873
+    },
+    {
+      "epoch": 1.4956723787394715,
+      "grad_norm": 0.560592770576477,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 12874
+    },
+    {
+      "epoch": 1.495788556491432,
+      "grad_norm": 0.6519601345062256,
+      "learning_rate": 0.0001,
+      "loss": 1.4626,
+      "step": 12875
+    },
+    {
+      "epoch": 1.4959047342433924,
+      "grad_norm": 0.6099236607551575,
+      "learning_rate": 0.0001,
+      "loss": 1.4496,
+      "step": 12876
+    },
+    {
+      "epoch": 1.496020911995353,
+      "grad_norm": 0.5811101794242859,
+      "learning_rate": 0.0001,
+      "loss": 1.4068,
+      "step": 12877
+    },
+    {
+      "epoch": 1.4961370897473134,
+      "grad_norm": 0.5691701769828796,
+      "learning_rate": 0.0001,
+      "loss": 1.3746,
+      "step": 12878
+    },
+    {
+      "epoch": 1.496253267499274,
+      "grad_norm": 0.568541944026947,
+      "learning_rate": 0.0001,
+      "loss": 1.2409,
+      "step": 12879
+    },
+    {
+      "epoch": 1.4963694452512344,
+      "grad_norm": 0.6283146142959595,
+      "learning_rate": 0.0001,
+      "loss": 1.4772,
+      "step": 12880
+    },
+    {
+      "epoch": 1.4964856230031949,
+      "grad_norm": 0.6234946250915527,
+      "learning_rate": 0.0001,
+      "loss": 1.5406,
+      "step": 12881
+    },
+    {
+      "epoch": 1.4966018007551554,
+      "grad_norm": 0.6640174984931946,
+      "learning_rate": 0.0001,
+      "loss": 1.756,
+      "step": 12882
+    },
+    {
+      "epoch": 1.4967179785071159,
+      "grad_norm": 0.645535945892334,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 12883
+    },
+    {
+      "epoch": 1.4968341562590763,
+      "grad_norm": 0.615752637386322,
+      "learning_rate": 0.0001,
+      "loss": 1.5753,
+      "step": 12884
+    },
+    {
+      "epoch": 1.4969503340110368,
+      "grad_norm": 0.6437330842018127,
+      "learning_rate": 0.0001,
+      "loss": 1.764,
+      "step": 12885
+    },
+    {
+      "epoch": 1.4970665117629973,
+      "grad_norm": 0.633956789970398,
+      "learning_rate": 0.0001,
+      "loss": 1.6816,
+      "step": 12886
+    },
+    {
+      "epoch": 1.4971826895149578,
+      "grad_norm": 0.625065803527832,
+      "learning_rate": 0.0001,
+      "loss": 1.6142,
+      "step": 12887
+    },
+    {
+      "epoch": 1.4972988672669185,
+      "grad_norm": 0.5906104445457458,
+      "learning_rate": 0.0001,
+      "loss": 1.2654,
+      "step": 12888
+    },
+    {
+      "epoch": 1.497415045018879,
+      "grad_norm": 0.6069050431251526,
+      "learning_rate": 0.0001,
+      "loss": 1.5332,
+      "step": 12889
+    },
+    {
+      "epoch": 1.4975312227708395,
+      "grad_norm": 0.5757995247840881,
+      "learning_rate": 0.0001,
+      "loss": 1.2766,
+      "step": 12890
+    },
+    {
+      "epoch": 1.4976474005228,
+      "grad_norm": 0.6138774752616882,
+      "learning_rate": 0.0001,
+      "loss": 1.4732,
+      "step": 12891
+    },
+    {
+      "epoch": 1.4977635782747605,
+      "grad_norm": 0.6091035008430481,
+      "learning_rate": 0.0001,
+      "loss": 1.4733,
+      "step": 12892
+    },
+    {
+      "epoch": 1.497879756026721,
+      "grad_norm": 0.5950594544410706,
+      "learning_rate": 0.0001,
+      "loss": 1.524,
+      "step": 12893
+    },
+    {
+      "epoch": 1.4979959337786815,
+      "grad_norm": 0.6246311664581299,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 12894
+    },
+    {
+      "epoch": 1.498112111530642,
+      "grad_norm": 0.5842698216438293,
+      "learning_rate": 0.0001,
+      "loss": 1.365,
+      "step": 12895
+    },
+    {
+      "epoch": 1.4982282892826024,
+      "grad_norm": 0.6111376881599426,
+      "learning_rate": 0.0001,
+      "loss": 1.3392,
+      "step": 12896
+    },
+    {
+      "epoch": 1.498344467034563,
+      "grad_norm": 0.5652458667755127,
+      "learning_rate": 0.0001,
+      "loss": 1.2803,
+      "step": 12897
+    },
+    {
+      "epoch": 1.4984606447865234,
+      "grad_norm": 0.600919246673584,
+      "learning_rate": 0.0001,
+      "loss": 1.4203,
+      "step": 12898
+    },
+    {
+      "epoch": 1.498576822538484,
+      "grad_norm": 0.5875718593597412,
+      "learning_rate": 0.0001,
+      "loss": 1.2227,
+      "step": 12899
+    },
+    {
+      "epoch": 1.4986930002904444,
+      "grad_norm": 0.6111353635787964,
+      "learning_rate": 0.0001,
+      "loss": 1.3412,
+      "step": 12900
+    },
+    {
+      "epoch": 1.4988091780424049,
+      "grad_norm": 0.6344109177589417,
+      "learning_rate": 0.0001,
+      "loss": 1.7502,
+      "step": 12901
+    },
+    {
+      "epoch": 1.4989253557943654,
+      "grad_norm": 0.6019986867904663,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 12902
+    },
+    {
+      "epoch": 1.4990415335463259,
+      "grad_norm": 0.6078724265098572,
+      "learning_rate": 0.0001,
+      "loss": 1.5648,
+      "step": 12903
+    },
+    {
+      "epoch": 1.4991577112982863,
+      "grad_norm": 0.6435593962669373,
+      "learning_rate": 0.0001,
+      "loss": 1.5631,
+      "step": 12904
+    },
+    {
+      "epoch": 1.4992738890502468,
+      "grad_norm": 0.661933422088623,
+      "learning_rate": 0.0001,
+      "loss": 1.4889,
+      "step": 12905
+    },
+    {
+      "epoch": 1.4993900668022073,
+      "grad_norm": 0.6397930383682251,
+      "learning_rate": 0.0001,
+      "loss": 1.6223,
+      "step": 12906
+    },
+    {
+      "epoch": 1.4995062445541678,
+      "grad_norm": 0.5697651505470276,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 12907
+    },
+    {
+      "epoch": 1.4996224223061283,
+      "grad_norm": 0.5920810103416443,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 12908
+    },
+    {
+      "epoch": 1.4997386000580888,
+      "grad_norm": 0.6010572910308838,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 12909
+    },
+    {
+      "epoch": 1.4998547778100493,
+      "grad_norm": 0.5660801529884338,
+      "learning_rate": 0.0001,
+      "loss": 1.4184,
+      "step": 12910
+    },
+    {
+      "epoch": 1.4999709555620098,
+      "grad_norm": 0.5613225102424622,
+      "learning_rate": 0.0001,
+      "loss": 1.3346,
+      "step": 12911
+    },
+    {
+      "epoch": 1.5000871333139703,
+      "grad_norm": 0.6039864420890808,
+      "learning_rate": 0.0001,
+      "loss": 1.495,
+      "step": 12912
+    },
+    {
+      "epoch": 1.5002033110659307,
+      "grad_norm": 0.6184871196746826,
+      "learning_rate": 0.0001,
+      "loss": 1.595,
+      "step": 12913
+    },
+    {
+      "epoch": 1.5003194888178912,
+      "grad_norm": 0.602561891078949,
+      "learning_rate": 0.0001,
+      "loss": 1.4573,
+      "step": 12914
+    },
+    {
+      "epoch": 1.5004356665698517,
+      "grad_norm": 0.597723126411438,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 12915
+    },
+    {
+      "epoch": 1.5005518443218122,
+      "grad_norm": 0.6435630917549133,
+      "learning_rate": 0.0001,
+      "loss": 1.5864,
+      "step": 12916
+    },
+    {
+      "epoch": 1.500668022073773,
+      "grad_norm": 0.6061123013496399,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 12917
+    },
+    {
+      "epoch": 1.5007841998257334,
+      "grad_norm": 0.5902183055877686,
+      "learning_rate": 0.0001,
+      "loss": 1.5089,
+      "step": 12918
+    },
+    {
+      "epoch": 1.500900377577694,
+      "grad_norm": 0.6924389600753784,
+      "learning_rate": 0.0001,
+      "loss": 1.552,
+      "step": 12919
+    },
+    {
+      "epoch": 1.5010165553296544,
+      "grad_norm": 0.6402321457862854,
+      "learning_rate": 0.0001,
+      "loss": 1.4686,
+      "step": 12920
+    },
+    {
+      "epoch": 1.5011327330816149,
+      "grad_norm": 0.6038404703140259,
+      "learning_rate": 0.0001,
+      "loss": 1.452,
+      "step": 12921
+    },
+    {
+      "epoch": 1.5012489108335754,
+      "grad_norm": 0.5710585713386536,
+      "learning_rate": 0.0001,
+      "loss": 1.5174,
+      "step": 12922
+    },
+    {
+      "epoch": 1.5013650885855359,
+      "grad_norm": 0.6276952624320984,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 12923
+    },
+    {
+      "epoch": 1.5014812663374963,
+      "grad_norm": 0.6179443001747131,
+      "learning_rate": 0.0001,
+      "loss": 1.4969,
+      "step": 12924
+    },
+    {
+      "epoch": 1.5015974440894568,
+      "grad_norm": 0.648830235004425,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 12925
+    },
+    {
+      "epoch": 1.5017136218414173,
+      "grad_norm": 0.6248328685760498,
+      "learning_rate": 0.0001,
+      "loss": 1.6314,
+      "step": 12926
+    },
+    {
+      "epoch": 1.501829799593378,
+      "grad_norm": 0.6510769724845886,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 12927
+    },
+    {
+      "epoch": 1.5019459773453385,
+      "grad_norm": 0.6090167760848999,
+      "learning_rate": 0.0001,
+      "loss": 1.3349,
+      "step": 12928
+    },
+    {
+      "epoch": 1.502062155097299,
+      "grad_norm": 0.5935208201408386,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 12929
+    },
+    {
+      "epoch": 1.5021783328492595,
+      "grad_norm": 0.6401615142822266,
+      "learning_rate": 0.0001,
+      "loss": 1.4513,
+      "step": 12930
+    },
+    {
+      "epoch": 1.50229451060122,
+      "grad_norm": 0.6500523090362549,
+      "learning_rate": 0.0001,
+      "loss": 1.5076,
+      "step": 12931
+    },
+    {
+      "epoch": 1.5024106883531805,
+      "grad_norm": 0.6160486936569214,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 12932
+    },
+    {
+      "epoch": 1.502526866105141,
+      "grad_norm": 0.6370952129364014,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 12933
+    },
+    {
+      "epoch": 1.5026430438571015,
+      "grad_norm": 0.6076738834381104,
+      "learning_rate": 0.0001,
+      "loss": 1.5261,
+      "step": 12934
+    },
+    {
+      "epoch": 1.502759221609062,
+      "grad_norm": 0.5521711707115173,
+      "learning_rate": 0.0001,
+      "loss": 1.3567,
+      "step": 12935
+    },
+    {
+      "epoch": 1.5028753993610224,
+      "grad_norm": 0.571445643901825,
+      "learning_rate": 0.0001,
+      "loss": 1.198,
+      "step": 12936
+    },
+    {
+      "epoch": 1.502991577112983,
+      "grad_norm": 0.6402273178100586,
+      "learning_rate": 0.0001,
+      "loss": 1.5463,
+      "step": 12937
+    },
+    {
+      "epoch": 1.5031077548649434,
+      "grad_norm": 0.6269349455833435,
+      "learning_rate": 0.0001,
+      "loss": 1.5242,
+      "step": 12938
+    },
+    {
+      "epoch": 1.503223932616904,
+      "grad_norm": 0.6312558054924011,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 12939
+    },
+    {
+      "epoch": 1.5033401103688644,
+      "grad_norm": 0.6417350172996521,
+      "learning_rate": 0.0001,
+      "loss": 1.5901,
+      "step": 12940
+    },
+    {
+      "epoch": 1.5034562881208249,
+      "grad_norm": 0.6578287482261658,
+      "learning_rate": 0.0001,
+      "loss": 1.5081,
+      "step": 12941
+    },
+    {
+      "epoch": 1.5035724658727854,
+      "grad_norm": 0.5819112062454224,
+      "learning_rate": 0.0001,
+      "loss": 1.3784,
+      "step": 12942
+    },
+    {
+      "epoch": 1.5036886436247459,
+      "grad_norm": 0.6250476837158203,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 12943
+    },
+    {
+      "epoch": 1.5038048213767063,
+      "grad_norm": 0.7036281824111938,
+      "learning_rate": 0.0001,
+      "loss": 1.5268,
+      "step": 12944
+    },
+    {
+      "epoch": 1.5039209991286668,
+      "grad_norm": 0.5976950526237488,
+      "learning_rate": 0.0001,
+      "loss": 1.3829,
+      "step": 12945
+    },
+    {
+      "epoch": 1.5040371768806273,
+      "grad_norm": 0.6492063403129578,
+      "learning_rate": 0.0001,
+      "loss": 1.5779,
+      "step": 12946
+    },
+    {
+      "epoch": 1.5041533546325878,
+      "grad_norm": 0.5875254273414612,
+      "learning_rate": 0.0001,
+      "loss": 1.4128,
+      "step": 12947
+    },
+    {
+      "epoch": 1.5042695323845483,
+      "grad_norm": 0.6553013324737549,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 12948
+    },
+    {
+      "epoch": 1.5043857101365088,
+      "grad_norm": 0.6260391473770142,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 12949
+    },
+    {
+      "epoch": 1.5045018878884693,
+      "grad_norm": 0.7011594772338867,
+      "learning_rate": 0.0001,
+      "loss": 1.5969,
+      "step": 12950
+    },
+    {
+      "epoch": 1.5046180656404298,
+      "grad_norm": 0.6216688752174377,
+      "learning_rate": 0.0001,
+      "loss": 1.4793,
+      "step": 12951
+    },
+    {
+      "epoch": 1.5047342433923903,
+      "grad_norm": 0.6175325512886047,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 12952
+    },
+    {
+      "epoch": 1.5048504211443507,
+      "grad_norm": 0.657291829586029,
+      "learning_rate": 0.0001,
+      "loss": 1.5744,
+      "step": 12953
+    },
+    {
+      "epoch": 1.5049665988963112,
+      "grad_norm": 0.6169841289520264,
+      "learning_rate": 0.0001,
+      "loss": 1.4183,
+      "step": 12954
+    },
+    {
+      "epoch": 1.5050827766482717,
+      "grad_norm": 0.6225051879882812,
+      "learning_rate": 0.0001,
+      "loss": 1.3909,
+      "step": 12955
+    },
+    {
+      "epoch": 1.5051989544002322,
+      "grad_norm": 0.7212052345275879,
+      "learning_rate": 0.0001,
+      "loss": 1.5745,
+      "step": 12956
+    },
+    {
+      "epoch": 1.5053151321521927,
+      "grad_norm": 0.6326794028282166,
+      "learning_rate": 0.0001,
+      "loss": 1.4791,
+      "step": 12957
+    },
+    {
+      "epoch": 1.5054313099041532,
+      "grad_norm": 0.6327944397926331,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 12958
+    },
+    {
+      "epoch": 1.505547487656114,
+      "grad_norm": 0.580574095249176,
+      "learning_rate": 0.0001,
+      "loss": 1.3895,
+      "step": 12959
+    },
+    {
+      "epoch": 1.5056636654080744,
+      "grad_norm": 0.7401102781295776,
+      "learning_rate": 0.0001,
+      "loss": 1.5773,
+      "step": 12960
+    },
+    {
+      "epoch": 1.5057798431600349,
+      "grad_norm": 0.5871754288673401,
+      "learning_rate": 0.0001,
+      "loss": 1.4968,
+      "step": 12961
+    },
+    {
+      "epoch": 1.5058960209119954,
+      "grad_norm": 0.6001573801040649,
+      "learning_rate": 0.0001,
+      "loss": 1.4837,
+      "step": 12962
+    },
+    {
+      "epoch": 1.5060121986639559,
+      "grad_norm": 0.537681519985199,
+      "learning_rate": 0.0001,
+      "loss": 1.371,
+      "step": 12963
+    },
+    {
+      "epoch": 1.5061283764159163,
+      "grad_norm": 0.6001483201980591,
+      "learning_rate": 0.0001,
+      "loss": 1.4206,
+      "step": 12964
+    },
+    {
+      "epoch": 1.5062445541678768,
+      "grad_norm": 0.6050352454185486,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 12965
+    },
+    {
+      "epoch": 1.5063607319198373,
+      "grad_norm": 0.5874537825584412,
+      "learning_rate": 0.0001,
+      "loss": 1.3447,
+      "step": 12966
+    },
+    {
+      "epoch": 1.5064769096717978,
+      "grad_norm": 0.6636204123497009,
+      "learning_rate": 0.0001,
+      "loss": 1.4919,
+      "step": 12967
+    },
+    {
+      "epoch": 1.5065930874237583,
+      "grad_norm": 0.6525164246559143,
+      "learning_rate": 0.0001,
+      "loss": 1.7726,
+      "step": 12968
+    },
+    {
+      "epoch": 1.506709265175719,
+      "grad_norm": 0.6085329055786133,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 12969
+    },
+    {
+      "epoch": 1.5068254429276795,
+      "grad_norm": 0.6257724165916443,
+      "learning_rate": 0.0001,
+      "loss": 1.5168,
+      "step": 12970
+    },
+    {
+      "epoch": 1.50694162067964,
+      "grad_norm": 0.6245921850204468,
+      "learning_rate": 0.0001,
+      "loss": 1.4173,
+      "step": 12971
+    },
+    {
+      "epoch": 1.5070577984316005,
+      "grad_norm": 0.6131578087806702,
+      "learning_rate": 0.0001,
+      "loss": 1.3337,
+      "step": 12972
+    },
+    {
+      "epoch": 1.507173976183561,
+      "grad_norm": 0.6182470321655273,
+      "learning_rate": 0.0001,
+      "loss": 1.5111,
+      "step": 12973
+    },
+    {
+      "epoch": 1.5072901539355215,
+      "grad_norm": 0.6009563207626343,
+      "learning_rate": 0.0001,
+      "loss": 1.4442,
+      "step": 12974
+    },
+    {
+      "epoch": 1.507406331687482,
+      "grad_norm": 0.5851460099220276,
+      "learning_rate": 0.0001,
+      "loss": 1.4297,
+      "step": 12975
+    },
+    {
+      "epoch": 1.5075225094394424,
+      "grad_norm": 0.5955979824066162,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 12976
+    },
+    {
+      "epoch": 1.507638687191403,
+      "grad_norm": 0.6064378619194031,
+      "learning_rate": 0.0001,
+      "loss": 1.4586,
+      "step": 12977
+    },
+    {
+      "epoch": 1.5077548649433634,
+      "grad_norm": 0.5852058529853821,
+      "learning_rate": 0.0001,
+      "loss": 1.2852,
+      "step": 12978
+    },
+    {
+      "epoch": 1.507871042695324,
+      "grad_norm": 0.5747120380401611,
+      "learning_rate": 0.0001,
+      "loss": 1.4461,
+      "step": 12979
+    },
+    {
+      "epoch": 1.5079872204472844,
+      "grad_norm": 0.566041886806488,
+      "learning_rate": 0.0001,
+      "loss": 1.4831,
+      "step": 12980
+    },
+    {
+      "epoch": 1.5081033981992449,
+      "grad_norm": 0.6108693480491638,
+      "learning_rate": 0.0001,
+      "loss": 1.3197,
+      "step": 12981
+    },
+    {
+      "epoch": 1.5082195759512054,
+      "grad_norm": 0.6155402660369873,
+      "learning_rate": 0.0001,
+      "loss": 1.5253,
+      "step": 12982
+    },
+    {
+      "epoch": 1.5083357537031659,
+      "grad_norm": 0.6545661687850952,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 12983
+    },
+    {
+      "epoch": 1.5084519314551263,
+      "grad_norm": 0.6387633085250854,
+      "learning_rate": 0.0001,
+      "loss": 1.3534,
+      "step": 12984
+    },
+    {
+      "epoch": 1.5085681092070868,
+      "grad_norm": 0.6138156056404114,
+      "learning_rate": 0.0001,
+      "loss": 1.4418,
+      "step": 12985
+    },
+    {
+      "epoch": 1.5086842869590473,
+      "grad_norm": 0.6812984347343445,
+      "learning_rate": 0.0001,
+      "loss": 1.5453,
+      "step": 12986
+    },
+    {
+      "epoch": 1.5088004647110078,
+      "grad_norm": 0.5979045033454895,
+      "learning_rate": 0.0001,
+      "loss": 1.3903,
+      "step": 12987
+    },
+    {
+      "epoch": 1.5089166424629683,
+      "grad_norm": 0.5794659852981567,
+      "learning_rate": 0.0001,
+      "loss": 1.4501,
+      "step": 12988
+    },
+    {
+      "epoch": 1.5090328202149288,
+      "grad_norm": 0.6213316321372986,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 12989
+    },
+    {
+      "epoch": 1.5091489979668893,
+      "grad_norm": 0.6868020296096802,
+      "learning_rate": 0.0001,
+      "loss": 1.6787,
+      "step": 12990
+    },
+    {
+      "epoch": 1.5092651757188498,
+      "grad_norm": 0.6285677552223206,
+      "learning_rate": 0.0001,
+      "loss": 1.5757,
+      "step": 12991
+    },
+    {
+      "epoch": 1.5093813534708103,
+      "grad_norm": 0.6092104315757751,
+      "learning_rate": 0.0001,
+      "loss": 1.5592,
+      "step": 12992
+    },
+    {
+      "epoch": 1.5094975312227707,
+      "grad_norm": 0.5800586342811584,
+      "learning_rate": 0.0001,
+      "loss": 1.413,
+      "step": 12993
+    },
+    {
+      "epoch": 1.5096137089747312,
+      "grad_norm": 0.6098979711532593,
+      "learning_rate": 0.0001,
+      "loss": 1.6554,
+      "step": 12994
+    },
+    {
+      "epoch": 1.5097298867266917,
+      "grad_norm": 0.5925366878509521,
+      "learning_rate": 0.0001,
+      "loss": 1.6048,
+      "step": 12995
+    },
+    {
+      "epoch": 1.5098460644786522,
+      "grad_norm": 0.5455504059791565,
+      "learning_rate": 0.0001,
+      "loss": 1.3888,
+      "step": 12996
+    },
+    {
+      "epoch": 1.5099622422306127,
+      "grad_norm": 0.6356263160705566,
+      "learning_rate": 0.0001,
+      "loss": 1.6256,
+      "step": 12997
+    },
+    {
+      "epoch": 1.5100784199825732,
+      "grad_norm": 0.6532312035560608,
+      "learning_rate": 0.0001,
+      "loss": 1.6825,
+      "step": 12998
+    },
+    {
+      "epoch": 1.5101945977345337,
+      "grad_norm": 0.6413785219192505,
+      "learning_rate": 0.0001,
+      "loss": 1.5933,
+      "step": 12999
+    },
+    {
+      "epoch": 1.5103107754864942,
+      "grad_norm": 0.5889772176742554,
+      "learning_rate": 0.0001,
+      "loss": 1.3492,
+      "step": 13000
+    },
+    {
+      "epoch": 1.5104269532384549,
+      "grad_norm": 0.5992516279220581,
+      "learning_rate": 0.0001,
+      "loss": 1.4146,
+      "step": 13001
+    },
+    {
+      "epoch": 1.5105431309904154,
+      "grad_norm": 0.5487164855003357,
+      "learning_rate": 0.0001,
+      "loss": 1.393,
+      "step": 13002
+    },
+    {
+      "epoch": 1.5106593087423759,
+      "grad_norm": 0.6081210970878601,
+      "learning_rate": 0.0001,
+      "loss": 1.3813,
+      "step": 13003
+    },
+    {
+      "epoch": 1.5107754864943364,
+      "grad_norm": 0.6229928731918335,
+      "learning_rate": 0.0001,
+      "loss": 1.4805,
+      "step": 13004
+    },
+    {
+      "epoch": 1.5108916642462968,
+      "grad_norm": 0.6931246519088745,
+      "learning_rate": 0.0001,
+      "loss": 1.8908,
+      "step": 13005
+    },
+    {
+      "epoch": 1.5110078419982573,
+      "grad_norm": 0.5892548561096191,
+      "learning_rate": 0.0001,
+      "loss": 1.4452,
+      "step": 13006
+    },
+    {
+      "epoch": 1.5111240197502178,
+      "grad_norm": 0.6359273791313171,
+      "learning_rate": 0.0001,
+      "loss": 1.4474,
+      "step": 13007
+    },
+    {
+      "epoch": 1.5112401975021783,
+      "grad_norm": 0.6264612078666687,
+      "learning_rate": 0.0001,
+      "loss": 1.5436,
+      "step": 13008
+    },
+    {
+      "epoch": 1.5113563752541388,
+      "grad_norm": 0.6442233324050903,
+      "learning_rate": 0.0001,
+      "loss": 1.3834,
+      "step": 13009
+    },
+    {
+      "epoch": 1.5114725530060993,
+      "grad_norm": 0.6219121217727661,
+      "learning_rate": 0.0001,
+      "loss": 1.2729,
+      "step": 13010
+    },
+    {
+      "epoch": 1.51158873075806,
+      "grad_norm": 0.6267338991165161,
+      "learning_rate": 0.0001,
+      "loss": 1.4941,
+      "step": 13011
+    },
+    {
+      "epoch": 1.5117049085100205,
+      "grad_norm": 0.5808162689208984,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 13012
+    },
+    {
+      "epoch": 1.511821086261981,
+      "grad_norm": 0.6579829454421997,
+      "learning_rate": 0.0001,
+      "loss": 1.4468,
+      "step": 13013
+    },
+    {
+      "epoch": 1.5119372640139415,
+      "grad_norm": 0.593537449836731,
+      "learning_rate": 0.0001,
+      "loss": 1.4952,
+      "step": 13014
+    },
+    {
+      "epoch": 1.512053441765902,
+      "grad_norm": 0.538985550403595,
+      "learning_rate": 0.0001,
+      "loss": 1.2365,
+      "step": 13015
+    },
+    {
+      "epoch": 1.5121696195178624,
+      "grad_norm": 0.6082078814506531,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 13016
+    },
+    {
+      "epoch": 1.512285797269823,
+      "grad_norm": 0.6273752450942993,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 13017
+    },
+    {
+      "epoch": 1.5124019750217834,
+      "grad_norm": 0.6424401998519897,
+      "learning_rate": 0.0001,
+      "loss": 1.5122,
+      "step": 13018
+    },
+    {
+      "epoch": 1.512518152773744,
+      "grad_norm": 0.5700364708900452,
+      "learning_rate": 0.0001,
+      "loss": 1.2343,
+      "step": 13019
+    },
+    {
+      "epoch": 1.5126343305257044,
+      "grad_norm": 0.6058292388916016,
+      "learning_rate": 0.0001,
+      "loss": 1.3065,
+      "step": 13020
+    },
+    {
+      "epoch": 1.5127505082776649,
+      "grad_norm": 0.5938413143157959,
+      "learning_rate": 0.0001,
+      "loss": 1.4116,
+      "step": 13021
+    },
+    {
+      "epoch": 1.5128666860296254,
+      "grad_norm": 0.5882897973060608,
+      "learning_rate": 0.0001,
+      "loss": 1.4321,
+      "step": 13022
+    },
+    {
+      "epoch": 1.5129828637815859,
+      "grad_norm": 0.6097550988197327,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 13023
+    },
+    {
+      "epoch": 1.5130990415335464,
+      "grad_norm": 0.6150256395339966,
+      "learning_rate": 0.0001,
+      "loss": 1.4294,
+      "step": 13024
+    },
+    {
+      "epoch": 1.5132152192855068,
+      "grad_norm": 0.5661260485649109,
+      "learning_rate": 0.0001,
+      "loss": 1.3101,
+      "step": 13025
+    },
+    {
+      "epoch": 1.5133313970374673,
+      "grad_norm": 0.6101265549659729,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 13026
+    },
+    {
+      "epoch": 1.5134475747894278,
+      "grad_norm": 0.5510701537132263,
+      "learning_rate": 0.0001,
+      "loss": 1.2253,
+      "step": 13027
+    },
+    {
+      "epoch": 1.5135637525413883,
+      "grad_norm": 0.5895761847496033,
+      "learning_rate": 0.0001,
+      "loss": 1.2103,
+      "step": 13028
+    },
+    {
+      "epoch": 1.5136799302933488,
+      "grad_norm": 0.6410093903541565,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 13029
+    },
+    {
+      "epoch": 1.5137961080453093,
+      "grad_norm": 0.6049452424049377,
+      "learning_rate": 0.0001,
+      "loss": 1.4493,
+      "step": 13030
+    },
+    {
+      "epoch": 1.5139122857972698,
+      "grad_norm": 0.5991563200950623,
+      "learning_rate": 0.0001,
+      "loss": 1.369,
+      "step": 13031
+    },
+    {
+      "epoch": 1.5140284635492303,
+      "grad_norm": 0.6251417398452759,
+      "learning_rate": 0.0001,
+      "loss": 1.4542,
+      "step": 13032
+    },
+    {
+      "epoch": 1.5141446413011908,
+      "grad_norm": 0.6023422479629517,
+      "learning_rate": 0.0001,
+      "loss": 1.3901,
+      "step": 13033
+    },
+    {
+      "epoch": 1.5142608190531512,
+      "grad_norm": 0.6075770854949951,
+      "learning_rate": 0.0001,
+      "loss": 1.4216,
+      "step": 13034
+    },
+    {
+      "epoch": 1.5143769968051117,
+      "grad_norm": 0.5965530872344971,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 13035
+    },
+    {
+      "epoch": 1.5144931745570722,
+      "grad_norm": 0.5827516913414001,
+      "learning_rate": 0.0001,
+      "loss": 1.3238,
+      "step": 13036
+    },
+    {
+      "epoch": 1.5146093523090327,
+      "grad_norm": 0.5984660387039185,
+      "learning_rate": 0.0001,
+      "loss": 1.5236,
+      "step": 13037
+    },
+    {
+      "epoch": 1.5147255300609932,
+      "grad_norm": 0.6355912685394287,
+      "learning_rate": 0.0001,
+      "loss": 1.6367,
+      "step": 13038
+    },
+    {
+      "epoch": 1.5148417078129537,
+      "grad_norm": 0.6122537851333618,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 13039
+    },
+    {
+      "epoch": 1.5149578855649142,
+      "grad_norm": 0.5791593790054321,
+      "learning_rate": 0.0001,
+      "loss": 1.3258,
+      "step": 13040
+    },
+    {
+      "epoch": 1.5150740633168747,
+      "grad_norm": 0.604680061340332,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 13041
+    },
+    {
+      "epoch": 1.5151902410688352,
+      "grad_norm": 0.610745370388031,
+      "learning_rate": 0.0001,
+      "loss": 1.2757,
+      "step": 13042
+    },
+    {
+      "epoch": 1.5153064188207959,
+      "grad_norm": 0.5982139706611633,
+      "learning_rate": 0.0001,
+      "loss": 1.5444,
+      "step": 13043
+    },
+    {
+      "epoch": 1.5154225965727564,
+      "grad_norm": 0.5831255316734314,
+      "learning_rate": 0.0001,
+      "loss": 1.366,
+      "step": 13044
+    },
+    {
+      "epoch": 1.5155387743247168,
+      "grad_norm": 0.6067568063735962,
+      "learning_rate": 0.0001,
+      "loss": 1.3268,
+      "step": 13045
+    },
+    {
+      "epoch": 1.5156549520766773,
+      "grad_norm": 0.575416088104248,
+      "learning_rate": 0.0001,
+      "loss": 1.4809,
+      "step": 13046
+    },
+    {
+      "epoch": 1.5157711298286378,
+      "grad_norm": 0.6012206673622131,
+      "learning_rate": 0.0001,
+      "loss": 1.4559,
+      "step": 13047
+    },
+    {
+      "epoch": 1.5158873075805983,
+      "grad_norm": 0.5732565522193909,
+      "learning_rate": 0.0001,
+      "loss": 1.3186,
+      "step": 13048
+    },
+    {
+      "epoch": 1.5160034853325588,
+      "grad_norm": 0.5650144815444946,
+      "learning_rate": 0.0001,
+      "loss": 1.4331,
+      "step": 13049
+    },
+    {
+      "epoch": 1.5161196630845193,
+      "grad_norm": 0.6107412576675415,
+      "learning_rate": 0.0001,
+      "loss": 1.388,
+      "step": 13050
+    },
+    {
+      "epoch": 1.5162358408364798,
+      "grad_norm": 0.6341642737388611,
+      "learning_rate": 0.0001,
+      "loss": 1.4583,
+      "step": 13051
+    },
+    {
+      "epoch": 1.5163520185884403,
+      "grad_norm": 0.6973785758018494,
+      "learning_rate": 0.0001,
+      "loss": 1.5455,
+      "step": 13052
+    },
+    {
+      "epoch": 1.516468196340401,
+      "grad_norm": 0.6058968901634216,
+      "learning_rate": 0.0001,
+      "loss": 1.1763,
+      "step": 13053
+    },
+    {
+      "epoch": 1.5165843740923615,
+      "grad_norm": 0.6408393383026123,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 13054
+    },
+    {
+      "epoch": 1.516700551844322,
+      "grad_norm": 0.6071977019309998,
+      "learning_rate": 0.0001,
+      "loss": 1.3627,
+      "step": 13055
+    },
+    {
+      "epoch": 1.5168167295962824,
+      "grad_norm": 0.6299028396606445,
+      "learning_rate": 0.0001,
+      "loss": 1.5713,
+      "step": 13056
+    },
+    {
+      "epoch": 1.516932907348243,
+      "grad_norm": 0.6032760143280029,
+      "learning_rate": 0.0001,
+      "loss": 1.4244,
+      "step": 13057
+    },
+    {
+      "epoch": 1.5170490851002034,
+      "grad_norm": 0.5748342871665955,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 13058
+    },
+    {
+      "epoch": 1.517165262852164,
+      "grad_norm": 0.5807141661643982,
+      "learning_rate": 0.0001,
+      "loss": 1.4516,
+      "step": 13059
+    },
+    {
+      "epoch": 1.5172814406041244,
+      "grad_norm": 0.6238056421279907,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 13060
+    },
+    {
+      "epoch": 1.5173976183560849,
+      "grad_norm": 0.6162986159324646,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 13061
+    },
+    {
+      "epoch": 1.5175137961080454,
+      "grad_norm": 0.6011757254600525,
+      "learning_rate": 0.0001,
+      "loss": 1.4601,
+      "step": 13062
+    },
+    {
+      "epoch": 1.5176299738600059,
+      "grad_norm": 0.6143328547477722,
+      "learning_rate": 0.0001,
+      "loss": 1.4704,
+      "step": 13063
+    },
+    {
+      "epoch": 1.5177461516119664,
+      "grad_norm": 0.5976557731628418,
+      "learning_rate": 0.0001,
+      "loss": 1.3611,
+      "step": 13064
+    },
+    {
+      "epoch": 1.5178623293639268,
+      "grad_norm": 0.6380662322044373,
+      "learning_rate": 0.0001,
+      "loss": 1.5687,
+      "step": 13065
+    },
+    {
+      "epoch": 1.5179785071158873,
+      "grad_norm": 0.6185310482978821,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 13066
+    },
+    {
+      "epoch": 1.5180946848678478,
+      "grad_norm": 0.5588987469673157,
+      "learning_rate": 0.0001,
+      "loss": 1.2781,
+      "step": 13067
+    },
+    {
+      "epoch": 1.5182108626198083,
+      "grad_norm": 0.6087438464164734,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 13068
+    },
+    {
+      "epoch": 1.5183270403717688,
+      "grad_norm": 0.6939107775688171,
+      "learning_rate": 0.0001,
+      "loss": 1.6837,
+      "step": 13069
+    },
+    {
+      "epoch": 1.5184432181237293,
+      "grad_norm": 0.5763146877288818,
+      "learning_rate": 0.0001,
+      "loss": 1.3932,
+      "step": 13070
+    },
+    {
+      "epoch": 1.5185593958756898,
+      "grad_norm": 0.5791681408882141,
+      "learning_rate": 0.0001,
+      "loss": 1.3638,
+      "step": 13071
+    },
+    {
+      "epoch": 1.5186755736276503,
+      "grad_norm": 0.6180105209350586,
+      "learning_rate": 0.0001,
+      "loss": 1.5888,
+      "step": 13072
+    },
+    {
+      "epoch": 1.5187917513796108,
+      "grad_norm": 0.6183565258979797,
+      "learning_rate": 0.0001,
+      "loss": 1.5131,
+      "step": 13073
+    },
+    {
+      "epoch": 1.5189079291315712,
+      "grad_norm": 0.6181787848472595,
+      "learning_rate": 0.0001,
+      "loss": 1.4906,
+      "step": 13074
+    },
+    {
+      "epoch": 1.5190241068835317,
+      "grad_norm": 0.6255152225494385,
+      "learning_rate": 0.0001,
+      "loss": 1.472,
+      "step": 13075
+    },
+    {
+      "epoch": 1.5191402846354922,
+      "grad_norm": 0.57309889793396,
+      "learning_rate": 0.0001,
+      "loss": 1.3557,
+      "step": 13076
+    },
+    {
+      "epoch": 1.5192564623874527,
+      "grad_norm": 0.5992699861526489,
+      "learning_rate": 0.0001,
+      "loss": 1.2967,
+      "step": 13077
+    },
+    {
+      "epoch": 1.5193726401394132,
+      "grad_norm": 0.6498000025749207,
+      "learning_rate": 0.0001,
+      "loss": 1.6267,
+      "step": 13078
+    },
+    {
+      "epoch": 1.5194888178913737,
+      "grad_norm": 0.6142323017120361,
+      "learning_rate": 0.0001,
+      "loss": 1.3953,
+      "step": 13079
+    },
+    {
+      "epoch": 1.5196049956433342,
+      "grad_norm": 0.5629411935806274,
+      "learning_rate": 0.0001,
+      "loss": 1.4722,
+      "step": 13080
+    },
+    {
+      "epoch": 1.5197211733952947,
+      "grad_norm": 0.6026464104652405,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 13081
+    },
+    {
+      "epoch": 1.5198373511472552,
+      "grad_norm": 0.5876480340957642,
+      "learning_rate": 0.0001,
+      "loss": 1.42,
+      "step": 13082
+    },
+    {
+      "epoch": 1.5199535288992156,
+      "grad_norm": 0.6540488600730896,
+      "learning_rate": 0.0001,
+      "loss": 1.5638,
+      "step": 13083
+    },
+    {
+      "epoch": 1.5200697066511761,
+      "grad_norm": 0.5790266990661621,
+      "learning_rate": 0.0001,
+      "loss": 1.4935,
+      "step": 13084
+    },
+    {
+      "epoch": 1.5201858844031368,
+      "grad_norm": 0.5915881991386414,
+      "learning_rate": 0.0001,
+      "loss": 1.5506,
+      "step": 13085
+    },
+    {
+      "epoch": 1.5203020621550973,
+      "grad_norm": 0.601015567779541,
+      "learning_rate": 0.0001,
+      "loss": 1.554,
+      "step": 13086
+    },
+    {
+      "epoch": 1.5204182399070578,
+      "grad_norm": 0.607069730758667,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 13087
+    },
+    {
+      "epoch": 1.5205344176590183,
+      "grad_norm": 0.5643183588981628,
+      "learning_rate": 0.0001,
+      "loss": 1.4889,
+      "step": 13088
+    },
+    {
+      "epoch": 1.5206505954109788,
+      "grad_norm": 0.6405245065689087,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 13089
+    },
+    {
+      "epoch": 1.5207667731629393,
+      "grad_norm": 0.618649959564209,
+      "learning_rate": 0.0001,
+      "loss": 1.4619,
+      "step": 13090
+    },
+    {
+      "epoch": 1.5208829509148998,
+      "grad_norm": 0.596312403678894,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 13091
+    },
+    {
+      "epoch": 1.5209991286668603,
+      "grad_norm": 0.6227704286575317,
+      "learning_rate": 0.0001,
+      "loss": 1.6598,
+      "step": 13092
+    },
+    {
+      "epoch": 1.5211153064188208,
+      "grad_norm": 0.5877798795700073,
+      "learning_rate": 0.0001,
+      "loss": 1.5022,
+      "step": 13093
+    },
+    {
+      "epoch": 1.5212314841707812,
+      "grad_norm": 0.5980353355407715,
+      "learning_rate": 0.0001,
+      "loss": 1.5944,
+      "step": 13094
+    },
+    {
+      "epoch": 1.521347661922742,
+      "grad_norm": 0.6470990777015686,
+      "learning_rate": 0.0001,
+      "loss": 1.6384,
+      "step": 13095
+    },
+    {
+      "epoch": 1.5214638396747024,
+      "grad_norm": 0.6021008491516113,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 13096
+    },
+    {
+      "epoch": 1.521580017426663,
+      "grad_norm": 0.5772629380226135,
+      "learning_rate": 0.0001,
+      "loss": 1.4815,
+      "step": 13097
+    },
+    {
+      "epoch": 1.5216961951786234,
+      "grad_norm": 0.5909640192985535,
+      "learning_rate": 0.0001,
+      "loss": 1.4743,
+      "step": 13098
+    },
+    {
+      "epoch": 1.521812372930584,
+      "grad_norm": 0.6111878752708435,
+      "learning_rate": 0.0001,
+      "loss": 1.5198,
+      "step": 13099
+    },
+    {
+      "epoch": 1.5219285506825444,
+      "grad_norm": 0.5904512405395508,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 13100
+    },
+    {
+      "epoch": 1.522044728434505,
+      "grad_norm": 0.5810577869415283,
+      "learning_rate": 0.0001,
+      "loss": 1.3642,
+      "step": 13101
+    },
+    {
+      "epoch": 1.5221609061864654,
+      "grad_norm": 0.6800036430358887,
+      "learning_rate": 0.0001,
+      "loss": 1.5471,
+      "step": 13102
+    },
+    {
+      "epoch": 1.5222770839384259,
+      "grad_norm": 0.6236268877983093,
+      "learning_rate": 0.0001,
+      "loss": 1.6704,
+      "step": 13103
+    },
+    {
+      "epoch": 1.5223932616903864,
+      "grad_norm": 0.6043499708175659,
+      "learning_rate": 0.0001,
+      "loss": 1.4506,
+      "step": 13104
+    },
+    {
+      "epoch": 1.5225094394423468,
+      "grad_norm": 0.609677255153656,
+      "learning_rate": 0.0001,
+      "loss": 1.45,
+      "step": 13105
+    },
+    {
+      "epoch": 1.5226256171943073,
+      "grad_norm": 0.5774016976356506,
+      "learning_rate": 0.0001,
+      "loss": 1.4479,
+      "step": 13106
+    },
+    {
+      "epoch": 1.5227417949462678,
+      "grad_norm": 0.6218743324279785,
+      "learning_rate": 0.0001,
+      "loss": 1.5468,
+      "step": 13107
+    },
+    {
+      "epoch": 1.5228579726982283,
+      "grad_norm": 0.5599686503410339,
+      "learning_rate": 0.0001,
+      "loss": 1.3028,
+      "step": 13108
+    },
+    {
+      "epoch": 1.5229741504501888,
+      "grad_norm": 0.5793375372886658,
+      "learning_rate": 0.0001,
+      "loss": 1.3442,
+      "step": 13109
+    },
+    {
+      "epoch": 1.5230903282021493,
+      "grad_norm": 0.5650421380996704,
+      "learning_rate": 0.0001,
+      "loss": 1.1701,
+      "step": 13110
+    },
+    {
+      "epoch": 1.5232065059541098,
+      "grad_norm": 0.628027617931366,
+      "learning_rate": 0.0001,
+      "loss": 1.4018,
+      "step": 13111
+    },
+    {
+      "epoch": 1.5233226837060703,
+      "grad_norm": 0.6404212117195129,
+      "learning_rate": 0.0001,
+      "loss": 1.5248,
+      "step": 13112
+    },
+    {
+      "epoch": 1.5234388614580308,
+      "grad_norm": 0.6501476168632507,
+      "learning_rate": 0.0001,
+      "loss": 1.6086,
+      "step": 13113
+    },
+    {
+      "epoch": 1.5235550392099912,
+      "grad_norm": 0.6205679774284363,
+      "learning_rate": 0.0001,
+      "loss": 1.5886,
+      "step": 13114
+    },
+    {
+      "epoch": 1.5236712169619517,
+      "grad_norm": 0.6271020174026489,
+      "learning_rate": 0.0001,
+      "loss": 1.4695,
+      "step": 13115
+    },
+    {
+      "epoch": 1.5237873947139122,
+      "grad_norm": 0.6364188194274902,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 13116
+    },
+    {
+      "epoch": 1.5239035724658727,
+      "grad_norm": 0.6291508674621582,
+      "learning_rate": 0.0001,
+      "loss": 1.4255,
+      "step": 13117
+    },
+    {
+      "epoch": 1.5240197502178332,
+      "grad_norm": 0.6832680106163025,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 13118
+    },
+    {
+      "epoch": 1.5241359279697937,
+      "grad_norm": 0.6287549138069153,
+      "learning_rate": 0.0001,
+      "loss": 1.6068,
+      "step": 13119
+    },
+    {
+      "epoch": 1.5242521057217542,
+      "grad_norm": 0.6037324070930481,
+      "learning_rate": 0.0001,
+      "loss": 1.3475,
+      "step": 13120
+    },
+    {
+      "epoch": 1.5243682834737147,
+      "grad_norm": 0.6374009847640991,
+      "learning_rate": 0.0001,
+      "loss": 1.544,
+      "step": 13121
+    },
+    {
+      "epoch": 1.5244844612256752,
+      "grad_norm": 0.5533431172370911,
+      "learning_rate": 0.0001,
+      "loss": 1.4256,
+      "step": 13122
+    },
+    {
+      "epoch": 1.5246006389776356,
+      "grad_norm": 0.6008569002151489,
+      "learning_rate": 0.0001,
+      "loss": 1.4351,
+      "step": 13123
+    },
+    {
+      "epoch": 1.5247168167295961,
+      "grad_norm": 0.5492884516716003,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 13124
+    },
+    {
+      "epoch": 1.5248329944815566,
+      "grad_norm": 0.6348086595535278,
+      "learning_rate": 0.0001,
+      "loss": 1.6349,
+      "step": 13125
+    },
+    {
+      "epoch": 1.5249491722335173,
+      "grad_norm": 0.5985419750213623,
+      "learning_rate": 0.0001,
+      "loss": 1.3882,
+      "step": 13126
+    },
+    {
+      "epoch": 1.5250653499854778,
+      "grad_norm": 0.588337242603302,
+      "learning_rate": 0.0001,
+      "loss": 1.4547,
+      "step": 13127
+    },
+    {
+      "epoch": 1.5251815277374383,
+      "grad_norm": 0.6453753113746643,
+      "learning_rate": 0.0001,
+      "loss": 1.593,
+      "step": 13128
+    },
+    {
+      "epoch": 1.5252977054893988,
+      "grad_norm": 0.6236133575439453,
+      "learning_rate": 0.0001,
+      "loss": 1.3412,
+      "step": 13129
+    },
+    {
+      "epoch": 1.5254138832413593,
+      "grad_norm": 0.6044201850891113,
+      "learning_rate": 0.0001,
+      "loss": 1.4729,
+      "step": 13130
+    },
+    {
+      "epoch": 1.5255300609933198,
+      "grad_norm": 0.5731863975524902,
+      "learning_rate": 0.0001,
+      "loss": 1.3612,
+      "step": 13131
+    },
+    {
+      "epoch": 1.5256462387452803,
+      "grad_norm": 0.6380339860916138,
+      "learning_rate": 0.0001,
+      "loss": 1.5127,
+      "step": 13132
+    },
+    {
+      "epoch": 1.5257624164972408,
+      "grad_norm": 0.6082126498222351,
+      "learning_rate": 0.0001,
+      "loss": 1.5667,
+      "step": 13133
+    },
+    {
+      "epoch": 1.5258785942492012,
+      "grad_norm": 0.6267526745796204,
+      "learning_rate": 0.0001,
+      "loss": 1.6277,
+      "step": 13134
+    },
+    {
+      "epoch": 1.5259947720011617,
+      "grad_norm": 0.5912509560585022,
+      "learning_rate": 0.0001,
+      "loss": 1.2835,
+      "step": 13135
+    },
+    {
+      "epoch": 1.5261109497531222,
+      "grad_norm": 0.6406939029693604,
+      "learning_rate": 0.0001,
+      "loss": 1.6174,
+      "step": 13136
+    },
+    {
+      "epoch": 1.526227127505083,
+      "grad_norm": 0.6724519729614258,
+      "learning_rate": 0.0001,
+      "loss": 1.4591,
+      "step": 13137
+    },
+    {
+      "epoch": 1.5263433052570434,
+      "grad_norm": 0.6085923314094543,
+      "learning_rate": 0.0001,
+      "loss": 1.4962,
+      "step": 13138
+    },
+    {
+      "epoch": 1.526459483009004,
+      "grad_norm": 0.5566753149032593,
+      "learning_rate": 0.0001,
+      "loss": 1.1431,
+      "step": 13139
+    },
+    {
+      "epoch": 1.5265756607609644,
+      "grad_norm": 0.5759755969047546,
+      "learning_rate": 0.0001,
+      "loss": 1.3858,
+      "step": 13140
+    },
+    {
+      "epoch": 1.526691838512925,
+      "grad_norm": 0.5898522138595581,
+      "learning_rate": 0.0001,
+      "loss": 1.4425,
+      "step": 13141
+    },
+    {
+      "epoch": 1.5268080162648854,
+      "grad_norm": 0.614996075630188,
+      "learning_rate": 0.0001,
+      "loss": 1.5029,
+      "step": 13142
+    },
+    {
+      "epoch": 1.5269241940168459,
+      "grad_norm": 0.5954774022102356,
+      "learning_rate": 0.0001,
+      "loss": 1.4092,
+      "step": 13143
+    },
+    {
+      "epoch": 1.5270403717688064,
+      "grad_norm": 0.6323831081390381,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 13144
+    },
+    {
+      "epoch": 1.5271565495207668,
+      "grad_norm": 0.6423251628875732,
+      "learning_rate": 0.0001,
+      "loss": 1.7558,
+      "step": 13145
+    },
+    {
+      "epoch": 1.5272727272727273,
+      "grad_norm": 0.6106740236282349,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 13146
+    },
+    {
+      "epoch": 1.5273889050246878,
+      "grad_norm": 0.5833632349967957,
+      "learning_rate": 0.0001,
+      "loss": 1.4143,
+      "step": 13147
+    },
+    {
+      "epoch": 1.5275050827766483,
+      "grad_norm": 0.580480694770813,
+      "learning_rate": 0.0001,
+      "loss": 1.4606,
+      "step": 13148
+    },
+    {
+      "epoch": 1.5276212605286088,
+      "grad_norm": 0.6620073318481445,
+      "learning_rate": 0.0001,
+      "loss": 1.5736,
+      "step": 13149
+    },
+    {
+      "epoch": 1.5277374382805693,
+      "grad_norm": 0.6580905914306641,
+      "learning_rate": 0.0001,
+      "loss": 1.6311,
+      "step": 13150
+    },
+    {
+      "epoch": 1.5278536160325298,
+      "grad_norm": 0.5711208581924438,
+      "learning_rate": 0.0001,
+      "loss": 1.4917,
+      "step": 13151
+    },
+    {
+      "epoch": 1.5279697937844903,
+      "grad_norm": 0.5903010964393616,
+      "learning_rate": 0.0001,
+      "loss": 1.3864,
+      "step": 13152
+    },
+    {
+      "epoch": 1.5280859715364508,
+      "grad_norm": 0.5728796720504761,
+      "learning_rate": 0.0001,
+      "loss": 1.4043,
+      "step": 13153
+    },
+    {
+      "epoch": 1.5282021492884112,
+      "grad_norm": 0.591552197933197,
+      "learning_rate": 0.0001,
+      "loss": 1.5419,
+      "step": 13154
+    },
+    {
+      "epoch": 1.5283183270403717,
+      "grad_norm": 0.5879071950912476,
+      "learning_rate": 0.0001,
+      "loss": 1.2898,
+      "step": 13155
+    },
+    {
+      "epoch": 1.5284345047923322,
+      "grad_norm": 0.6146317720413208,
+      "learning_rate": 0.0001,
+      "loss": 1.4951,
+      "step": 13156
+    },
+    {
+      "epoch": 1.5285506825442927,
+      "grad_norm": 0.6422489285469055,
+      "learning_rate": 0.0001,
+      "loss": 1.5282,
+      "step": 13157
+    },
+    {
+      "epoch": 1.5286668602962532,
+      "grad_norm": 0.6233516335487366,
+      "learning_rate": 0.0001,
+      "loss": 1.5457,
+      "step": 13158
+    },
+    {
+      "epoch": 1.5287830380482137,
+      "grad_norm": 0.7108170986175537,
+      "learning_rate": 0.0001,
+      "loss": 1.7262,
+      "step": 13159
+    },
+    {
+      "epoch": 1.5288992158001742,
+      "grad_norm": 0.6578541398048401,
+      "learning_rate": 0.0001,
+      "loss": 1.6322,
+      "step": 13160
+    },
+    {
+      "epoch": 1.5290153935521347,
+      "grad_norm": 0.6104475259780884,
+      "learning_rate": 0.0001,
+      "loss": 1.461,
+      "step": 13161
+    },
+    {
+      "epoch": 1.5291315713040952,
+      "grad_norm": 0.6033722758293152,
+      "learning_rate": 0.0001,
+      "loss": 1.68,
+      "step": 13162
+    },
+    {
+      "epoch": 1.5292477490560556,
+      "grad_norm": 0.5982484817504883,
+      "learning_rate": 0.0001,
+      "loss": 1.5397,
+      "step": 13163
+    },
+    {
+      "epoch": 1.5293639268080161,
+      "grad_norm": 0.5699301958084106,
+      "learning_rate": 0.0001,
+      "loss": 1.3097,
+      "step": 13164
+    },
+    {
+      "epoch": 1.5294801045599766,
+      "grad_norm": 0.5736358165740967,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 13165
+    },
+    {
+      "epoch": 1.5295962823119371,
+      "grad_norm": 0.5809690952301025,
+      "learning_rate": 0.0001,
+      "loss": 1.2927,
+      "step": 13166
+    },
+    {
+      "epoch": 1.5297124600638976,
+      "grad_norm": 0.64310622215271,
+      "learning_rate": 0.0001,
+      "loss": 1.6676,
+      "step": 13167
+    },
+    {
+      "epoch": 1.5298286378158583,
+      "grad_norm": 0.6130411624908447,
+      "learning_rate": 0.0001,
+      "loss": 1.3156,
+      "step": 13168
+    },
+    {
+      "epoch": 1.5299448155678188,
+      "grad_norm": 0.611953854560852,
+      "learning_rate": 0.0001,
+      "loss": 1.6393,
+      "step": 13169
+    },
+    {
+      "epoch": 1.5300609933197793,
+      "grad_norm": 0.6218355894088745,
+      "learning_rate": 0.0001,
+      "loss": 1.3941,
+      "step": 13170
+    },
+    {
+      "epoch": 1.5301771710717398,
+      "grad_norm": 0.6101202964782715,
+      "learning_rate": 0.0001,
+      "loss": 1.405,
+      "step": 13171
+    },
+    {
+      "epoch": 1.5302933488237003,
+      "grad_norm": 0.6130668520927429,
+      "learning_rate": 0.0001,
+      "loss": 1.6009,
+      "step": 13172
+    },
+    {
+      "epoch": 1.5304095265756608,
+      "grad_norm": 0.5833741426467896,
+      "learning_rate": 0.0001,
+      "loss": 1.5686,
+      "step": 13173
+    },
+    {
+      "epoch": 1.5305257043276213,
+      "grad_norm": 0.5890129208564758,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 13174
+    },
+    {
+      "epoch": 1.5306418820795817,
+      "grad_norm": 0.5778335928916931,
+      "learning_rate": 0.0001,
+      "loss": 1.3429,
+      "step": 13175
+    },
+    {
+      "epoch": 1.5307580598315422,
+      "grad_norm": 0.6423119306564331,
+      "learning_rate": 0.0001,
+      "loss": 1.58,
+      "step": 13176
+    },
+    {
+      "epoch": 1.5308742375835027,
+      "grad_norm": 0.594713032245636,
+      "learning_rate": 0.0001,
+      "loss": 1.566,
+      "step": 13177
+    },
+    {
+      "epoch": 1.5309904153354632,
+      "grad_norm": 0.662578821182251,
+      "learning_rate": 0.0001,
+      "loss": 1.6117,
+      "step": 13178
+    },
+    {
+      "epoch": 1.531106593087424,
+      "grad_norm": 0.6052879095077515,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 13179
+    },
+    {
+      "epoch": 1.5312227708393844,
+      "grad_norm": 0.580970287322998,
+      "learning_rate": 0.0001,
+      "loss": 1.5377,
+      "step": 13180
+    },
+    {
+      "epoch": 1.531338948591345,
+      "grad_norm": 0.6133270859718323,
+      "learning_rate": 0.0001,
+      "loss": 1.5026,
+      "step": 13181
+    },
+    {
+      "epoch": 1.5314551263433054,
+      "grad_norm": 0.5756634473800659,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 13182
+    },
+    {
+      "epoch": 1.5315713040952659,
+      "grad_norm": 0.588685154914856,
+      "learning_rate": 0.0001,
+      "loss": 1.3341,
+      "step": 13183
+    },
+    {
+      "epoch": 1.5316874818472264,
+      "grad_norm": 0.6559744477272034,
+      "learning_rate": 0.0001,
+      "loss": 1.4987,
+      "step": 13184
+    },
+    {
+      "epoch": 1.5318036595991869,
+      "grad_norm": 0.6924941539764404,
+      "learning_rate": 0.0001,
+      "loss": 1.6968,
+      "step": 13185
+    },
+    {
+      "epoch": 1.5319198373511473,
+      "grad_norm": 0.5851516723632812,
+      "learning_rate": 0.0001,
+      "loss": 1.4279,
+      "step": 13186
+    },
+    {
+      "epoch": 1.5320360151031078,
+      "grad_norm": 0.6089817881584167,
+      "learning_rate": 0.0001,
+      "loss": 1.4143,
+      "step": 13187
+    },
+    {
+      "epoch": 1.5321521928550683,
+      "grad_norm": 0.6336711049079895,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 13188
+    },
+    {
+      "epoch": 1.5322683706070288,
+      "grad_norm": 0.5794046521186829,
+      "learning_rate": 0.0001,
+      "loss": 1.4202,
+      "step": 13189
+    },
+    {
+      "epoch": 1.5323845483589893,
+      "grad_norm": 0.614953875541687,
+      "learning_rate": 0.0001,
+      "loss": 1.4362,
+      "step": 13190
+    },
+    {
+      "epoch": 1.5325007261109498,
+      "grad_norm": 0.6357384324073792,
+      "learning_rate": 0.0001,
+      "loss": 1.3925,
+      "step": 13191
+    },
+    {
+      "epoch": 1.5326169038629103,
+      "grad_norm": 0.617169201374054,
+      "learning_rate": 0.0001,
+      "loss": 1.4264,
+      "step": 13192
+    },
+    {
+      "epoch": 1.5327330816148708,
+      "grad_norm": 0.5910585522651672,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 13193
+    },
+    {
+      "epoch": 1.5328492593668313,
+      "grad_norm": 0.5722532868385315,
+      "learning_rate": 0.0001,
+      "loss": 1.3988,
+      "step": 13194
+    },
+    {
+      "epoch": 1.5329654371187917,
+      "grad_norm": 0.5971508622169495,
+      "learning_rate": 0.0001,
+      "loss": 1.4986,
+      "step": 13195
+    },
+    {
+      "epoch": 1.5330816148707522,
+      "grad_norm": 0.6081166863441467,
+      "learning_rate": 0.0001,
+      "loss": 1.4698,
+      "step": 13196
+    },
+    {
+      "epoch": 1.5331977926227127,
+      "grad_norm": 0.6401707530021667,
+      "learning_rate": 0.0001,
+      "loss": 1.3302,
+      "step": 13197
+    },
+    {
+      "epoch": 1.5333139703746732,
+      "grad_norm": 0.6085423827171326,
+      "learning_rate": 0.0001,
+      "loss": 1.3921,
+      "step": 13198
+    },
+    {
+      "epoch": 1.5334301481266337,
+      "grad_norm": 0.6252439618110657,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 13199
+    },
+    {
+      "epoch": 1.5335463258785942,
+      "grad_norm": 0.6168279051780701,
+      "learning_rate": 0.0001,
+      "loss": 1.4045,
+      "step": 13200
+    },
+    {
+      "epoch": 1.5336625036305547,
+      "grad_norm": 0.5795454382896423,
+      "learning_rate": 0.0001,
+      "loss": 1.4236,
+      "step": 13201
+    },
+    {
+      "epoch": 1.5337786813825152,
+      "grad_norm": 0.6378520131111145,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 13202
+    },
+    {
+      "epoch": 1.5338948591344757,
+      "grad_norm": 0.616070568561554,
+      "learning_rate": 0.0001,
+      "loss": 1.5371,
+      "step": 13203
+    },
+    {
+      "epoch": 1.5340110368864361,
+      "grad_norm": 0.670217752456665,
+      "learning_rate": 0.0001,
+      "loss": 1.5229,
+      "step": 13204
+    },
+    {
+      "epoch": 1.5341272146383966,
+      "grad_norm": 0.6514032483100891,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 13205
+    },
+    {
+      "epoch": 1.5342433923903571,
+      "grad_norm": 0.5641196966171265,
+      "learning_rate": 0.0001,
+      "loss": 1.4328,
+      "step": 13206
+    },
+    {
+      "epoch": 1.5343595701423176,
+      "grad_norm": 0.5916898250579834,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 13207
+    },
+    {
+      "epoch": 1.534475747894278,
+      "grad_norm": 0.5729571580886841,
+      "learning_rate": 0.0001,
+      "loss": 1.3437,
+      "step": 13208
+    },
+    {
+      "epoch": 1.5345919256462386,
+      "grad_norm": 0.6877140402793884,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 13209
+    },
+    {
+      "epoch": 1.5347081033981993,
+      "grad_norm": 0.628453254699707,
+      "learning_rate": 0.0001,
+      "loss": 1.3626,
+      "step": 13210
+    },
+    {
+      "epoch": 1.5348242811501598,
+      "grad_norm": 0.5816517472267151,
+      "learning_rate": 0.0001,
+      "loss": 1.3589,
+      "step": 13211
+    },
+    {
+      "epoch": 1.5349404589021203,
+      "grad_norm": 0.5863345861434937,
+      "learning_rate": 0.0001,
+      "loss": 1.4668,
+      "step": 13212
+    },
+    {
+      "epoch": 1.5350566366540808,
+      "grad_norm": 0.5848140120506287,
+      "learning_rate": 0.0001,
+      "loss": 1.4771,
+      "step": 13213
+    },
+    {
+      "epoch": 1.5351728144060413,
+      "grad_norm": 0.6075370907783508,
+      "learning_rate": 0.0001,
+      "loss": 1.4385,
+      "step": 13214
+    },
+    {
+      "epoch": 1.5352889921580017,
+      "grad_norm": 0.6305372714996338,
+      "learning_rate": 0.0001,
+      "loss": 1.5449,
+      "step": 13215
+    },
+    {
+      "epoch": 1.5354051699099622,
+      "grad_norm": 0.5978954434394836,
+      "learning_rate": 0.0001,
+      "loss": 1.514,
+      "step": 13216
+    },
+    {
+      "epoch": 1.5355213476619227,
+      "grad_norm": 0.6035651564598083,
+      "learning_rate": 0.0001,
+      "loss": 1.5109,
+      "step": 13217
+    },
+    {
+      "epoch": 1.5356375254138832,
+      "grad_norm": 0.5722531080245972,
+      "learning_rate": 0.0001,
+      "loss": 1.5771,
+      "step": 13218
+    },
+    {
+      "epoch": 1.5357537031658437,
+      "grad_norm": 0.5960158705711365,
+      "learning_rate": 0.0001,
+      "loss": 1.4061,
+      "step": 13219
+    },
+    {
+      "epoch": 1.5358698809178042,
+      "grad_norm": 0.6274595260620117,
+      "learning_rate": 0.0001,
+      "loss": 1.5707,
+      "step": 13220
+    },
+    {
+      "epoch": 1.535986058669765,
+      "grad_norm": 0.6137160062789917,
+      "learning_rate": 0.0001,
+      "loss": 1.1475,
+      "step": 13221
+    },
+    {
+      "epoch": 1.5361022364217254,
+      "grad_norm": 0.6096351742744446,
+      "learning_rate": 0.0001,
+      "loss": 1.2318,
+      "step": 13222
+    },
+    {
+      "epoch": 1.5362184141736859,
+      "grad_norm": 0.642091691493988,
+      "learning_rate": 0.0001,
+      "loss": 1.5027,
+      "step": 13223
+    },
+    {
+      "epoch": 1.5363345919256464,
+      "grad_norm": 0.6222463250160217,
+      "learning_rate": 0.0001,
+      "loss": 1.4155,
+      "step": 13224
+    },
+    {
+      "epoch": 1.5364507696776069,
+      "grad_norm": 0.5922936201095581,
+      "learning_rate": 0.0001,
+      "loss": 1.3679,
+      "step": 13225
+    },
+    {
+      "epoch": 1.5365669474295673,
+      "grad_norm": 0.565569281578064,
+      "learning_rate": 0.0001,
+      "loss": 1.2068,
+      "step": 13226
+    },
+    {
+      "epoch": 1.5366831251815278,
+      "grad_norm": 0.6304237842559814,
+      "learning_rate": 0.0001,
+      "loss": 1.4687,
+      "step": 13227
+    },
+    {
+      "epoch": 1.5367993029334883,
+      "grad_norm": 0.6497960686683655,
+      "learning_rate": 0.0001,
+      "loss": 1.5075,
+      "step": 13228
+    },
+    {
+      "epoch": 1.5369154806854488,
+      "grad_norm": 0.6448734402656555,
+      "learning_rate": 0.0001,
+      "loss": 1.4774,
+      "step": 13229
+    },
+    {
+      "epoch": 1.5370316584374093,
+      "grad_norm": 0.5980844497680664,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 13230
+    },
+    {
+      "epoch": 1.5371478361893698,
+      "grad_norm": 0.5994159579277039,
+      "learning_rate": 0.0001,
+      "loss": 1.5558,
+      "step": 13231
+    },
+    {
+      "epoch": 1.5372640139413303,
+      "grad_norm": 0.58444744348526,
+      "learning_rate": 0.0001,
+      "loss": 1.4855,
+      "step": 13232
+    },
+    {
+      "epoch": 1.5373801916932908,
+      "grad_norm": 0.6289106607437134,
+      "learning_rate": 0.0001,
+      "loss": 1.5868,
+      "step": 13233
+    },
+    {
+      "epoch": 1.5374963694452513,
+      "grad_norm": 0.6375524997711182,
+      "learning_rate": 0.0001,
+      "loss": 1.6756,
+      "step": 13234
+    },
+    {
+      "epoch": 1.5376125471972117,
+      "grad_norm": 0.6511433720588684,
+      "learning_rate": 0.0001,
+      "loss": 1.4251,
+      "step": 13235
+    },
+    {
+      "epoch": 1.5377287249491722,
+      "grad_norm": 0.5981631875038147,
+      "learning_rate": 0.0001,
+      "loss": 1.4202,
+      "step": 13236
+    },
+    {
+      "epoch": 1.5378449027011327,
+      "grad_norm": 0.5719684958457947,
+      "learning_rate": 0.0001,
+      "loss": 1.4421,
+      "step": 13237
+    },
+    {
+      "epoch": 1.5379610804530932,
+      "grad_norm": 0.6303753852844238,
+      "learning_rate": 0.0001,
+      "loss": 1.2953,
+      "step": 13238
+    },
+    {
+      "epoch": 1.5380772582050537,
+      "grad_norm": 0.6014121174812317,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 13239
+    },
+    {
+      "epoch": 1.5381934359570142,
+      "grad_norm": 0.5993248820304871,
+      "learning_rate": 0.0001,
+      "loss": 1.4158,
+      "step": 13240
+    },
+    {
+      "epoch": 1.5383096137089747,
+      "grad_norm": 0.6195975542068481,
+      "learning_rate": 0.0001,
+      "loss": 1.4652,
+      "step": 13241
+    },
+    {
+      "epoch": 1.5384257914609352,
+      "grad_norm": 0.5984340906143188,
+      "learning_rate": 0.0001,
+      "loss": 1.5031,
+      "step": 13242
+    },
+    {
+      "epoch": 1.5385419692128957,
+      "grad_norm": 0.610115647315979,
+      "learning_rate": 0.0001,
+      "loss": 1.4412,
+      "step": 13243
+    },
+    {
+      "epoch": 1.5386581469648561,
+      "grad_norm": 0.6320290565490723,
+      "learning_rate": 0.0001,
+      "loss": 1.5107,
+      "step": 13244
+    },
+    {
+      "epoch": 1.5387743247168166,
+      "grad_norm": 0.6114384531974792,
+      "learning_rate": 0.0001,
+      "loss": 1.569,
+      "step": 13245
+    },
+    {
+      "epoch": 1.5388905024687771,
+      "grad_norm": 0.6696942448616028,
+      "learning_rate": 0.0001,
+      "loss": 1.5754,
+      "step": 13246
+    },
+    {
+      "epoch": 1.5390066802207376,
+      "grad_norm": 0.6278484463691711,
+      "learning_rate": 0.0001,
+      "loss": 1.3396,
+      "step": 13247
+    },
+    {
+      "epoch": 1.539122857972698,
+      "grad_norm": 0.5679048895835876,
+      "learning_rate": 0.0001,
+      "loss": 1.3174,
+      "step": 13248
+    },
+    {
+      "epoch": 1.5392390357246586,
+      "grad_norm": 0.6171903610229492,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 13249
+    },
+    {
+      "epoch": 1.539355213476619,
+      "grad_norm": 0.6222214698791504,
+      "learning_rate": 0.0001,
+      "loss": 1.439,
+      "step": 13250
+    },
+    {
+      "epoch": 1.5394713912285796,
+      "grad_norm": 0.5891657471656799,
+      "learning_rate": 0.0001,
+      "loss": 1.4478,
+      "step": 13251
+    },
+    {
+      "epoch": 1.5395875689805403,
+      "grad_norm": 0.6132618188858032,
+      "learning_rate": 0.0001,
+      "loss": 1.4395,
+      "step": 13252
+    },
+    {
+      "epoch": 1.5397037467325008,
+      "grad_norm": 0.5770408511161804,
+      "learning_rate": 0.0001,
+      "loss": 1.443,
+      "step": 13253
+    },
+    {
+      "epoch": 1.5398199244844613,
+      "grad_norm": 0.6150637269020081,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 13254
+    },
+    {
+      "epoch": 1.5399361022364217,
+      "grad_norm": 0.6168022155761719,
+      "learning_rate": 0.0001,
+      "loss": 1.3776,
+      "step": 13255
+    },
+    {
+      "epoch": 1.5400522799883822,
+      "grad_norm": 0.6376373767852783,
+      "learning_rate": 0.0001,
+      "loss": 1.6387,
+      "step": 13256
+    },
+    {
+      "epoch": 1.5401684577403427,
+      "grad_norm": 0.5829399824142456,
+      "learning_rate": 0.0001,
+      "loss": 1.2685,
+      "step": 13257
+    },
+    {
+      "epoch": 1.5402846354923032,
+      "grad_norm": 0.6367193460464478,
+      "learning_rate": 0.0001,
+      "loss": 1.5424,
+      "step": 13258
+    },
+    {
+      "epoch": 1.5404008132442637,
+      "grad_norm": 0.5960088968276978,
+      "learning_rate": 0.0001,
+      "loss": 1.3126,
+      "step": 13259
+    },
+    {
+      "epoch": 1.5405169909962242,
+      "grad_norm": 0.6211891770362854,
+      "learning_rate": 0.0001,
+      "loss": 1.5361,
+      "step": 13260
+    },
+    {
+      "epoch": 1.5406331687481847,
+      "grad_norm": 0.5986420512199402,
+      "learning_rate": 0.0001,
+      "loss": 1.5322,
+      "step": 13261
+    },
+    {
+      "epoch": 1.5407493465001452,
+      "grad_norm": 0.6235995888710022,
+      "learning_rate": 0.0001,
+      "loss": 1.4277,
+      "step": 13262
+    },
+    {
+      "epoch": 1.5408655242521059,
+      "grad_norm": 0.613019585609436,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 13263
+    },
+    {
+      "epoch": 1.5409817020040664,
+      "grad_norm": 0.630034327507019,
+      "learning_rate": 0.0001,
+      "loss": 1.6762,
+      "step": 13264
+    },
+    {
+      "epoch": 1.5410978797560269,
+      "grad_norm": 0.5918728709220886,
+      "learning_rate": 0.0001,
+      "loss": 1.4057,
+      "step": 13265
+    },
+    {
+      "epoch": 1.5412140575079873,
+      "grad_norm": 0.5738247036933899,
+      "learning_rate": 0.0001,
+      "loss": 1.4588,
+      "step": 13266
+    },
+    {
+      "epoch": 1.5413302352599478,
+      "grad_norm": 0.6175058484077454,
+      "learning_rate": 0.0001,
+      "loss": 1.457,
+      "step": 13267
+    },
+    {
+      "epoch": 1.5414464130119083,
+      "grad_norm": 0.5984748601913452,
+      "learning_rate": 0.0001,
+      "loss": 1.5045,
+      "step": 13268
+    },
+    {
+      "epoch": 1.5415625907638688,
+      "grad_norm": 0.6336756944656372,
+      "learning_rate": 0.0001,
+      "loss": 1.5041,
+      "step": 13269
+    },
+    {
+      "epoch": 1.5416787685158293,
+      "grad_norm": 0.6561615467071533,
+      "learning_rate": 0.0001,
+      "loss": 1.4334,
+      "step": 13270
+    },
+    {
+      "epoch": 1.5417949462677898,
+      "grad_norm": 0.643224835395813,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 13271
+    },
+    {
+      "epoch": 1.5419111240197503,
+      "grad_norm": 0.5880582928657532,
+      "learning_rate": 0.0001,
+      "loss": 1.3817,
+      "step": 13272
+    },
+    {
+      "epoch": 1.5420273017717108,
+      "grad_norm": 0.6055521368980408,
+      "learning_rate": 0.0001,
+      "loss": 1.3452,
+      "step": 13273
+    },
+    {
+      "epoch": 1.5421434795236713,
+      "grad_norm": 0.590501070022583,
+      "learning_rate": 0.0001,
+      "loss": 1.3997,
+      "step": 13274
+    },
+    {
+      "epoch": 1.5422596572756317,
+      "grad_norm": 0.5980383157730103,
+      "learning_rate": 0.0001,
+      "loss": 1.5425,
+      "step": 13275
+    },
+    {
+      "epoch": 1.5423758350275922,
+      "grad_norm": 0.5758526921272278,
+      "learning_rate": 0.0001,
+      "loss": 1.348,
+      "step": 13276
+    },
+    {
+      "epoch": 1.5424920127795527,
+      "grad_norm": 0.5835914611816406,
+      "learning_rate": 0.0001,
+      "loss": 1.2751,
+      "step": 13277
+    },
+    {
+      "epoch": 1.5426081905315132,
+      "grad_norm": 0.5609144568443298,
+      "learning_rate": 0.0001,
+      "loss": 1.2751,
+      "step": 13278
+    },
+    {
+      "epoch": 1.5427243682834737,
+      "grad_norm": 0.5979585647583008,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 13279
+    },
+    {
+      "epoch": 1.5428405460354342,
+      "grad_norm": 0.6307063698768616,
+      "learning_rate": 0.0001,
+      "loss": 1.4208,
+      "step": 13280
+    },
+    {
+      "epoch": 1.5429567237873947,
+      "grad_norm": 0.6205479502677917,
+      "learning_rate": 0.0001,
+      "loss": 1.4094,
+      "step": 13281
+    },
+    {
+      "epoch": 1.5430729015393552,
+      "grad_norm": 0.6695151925086975,
+      "learning_rate": 0.0001,
+      "loss": 1.5445,
+      "step": 13282
+    },
+    {
+      "epoch": 1.5431890792913157,
+      "grad_norm": 0.5992878675460815,
+      "learning_rate": 0.0001,
+      "loss": 1.3733,
+      "step": 13283
+    },
+    {
+      "epoch": 1.5433052570432761,
+      "grad_norm": 0.5989102125167847,
+      "learning_rate": 0.0001,
+      "loss": 1.3398,
+      "step": 13284
+    },
+    {
+      "epoch": 1.5434214347952366,
+      "grad_norm": 0.6177709698677063,
+      "learning_rate": 0.0001,
+      "loss": 1.5911,
+      "step": 13285
+    },
+    {
+      "epoch": 1.5435376125471971,
+      "grad_norm": 0.5789451599121094,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 13286
+    },
+    {
+      "epoch": 1.5436537902991576,
+      "grad_norm": 0.6461418271064758,
+      "learning_rate": 0.0001,
+      "loss": 1.4955,
+      "step": 13287
+    },
+    {
+      "epoch": 1.543769968051118,
+      "grad_norm": 0.5908834934234619,
+      "learning_rate": 0.0001,
+      "loss": 1.4299,
+      "step": 13288
+    },
+    {
+      "epoch": 1.5438861458030786,
+      "grad_norm": 0.6150990724563599,
+      "learning_rate": 0.0001,
+      "loss": 1.5879,
+      "step": 13289
+    },
+    {
+      "epoch": 1.544002323555039,
+      "grad_norm": 0.5878303050994873,
+      "learning_rate": 0.0001,
+      "loss": 1.2382,
+      "step": 13290
+    },
+    {
+      "epoch": 1.5441185013069996,
+      "grad_norm": 0.6185941100120544,
+      "learning_rate": 0.0001,
+      "loss": 1.542,
+      "step": 13291
+    },
+    {
+      "epoch": 1.54423467905896,
+      "grad_norm": 0.7063065767288208,
+      "learning_rate": 0.0001,
+      "loss": 1.6306,
+      "step": 13292
+    },
+    {
+      "epoch": 1.5443508568109205,
+      "grad_norm": 0.6151609420776367,
+      "learning_rate": 0.0001,
+      "loss": 1.4068,
+      "step": 13293
+    },
+    {
+      "epoch": 1.5444670345628813,
+      "grad_norm": 0.5908089280128479,
+      "learning_rate": 0.0001,
+      "loss": 1.5094,
+      "step": 13294
+    },
+    {
+      "epoch": 1.5445832123148417,
+      "grad_norm": 0.5625834465026855,
+      "learning_rate": 0.0001,
+      "loss": 1.3472,
+      "step": 13295
+    },
+    {
+      "epoch": 1.5446993900668022,
+      "grad_norm": 0.6102145314216614,
+      "learning_rate": 0.0001,
+      "loss": 1.4495,
+      "step": 13296
+    },
+    {
+      "epoch": 1.5448155678187627,
+      "grad_norm": 0.6253021359443665,
+      "learning_rate": 0.0001,
+      "loss": 1.4255,
+      "step": 13297
+    },
+    {
+      "epoch": 1.5449317455707232,
+      "grad_norm": 0.588062047958374,
+      "learning_rate": 0.0001,
+      "loss": 1.3395,
+      "step": 13298
+    },
+    {
+      "epoch": 1.5450479233226837,
+      "grad_norm": 0.6264815926551819,
+      "learning_rate": 0.0001,
+      "loss": 1.3836,
+      "step": 13299
+    },
+    {
+      "epoch": 1.5451641010746442,
+      "grad_norm": 0.6290935277938843,
+      "learning_rate": 0.0001,
+      "loss": 1.4261,
+      "step": 13300
+    },
+    {
+      "epoch": 1.5452802788266047,
+      "grad_norm": 0.6513729691505432,
+      "learning_rate": 0.0001,
+      "loss": 1.4928,
+      "step": 13301
+    },
+    {
+      "epoch": 1.5453964565785652,
+      "grad_norm": 0.6446203589439392,
+      "learning_rate": 0.0001,
+      "loss": 1.4939,
+      "step": 13302
+    },
+    {
+      "epoch": 1.5455126343305257,
+      "grad_norm": 0.6368615031242371,
+      "learning_rate": 0.0001,
+      "loss": 1.5529,
+      "step": 13303
+    },
+    {
+      "epoch": 1.5456288120824864,
+      "grad_norm": 0.5730301141738892,
+      "learning_rate": 0.0001,
+      "loss": 1.5144,
+      "step": 13304
+    },
+    {
+      "epoch": 1.5457449898344469,
+      "grad_norm": 0.5296486020088196,
+      "learning_rate": 0.0001,
+      "loss": 1.4107,
+      "step": 13305
+    },
+    {
+      "epoch": 1.5458611675864073,
+      "grad_norm": 0.6226763725280762,
+      "learning_rate": 0.0001,
+      "loss": 1.4552,
+      "step": 13306
+    },
+    {
+      "epoch": 1.5459773453383678,
+      "grad_norm": 0.6259723901748657,
+      "learning_rate": 0.0001,
+      "loss": 1.4037,
+      "step": 13307
+    },
+    {
+      "epoch": 1.5460935230903283,
+      "grad_norm": 0.5425748825073242,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 13308
+    },
+    {
+      "epoch": 1.5462097008422888,
+      "grad_norm": 0.6267858743667603,
+      "learning_rate": 0.0001,
+      "loss": 1.2792,
+      "step": 13309
+    },
+    {
+      "epoch": 1.5463258785942493,
+      "grad_norm": 0.5841253399848938,
+      "learning_rate": 0.0001,
+      "loss": 1.5139,
+      "step": 13310
+    },
+    {
+      "epoch": 1.5464420563462098,
+      "grad_norm": 0.660430908203125,
+      "learning_rate": 0.0001,
+      "loss": 1.5256,
+      "step": 13311
+    },
+    {
+      "epoch": 1.5465582340981703,
+      "grad_norm": 0.5816801190376282,
+      "learning_rate": 0.0001,
+      "loss": 1.3198,
+      "step": 13312
+    },
+    {
+      "epoch": 1.5466744118501308,
+      "grad_norm": 0.6387170553207397,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 13313
+    },
+    {
+      "epoch": 1.5467905896020913,
+      "grad_norm": 0.6249637603759766,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 13314
+    },
+    {
+      "epoch": 1.5469067673540517,
+      "grad_norm": 0.6024580597877502,
+      "learning_rate": 0.0001,
+      "loss": 1.4078,
+      "step": 13315
+    },
+    {
+      "epoch": 1.5470229451060122,
+      "grad_norm": 0.6384835243225098,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 13316
+    },
+    {
+      "epoch": 1.5471391228579727,
+      "grad_norm": 0.5695124864578247,
+      "learning_rate": 0.0001,
+      "loss": 1.364,
+      "step": 13317
+    },
+    {
+      "epoch": 1.5472553006099332,
+      "grad_norm": 0.6180372834205627,
+      "learning_rate": 0.0001,
+      "loss": 1.5373,
+      "step": 13318
+    },
+    {
+      "epoch": 1.5473714783618937,
+      "grad_norm": 0.5775349140167236,
+      "learning_rate": 0.0001,
+      "loss": 1.3818,
+      "step": 13319
+    },
+    {
+      "epoch": 1.5474876561138542,
+      "grad_norm": 0.6106045842170715,
+      "learning_rate": 0.0001,
+      "loss": 1.3383,
+      "step": 13320
+    },
+    {
+      "epoch": 1.5476038338658147,
+      "grad_norm": 0.6502373814582825,
+      "learning_rate": 0.0001,
+      "loss": 1.5179,
+      "step": 13321
+    },
+    {
+      "epoch": 1.5477200116177752,
+      "grad_norm": 0.6484719514846802,
+      "learning_rate": 0.0001,
+      "loss": 1.4393,
+      "step": 13322
+    },
+    {
+      "epoch": 1.5478361893697357,
+      "grad_norm": 0.6394400000572205,
+      "learning_rate": 0.0001,
+      "loss": 1.646,
+      "step": 13323
+    },
+    {
+      "epoch": 1.5479523671216961,
+      "grad_norm": 0.5985630750656128,
+      "learning_rate": 0.0001,
+      "loss": 1.4505,
+      "step": 13324
+    },
+    {
+      "epoch": 1.5480685448736566,
+      "grad_norm": 0.616834282875061,
+      "learning_rate": 0.0001,
+      "loss": 1.5475,
+      "step": 13325
+    },
+    {
+      "epoch": 1.5481847226256171,
+      "grad_norm": 0.6069583892822266,
+      "learning_rate": 0.0001,
+      "loss": 1.5224,
+      "step": 13326
+    },
+    {
+      "epoch": 1.5483009003775776,
+      "grad_norm": 0.6398360133171082,
+      "learning_rate": 0.0001,
+      "loss": 1.4967,
+      "step": 13327
+    },
+    {
+      "epoch": 1.548417078129538,
+      "grad_norm": 0.5988773107528687,
+      "learning_rate": 0.0001,
+      "loss": 1.3364,
+      "step": 13328
+    },
+    {
+      "epoch": 1.5485332558814986,
+      "grad_norm": 0.6219581365585327,
+      "learning_rate": 0.0001,
+      "loss": 1.3877,
+      "step": 13329
+    },
+    {
+      "epoch": 1.548649433633459,
+      "grad_norm": 0.5958396792411804,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 13330
+    },
+    {
+      "epoch": 1.5487656113854196,
+      "grad_norm": 0.5914923548698425,
+      "learning_rate": 0.0001,
+      "loss": 1.4747,
+      "step": 13331
+    },
+    {
+      "epoch": 1.54888178913738,
+      "grad_norm": 0.6485493779182434,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 13332
+    },
+    {
+      "epoch": 1.5489979668893405,
+      "grad_norm": 0.6255422830581665,
+      "learning_rate": 0.0001,
+      "loss": 1.4731,
+      "step": 13333
+    },
+    {
+      "epoch": 1.549114144641301,
+      "grad_norm": 0.6094285845756531,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 13334
+    },
+    {
+      "epoch": 1.5492303223932615,
+      "grad_norm": 0.5891302824020386,
+      "learning_rate": 0.0001,
+      "loss": 1.4598,
+      "step": 13335
+    },
+    {
+      "epoch": 1.5493465001452222,
+      "grad_norm": 0.6349563002586365,
+      "learning_rate": 0.0001,
+      "loss": 1.5865,
+      "step": 13336
+    },
+    {
+      "epoch": 1.5494626778971827,
+      "grad_norm": 0.6088860034942627,
+      "learning_rate": 0.0001,
+      "loss": 1.4814,
+      "step": 13337
+    },
+    {
+      "epoch": 1.5495788556491432,
+      "grad_norm": 0.7181675434112549,
+      "learning_rate": 0.0001,
+      "loss": 1.6944,
+      "step": 13338
+    },
+    {
+      "epoch": 1.5496950334011037,
+      "grad_norm": 0.6011210083961487,
+      "learning_rate": 0.0001,
+      "loss": 1.4959,
+      "step": 13339
+    },
+    {
+      "epoch": 1.5498112111530642,
+      "grad_norm": 0.6319872140884399,
+      "learning_rate": 0.0001,
+      "loss": 1.4118,
+      "step": 13340
+    },
+    {
+      "epoch": 1.5499273889050247,
+      "grad_norm": 0.6266433596611023,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 13341
+    },
+    {
+      "epoch": 1.5500435666569852,
+      "grad_norm": 0.6468966603279114,
+      "learning_rate": 0.0001,
+      "loss": 1.5942,
+      "step": 13342
+    },
+    {
+      "epoch": 1.5501597444089457,
+      "grad_norm": 0.5942563414573669,
+      "learning_rate": 0.0001,
+      "loss": 1.5652,
+      "step": 13343
+    },
+    {
+      "epoch": 1.5502759221609062,
+      "grad_norm": 0.5953789949417114,
+      "learning_rate": 0.0001,
+      "loss": 1.2761,
+      "step": 13344
+    },
+    {
+      "epoch": 1.5503920999128666,
+      "grad_norm": 0.5696388483047485,
+      "learning_rate": 0.0001,
+      "loss": 1.3045,
+      "step": 13345
+    },
+    {
+      "epoch": 1.5505082776648274,
+      "grad_norm": 0.5958232879638672,
+      "learning_rate": 0.0001,
+      "loss": 1.4762,
+      "step": 13346
+    },
+    {
+      "epoch": 1.5506244554167878,
+      "grad_norm": 0.6455214619636536,
+      "learning_rate": 0.0001,
+      "loss": 1.5647,
+      "step": 13347
+    },
+    {
+      "epoch": 1.5507406331687483,
+      "grad_norm": 0.7005032896995544,
+      "learning_rate": 0.0001,
+      "loss": 1.4845,
+      "step": 13348
+    },
+    {
+      "epoch": 1.5508568109207088,
+      "grad_norm": 0.6717994809150696,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 13349
+    },
+    {
+      "epoch": 1.5509729886726693,
+      "grad_norm": 0.5751558542251587,
+      "learning_rate": 0.0001,
+      "loss": 1.3095,
+      "step": 13350
+    },
+    {
+      "epoch": 1.5510891664246298,
+      "grad_norm": 0.5958260297775269,
+      "learning_rate": 0.0001,
+      "loss": 1.4298,
+      "step": 13351
+    },
+    {
+      "epoch": 1.5512053441765903,
+      "grad_norm": 0.5926149487495422,
+      "learning_rate": 0.0001,
+      "loss": 1.4678,
+      "step": 13352
+    },
+    {
+      "epoch": 1.5513215219285508,
+      "grad_norm": 0.621460497379303,
+      "learning_rate": 0.0001,
+      "loss": 1.3896,
+      "step": 13353
+    },
+    {
+      "epoch": 1.5514376996805113,
+      "grad_norm": 0.6164433360099792,
+      "learning_rate": 0.0001,
+      "loss": 1.5632,
+      "step": 13354
+    },
+    {
+      "epoch": 1.5515538774324718,
+      "grad_norm": 0.6684780120849609,
+      "learning_rate": 0.0001,
+      "loss": 1.6347,
+      "step": 13355
+    },
+    {
+      "epoch": 1.5516700551844322,
+      "grad_norm": 0.6255985498428345,
+      "learning_rate": 0.0001,
+      "loss": 1.3428,
+      "step": 13356
+    },
+    {
+      "epoch": 1.5517862329363927,
+      "grad_norm": 0.571209728717804,
+      "learning_rate": 0.0001,
+      "loss": 1.3975,
+      "step": 13357
+    },
+    {
+      "epoch": 1.5519024106883532,
+      "grad_norm": 0.5906187891960144,
+      "learning_rate": 0.0001,
+      "loss": 1.4001,
+      "step": 13358
+    },
+    {
+      "epoch": 1.5520185884403137,
+      "grad_norm": 0.6360796093940735,
+      "learning_rate": 0.0001,
+      "loss": 1.6595,
+      "step": 13359
+    },
+    {
+      "epoch": 1.5521347661922742,
+      "grad_norm": 0.5564336180686951,
+      "learning_rate": 0.0001,
+      "loss": 1.2847,
+      "step": 13360
+    },
+    {
+      "epoch": 1.5522509439442347,
+      "grad_norm": 0.6803907155990601,
+      "learning_rate": 0.0001,
+      "loss": 1.5776,
+      "step": 13361
+    },
+    {
+      "epoch": 1.5523671216961952,
+      "grad_norm": 0.6163176894187927,
+      "learning_rate": 0.0001,
+      "loss": 1.5737,
+      "step": 13362
+    },
+    {
+      "epoch": 1.5524832994481557,
+      "grad_norm": 0.5890237092971802,
+      "learning_rate": 0.0001,
+      "loss": 1.3819,
+      "step": 13363
+    },
+    {
+      "epoch": 1.5525994772001162,
+      "grad_norm": 0.6154002547264099,
+      "learning_rate": 0.0001,
+      "loss": 1.2708,
+      "step": 13364
+    },
+    {
+      "epoch": 1.5527156549520766,
+      "grad_norm": 0.6088365316390991,
+      "learning_rate": 0.0001,
+      "loss": 1.5037,
+      "step": 13365
+    },
+    {
+      "epoch": 1.5528318327040371,
+      "grad_norm": 0.5878404378890991,
+      "learning_rate": 0.0001,
+      "loss": 1.3299,
+      "step": 13366
+    },
+    {
+      "epoch": 1.5529480104559976,
+      "grad_norm": 0.6358885765075684,
+      "learning_rate": 0.0001,
+      "loss": 1.6242,
+      "step": 13367
+    },
+    {
+      "epoch": 1.553064188207958,
+      "grad_norm": 0.6308332681655884,
+      "learning_rate": 0.0001,
+      "loss": 1.4362,
+      "step": 13368
+    },
+    {
+      "epoch": 1.5531803659599186,
+      "grad_norm": 0.5902178287506104,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 13369
+    },
+    {
+      "epoch": 1.553296543711879,
+      "grad_norm": 0.6629870533943176,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 13370
+    },
+    {
+      "epoch": 1.5534127214638396,
+      "grad_norm": 0.6822353601455688,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 13371
+    },
+    {
+      "epoch": 1.5535288992158,
+      "grad_norm": 0.5877573490142822,
+      "learning_rate": 0.0001,
+      "loss": 1.2979,
+      "step": 13372
+    },
+    {
+      "epoch": 1.5536450769677606,
+      "grad_norm": 0.5882243514060974,
+      "learning_rate": 0.0001,
+      "loss": 1.5042,
+      "step": 13373
+    },
+    {
+      "epoch": 1.553761254719721,
+      "grad_norm": 0.6118847727775574,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 13374
+    },
+    {
+      "epoch": 1.5538774324716815,
+      "grad_norm": 0.6584449410438538,
+      "learning_rate": 0.0001,
+      "loss": 1.6281,
+      "step": 13375
+    },
+    {
+      "epoch": 1.553993610223642,
+      "grad_norm": 0.6288265585899353,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 13376
+    },
+    {
+      "epoch": 1.5541097879756025,
+      "grad_norm": 0.637592077255249,
+      "learning_rate": 0.0001,
+      "loss": 1.5458,
+      "step": 13377
+    },
+    {
+      "epoch": 1.5542259657275632,
+      "grad_norm": 0.642940104007721,
+      "learning_rate": 0.0001,
+      "loss": 1.4834,
+      "step": 13378
+    },
+    {
+      "epoch": 1.5543421434795237,
+      "grad_norm": 0.6421939730644226,
+      "learning_rate": 0.0001,
+      "loss": 1.4557,
+      "step": 13379
+    },
+    {
+      "epoch": 1.5544583212314842,
+      "grad_norm": 0.5893421173095703,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 13380
+    },
+    {
+      "epoch": 1.5545744989834447,
+      "grad_norm": 0.6120336055755615,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 13381
+    },
+    {
+      "epoch": 1.5546906767354052,
+      "grad_norm": 0.6369379162788391,
+      "learning_rate": 0.0001,
+      "loss": 1.4525,
+      "step": 13382
+    },
+    {
+      "epoch": 1.5548068544873657,
+      "grad_norm": 0.6311773657798767,
+      "learning_rate": 0.0001,
+      "loss": 1.4666,
+      "step": 13383
+    },
+    {
+      "epoch": 1.5549230322393262,
+      "grad_norm": 0.6140812039375305,
+      "learning_rate": 0.0001,
+      "loss": 1.4224,
+      "step": 13384
+    },
+    {
+      "epoch": 1.5550392099912866,
+      "grad_norm": 0.6031202673912048,
+      "learning_rate": 0.0001,
+      "loss": 1.4273,
+      "step": 13385
+    },
+    {
+      "epoch": 1.5551553877432471,
+      "grad_norm": 0.6218429803848267,
+      "learning_rate": 0.0001,
+      "loss": 1.5215,
+      "step": 13386
+    },
+    {
+      "epoch": 1.5552715654952076,
+      "grad_norm": 0.6086384654045105,
+      "learning_rate": 0.0001,
+      "loss": 1.6033,
+      "step": 13387
+    },
+    {
+      "epoch": 1.5553877432471683,
+      "grad_norm": 0.6052895188331604,
+      "learning_rate": 0.0001,
+      "loss": 1.4336,
+      "step": 13388
+    },
+    {
+      "epoch": 1.5555039209991288,
+      "grad_norm": 0.6349013447761536,
+      "learning_rate": 0.0001,
+      "loss": 1.5928,
+      "step": 13389
+    },
+    {
+      "epoch": 1.5556200987510893,
+      "grad_norm": 0.6184483766555786,
+      "learning_rate": 0.0001,
+      "loss": 1.61,
+      "step": 13390
+    },
+    {
+      "epoch": 1.5557362765030498,
+      "grad_norm": 0.601908266544342,
+      "learning_rate": 0.0001,
+      "loss": 1.5301,
+      "step": 13391
+    },
+    {
+      "epoch": 1.5558524542550103,
+      "grad_norm": 0.6823452115058899,
+      "learning_rate": 0.0001,
+      "loss": 1.7363,
+      "step": 13392
+    },
+    {
+      "epoch": 1.5559686320069708,
+      "grad_norm": 0.6129056811332703,
+      "learning_rate": 0.0001,
+      "loss": 1.41,
+      "step": 13393
+    },
+    {
+      "epoch": 1.5560848097589313,
+      "grad_norm": 0.5627692937850952,
+      "learning_rate": 0.0001,
+      "loss": 1.3706,
+      "step": 13394
+    },
+    {
+      "epoch": 1.5562009875108918,
+      "grad_norm": 0.5908641815185547,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 13395
+    },
+    {
+      "epoch": 1.5563171652628522,
+      "grad_norm": 0.6511195302009583,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 13396
+    },
+    {
+      "epoch": 1.5564333430148127,
+      "grad_norm": 0.619094729423523,
+      "learning_rate": 0.0001,
+      "loss": 1.4406,
+      "step": 13397
+    },
+    {
+      "epoch": 1.5565495207667732,
+      "grad_norm": 0.6720662117004395,
+      "learning_rate": 0.0001,
+      "loss": 1.5678,
+      "step": 13398
+    },
+    {
+      "epoch": 1.5566656985187337,
+      "grad_norm": 0.7280980944633484,
+      "learning_rate": 0.0001,
+      "loss": 1.6025,
+      "step": 13399
+    },
+    {
+      "epoch": 1.5567818762706942,
+      "grad_norm": 0.612970232963562,
+      "learning_rate": 0.0001,
+      "loss": 1.2731,
+      "step": 13400
+    },
+    {
+      "epoch": 1.5568980540226547,
+      "grad_norm": 0.6179816722869873,
+      "learning_rate": 0.0001,
+      "loss": 1.4997,
+      "step": 13401
+    },
+    {
+      "epoch": 1.5570142317746152,
+      "grad_norm": 0.6261540651321411,
+      "learning_rate": 0.0001,
+      "loss": 1.3036,
+      "step": 13402
+    },
+    {
+      "epoch": 1.5571304095265757,
+      "grad_norm": 0.6171663403511047,
+      "learning_rate": 0.0001,
+      "loss": 1.5464,
+      "step": 13403
+    },
+    {
+      "epoch": 1.5572465872785362,
+      "grad_norm": 0.5964109301567078,
+      "learning_rate": 0.0001,
+      "loss": 1.538,
+      "step": 13404
+    },
+    {
+      "epoch": 1.5573627650304966,
+      "grad_norm": 0.6069731712341309,
+      "learning_rate": 0.0001,
+      "loss": 1.3542,
+      "step": 13405
+    },
+    {
+      "epoch": 1.5574789427824571,
+      "grad_norm": 0.61017906665802,
+      "learning_rate": 0.0001,
+      "loss": 1.5481,
+      "step": 13406
+    },
+    {
+      "epoch": 1.5575951205344176,
+      "grad_norm": 0.6241382956504822,
+      "learning_rate": 0.0001,
+      "loss": 1.5264,
+      "step": 13407
+    },
+    {
+      "epoch": 1.557711298286378,
+      "grad_norm": 0.6097595691680908,
+      "learning_rate": 0.0001,
+      "loss": 1.4519,
+      "step": 13408
+    },
+    {
+      "epoch": 1.5578274760383386,
+      "grad_norm": 0.6461699604988098,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 13409
+    },
+    {
+      "epoch": 1.557943653790299,
+      "grad_norm": 0.6478632688522339,
+      "learning_rate": 0.0001,
+      "loss": 1.6612,
+      "step": 13410
+    },
+    {
+      "epoch": 1.5580598315422596,
+      "grad_norm": 0.5949918627738953,
+      "learning_rate": 0.0001,
+      "loss": 1.3795,
+      "step": 13411
+    },
+    {
+      "epoch": 1.55817600929422,
+      "grad_norm": 0.6056655645370483,
+      "learning_rate": 0.0001,
+      "loss": 1.4087,
+      "step": 13412
+    },
+    {
+      "epoch": 1.5582921870461806,
+      "grad_norm": 0.5610243082046509,
+      "learning_rate": 0.0001,
+      "loss": 1.3303,
+      "step": 13413
+    },
+    {
+      "epoch": 1.558408364798141,
+      "grad_norm": 0.5745115280151367,
+      "learning_rate": 0.0001,
+      "loss": 1.3136,
+      "step": 13414
+    },
+    {
+      "epoch": 1.5585245425501015,
+      "grad_norm": 0.587050199508667,
+      "learning_rate": 0.0001,
+      "loss": 1.4801,
+      "step": 13415
+    },
+    {
+      "epoch": 1.558640720302062,
+      "grad_norm": 0.6332523226737976,
+      "learning_rate": 0.0001,
+      "loss": 1.5378,
+      "step": 13416
+    },
+    {
+      "epoch": 1.5587568980540225,
+      "grad_norm": 0.6112111210823059,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 13417
+    },
+    {
+      "epoch": 1.558873075805983,
+      "grad_norm": 0.6478002667427063,
+      "learning_rate": 0.0001,
+      "loss": 1.6473,
+      "step": 13418
+    },
+    {
+      "epoch": 1.5589892535579435,
+      "grad_norm": 0.6234623789787292,
+      "learning_rate": 0.0001,
+      "loss": 1.5801,
+      "step": 13419
+    },
+    {
+      "epoch": 1.5591054313099042,
+      "grad_norm": 0.6207946538925171,
+      "learning_rate": 0.0001,
+      "loss": 1.4796,
+      "step": 13420
+    },
+    {
+      "epoch": 1.5592216090618647,
+      "grad_norm": 0.6182363629341125,
+      "learning_rate": 0.0001,
+      "loss": 1.4773,
+      "step": 13421
+    },
+    {
+      "epoch": 1.5593377868138252,
+      "grad_norm": 0.6172978281974792,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 13422
+    },
+    {
+      "epoch": 1.5594539645657857,
+      "grad_norm": 0.59696364402771,
+      "learning_rate": 0.0001,
+      "loss": 1.5049,
+      "step": 13423
+    },
+    {
+      "epoch": 1.5595701423177462,
+      "grad_norm": 0.58649742603302,
+      "learning_rate": 0.0001,
+      "loss": 1.4482,
+      "step": 13424
+    },
+    {
+      "epoch": 1.5596863200697066,
+      "grad_norm": 0.6304432153701782,
+      "learning_rate": 0.0001,
+      "loss": 1.6529,
+      "step": 13425
+    },
+    {
+      "epoch": 1.5598024978216671,
+      "grad_norm": 0.5944958329200745,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 13426
+    },
+    {
+      "epoch": 1.5599186755736276,
+      "grad_norm": 0.5643553137779236,
+      "learning_rate": 0.0001,
+      "loss": 1.4529,
+      "step": 13427
+    },
+    {
+      "epoch": 1.5600348533255881,
+      "grad_norm": 0.6088239550590515,
+      "learning_rate": 0.0001,
+      "loss": 1.4445,
+      "step": 13428
+    },
+    {
+      "epoch": 1.5601510310775486,
+      "grad_norm": 0.6248722672462463,
+      "learning_rate": 0.0001,
+      "loss": 1.4446,
+      "step": 13429
+    },
+    {
+      "epoch": 1.5602672088295093,
+      "grad_norm": 0.6209125518798828,
+      "learning_rate": 0.0001,
+      "loss": 1.4512,
+      "step": 13430
+    },
+    {
+      "epoch": 1.5603833865814698,
+      "grad_norm": 0.6119517683982849,
+      "learning_rate": 0.0001,
+      "loss": 1.498,
+      "step": 13431
+    },
+    {
+      "epoch": 1.5604995643334303,
+      "grad_norm": 0.6059420704841614,
+      "learning_rate": 0.0001,
+      "loss": 1.3094,
+      "step": 13432
+    },
+    {
+      "epoch": 1.5606157420853908,
+      "grad_norm": 0.635593593120575,
+      "learning_rate": 0.0001,
+      "loss": 1.5286,
+      "step": 13433
+    },
+    {
+      "epoch": 1.5607319198373513,
+      "grad_norm": 0.63407963514328,
+      "learning_rate": 0.0001,
+      "loss": 1.6207,
+      "step": 13434
+    },
+    {
+      "epoch": 1.5608480975893118,
+      "grad_norm": 0.5982295274734497,
+      "learning_rate": 0.0001,
+      "loss": 1.4909,
+      "step": 13435
+    },
+    {
+      "epoch": 1.5609642753412722,
+      "grad_norm": 0.6256076097488403,
+      "learning_rate": 0.0001,
+      "loss": 1.4428,
+      "step": 13436
+    },
+    {
+      "epoch": 1.5610804530932327,
+      "grad_norm": 0.600486159324646,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 13437
+    },
+    {
+      "epoch": 1.5611966308451932,
+      "grad_norm": 0.6342299580574036,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 13438
+    },
+    {
+      "epoch": 1.5613128085971537,
+      "grad_norm": 0.625291109085083,
+      "learning_rate": 0.0001,
+      "loss": 1.679,
+      "step": 13439
+    },
+    {
+      "epoch": 1.5614289863491142,
+      "grad_norm": 0.6260406970977783,
+      "learning_rate": 0.0001,
+      "loss": 1.5073,
+      "step": 13440
+    },
+    {
+      "epoch": 1.5615451641010747,
+      "grad_norm": 0.6225281953811646,
+      "learning_rate": 0.0001,
+      "loss": 1.4857,
+      "step": 13441
+    },
+    {
+      "epoch": 1.5616613418530352,
+      "grad_norm": 0.5835169553756714,
+      "learning_rate": 0.0001,
+      "loss": 1.5368,
+      "step": 13442
+    },
+    {
+      "epoch": 1.5617775196049957,
+      "grad_norm": 0.6170945167541504,
+      "learning_rate": 0.0001,
+      "loss": 1.53,
+      "step": 13443
+    },
+    {
+      "epoch": 1.5618936973569562,
+      "grad_norm": 0.594443142414093,
+      "learning_rate": 0.0001,
+      "loss": 1.3711,
+      "step": 13444
+    },
+    {
+      "epoch": 1.5620098751089166,
+      "grad_norm": 0.6158294081687927,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 13445
+    },
+    {
+      "epoch": 1.5621260528608771,
+      "grad_norm": 0.6171496510505676,
+      "learning_rate": 0.0001,
+      "loss": 1.5523,
+      "step": 13446
+    },
+    {
+      "epoch": 1.5622422306128376,
+      "grad_norm": 0.6249264478683472,
+      "learning_rate": 0.0001,
+      "loss": 1.4543,
+      "step": 13447
+    },
+    {
+      "epoch": 1.5623584083647981,
+      "grad_norm": 0.6244696974754333,
+      "learning_rate": 0.0001,
+      "loss": 1.4413,
+      "step": 13448
+    },
+    {
+      "epoch": 1.5624745861167586,
+      "grad_norm": 0.6428230404853821,
+      "learning_rate": 0.0001,
+      "loss": 1.5077,
+      "step": 13449
+    },
+    {
+      "epoch": 1.562590763868719,
+      "grad_norm": 0.6612055897712708,
+      "learning_rate": 0.0001,
+      "loss": 1.5691,
+      "step": 13450
+    },
+    {
+      "epoch": 1.5627069416206796,
+      "grad_norm": 0.6030257344245911,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 13451
+    },
+    {
+      "epoch": 1.56282311937264,
+      "grad_norm": 0.5865484476089478,
+      "learning_rate": 0.0001,
+      "loss": 1.2867,
+      "step": 13452
+    },
+    {
+      "epoch": 1.5629392971246006,
+      "grad_norm": 0.6694547533988953,
+      "learning_rate": 0.0001,
+      "loss": 1.5695,
+      "step": 13453
+    },
+    {
+      "epoch": 1.563055474876561,
+      "grad_norm": 0.6105194687843323,
+      "learning_rate": 0.0001,
+      "loss": 1.3006,
+      "step": 13454
+    },
+    {
+      "epoch": 1.5631716526285215,
+      "grad_norm": 0.5813074707984924,
+      "learning_rate": 0.0001,
+      "loss": 1.3504,
+      "step": 13455
+    },
+    {
+      "epoch": 1.563287830380482,
+      "grad_norm": 0.6537028551101685,
+      "learning_rate": 0.0001,
+      "loss": 1.3702,
+      "step": 13456
+    },
+    {
+      "epoch": 1.5634040081324425,
+      "grad_norm": 0.6405291557312012,
+      "learning_rate": 0.0001,
+      "loss": 1.3698,
+      "step": 13457
+    },
+    {
+      "epoch": 1.563520185884403,
+      "grad_norm": 0.644826352596283,
+      "learning_rate": 0.0001,
+      "loss": 1.5245,
+      "step": 13458
+    },
+    {
+      "epoch": 1.5636363636363635,
+      "grad_norm": 0.6287594437599182,
+      "learning_rate": 0.0001,
+      "loss": 1.4445,
+      "step": 13459
+    },
+    {
+      "epoch": 1.563752541388324,
+      "grad_norm": 0.6116755604743958,
+      "learning_rate": 0.0001,
+      "loss": 1.42,
+      "step": 13460
+    },
+    {
+      "epoch": 1.5638687191402847,
+      "grad_norm": 0.6694396734237671,
+      "learning_rate": 0.0001,
+      "loss": 1.528,
+      "step": 13461
+    },
+    {
+      "epoch": 1.5639848968922452,
+      "grad_norm": 0.6518230438232422,
+      "learning_rate": 0.0001,
+      "loss": 1.5947,
+      "step": 13462
+    },
+    {
+      "epoch": 1.5641010746442057,
+      "grad_norm": 0.5904227495193481,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 13463
+    },
+    {
+      "epoch": 1.5642172523961662,
+      "grad_norm": 0.6056810617446899,
+      "learning_rate": 0.0001,
+      "loss": 1.5085,
+      "step": 13464
+    },
+    {
+      "epoch": 1.5643334301481266,
+      "grad_norm": 0.606863260269165,
+      "learning_rate": 0.0001,
+      "loss": 1.4822,
+      "step": 13465
+    },
+    {
+      "epoch": 1.5644496079000871,
+      "grad_norm": 0.61651211977005,
+      "learning_rate": 0.0001,
+      "loss": 1.4381,
+      "step": 13466
+    },
+    {
+      "epoch": 1.5645657856520476,
+      "grad_norm": 0.6035972833633423,
+      "learning_rate": 0.0001,
+      "loss": 1.4649,
+      "step": 13467
+    },
+    {
+      "epoch": 1.5646819634040081,
+      "grad_norm": 0.6274685859680176,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 13468
+    },
+    {
+      "epoch": 1.5647981411559686,
+      "grad_norm": 0.6115575432777405,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 13469
+    },
+    {
+      "epoch": 1.564914318907929,
+      "grad_norm": 0.6543807983398438,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 13470
+    },
+    {
+      "epoch": 1.5650304966598896,
+      "grad_norm": 0.6324952244758606,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 13471
+    },
+    {
+      "epoch": 1.5651466744118503,
+      "grad_norm": 0.6065995693206787,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 13472
+    },
+    {
+      "epoch": 1.5652628521638108,
+      "grad_norm": 0.6735640168190002,
+      "learning_rate": 0.0001,
+      "loss": 1.4644,
+      "step": 13473
+    },
+    {
+      "epoch": 1.5653790299157713,
+      "grad_norm": 0.6395497918128967,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 13474
+    },
+    {
+      "epoch": 1.5654952076677318,
+      "grad_norm": 0.6460095643997192,
+      "learning_rate": 0.0001,
+      "loss": 1.5333,
+      "step": 13475
+    },
+    {
+      "epoch": 1.5656113854196922,
+      "grad_norm": 0.6567825078964233,
+      "learning_rate": 0.0001,
+      "loss": 1.6167,
+      "step": 13476
+    },
+    {
+      "epoch": 1.5657275631716527,
+      "grad_norm": 0.6820036768913269,
+      "learning_rate": 0.0001,
+      "loss": 1.6258,
+      "step": 13477
+    },
+    {
+      "epoch": 1.5658437409236132,
+      "grad_norm": 0.6071247458457947,
+      "learning_rate": 0.0001,
+      "loss": 1.3688,
+      "step": 13478
+    },
+    {
+      "epoch": 1.5659599186755737,
+      "grad_norm": 0.608846127986908,
+      "learning_rate": 0.0001,
+      "loss": 1.3601,
+      "step": 13479
+    },
+    {
+      "epoch": 1.5660760964275342,
+      "grad_norm": 0.7005351781845093,
+      "learning_rate": 0.0001,
+      "loss": 1.535,
+      "step": 13480
+    },
+    {
+      "epoch": 1.5661922741794947,
+      "grad_norm": 0.6527760624885559,
+      "learning_rate": 0.0001,
+      "loss": 1.5133,
+      "step": 13481
+    },
+    {
+      "epoch": 1.5663084519314552,
+      "grad_norm": 0.593977153301239,
+      "learning_rate": 0.0001,
+      "loss": 1.4874,
+      "step": 13482
+    },
+    {
+      "epoch": 1.5664246296834157,
+      "grad_norm": 0.6437960267066956,
+      "learning_rate": 0.0001,
+      "loss": 1.4985,
+      "step": 13483
+    },
+    {
+      "epoch": 1.5665408074353762,
+      "grad_norm": 0.608905553817749,
+      "learning_rate": 0.0001,
+      "loss": 1.4795,
+      "step": 13484
+    },
+    {
+      "epoch": 1.5666569851873366,
+      "grad_norm": 0.5913368463516235,
+      "learning_rate": 0.0001,
+      "loss": 1.3551,
+      "step": 13485
+    },
+    {
+      "epoch": 1.5667731629392971,
+      "grad_norm": 0.6186370849609375,
+      "learning_rate": 0.0001,
+      "loss": 1.4108,
+      "step": 13486
+    },
+    {
+      "epoch": 1.5668893406912576,
+      "grad_norm": 0.6638941168785095,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 13487
+    },
+    {
+      "epoch": 1.5670055184432181,
+      "grad_norm": 0.7029228806495667,
+      "learning_rate": 0.0001,
+      "loss": 1.6166,
+      "step": 13488
+    },
+    {
+      "epoch": 1.5671216961951786,
+      "grad_norm": 0.6374664902687073,
+      "learning_rate": 0.0001,
+      "loss": 1.5843,
+      "step": 13489
+    },
+    {
+      "epoch": 1.567237873947139,
+      "grad_norm": 0.6297853589057922,
+      "learning_rate": 0.0001,
+      "loss": 1.4802,
+      "step": 13490
+    },
+    {
+      "epoch": 1.5673540516990996,
+      "grad_norm": 0.6086907386779785,
+      "learning_rate": 0.0001,
+      "loss": 1.3929,
+      "step": 13491
+    },
+    {
+      "epoch": 1.56747022945106,
+      "grad_norm": 0.6632092595100403,
+      "learning_rate": 0.0001,
+      "loss": 1.5099,
+      "step": 13492
+    },
+    {
+      "epoch": 1.5675864072030206,
+      "grad_norm": 0.6230993270874023,
+      "learning_rate": 0.0001,
+      "loss": 1.497,
+      "step": 13493
+    },
+    {
+      "epoch": 1.567702584954981,
+      "grad_norm": 0.6114823818206787,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 13494
+    },
+    {
+      "epoch": 1.5678187627069415,
+      "grad_norm": 0.645427405834198,
+      "learning_rate": 0.0001,
+      "loss": 1.3659,
+      "step": 13495
+    },
+    {
+      "epoch": 1.567934940458902,
+      "grad_norm": 0.5928539633750916,
+      "learning_rate": 0.0001,
+      "loss": 1.3554,
+      "step": 13496
+    },
+    {
+      "epoch": 1.5680511182108625,
+      "grad_norm": 0.6389637589454651,
+      "learning_rate": 0.0001,
+      "loss": 1.3468,
+      "step": 13497
+    },
+    {
+      "epoch": 1.568167295962823,
+      "grad_norm": 0.6235938668251038,
+      "learning_rate": 0.0001,
+      "loss": 1.4206,
+      "step": 13498
+    },
+    {
+      "epoch": 1.5682834737147835,
+      "grad_norm": 0.6196651458740234,
+      "learning_rate": 0.0001,
+      "loss": 1.5044,
+      "step": 13499
+    },
+    {
+      "epoch": 1.568399651466744,
+      "grad_norm": 0.6908482313156128,
+      "learning_rate": 0.0001,
+      "loss": 1.6682,
+      "step": 13500
+    },
+    {
+      "epoch": 1.5685158292187045,
+      "grad_norm": 0.5792660713195801,
+      "learning_rate": 0.0001,
+      "loss": 1.4432,
+      "step": 13501
+    },
+    {
+      "epoch": 1.568632006970665,
+      "grad_norm": 0.6126808524131775,
+      "learning_rate": 0.0001,
+      "loss": 1.5117,
+      "step": 13502
+    },
+    {
+      "epoch": 1.5687481847226257,
+      "grad_norm": 0.6406491994857788,
+      "learning_rate": 0.0001,
+      "loss": 1.484,
+      "step": 13503
+    },
+    {
+      "epoch": 1.5688643624745862,
+      "grad_norm": 0.6253640055656433,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 13504
+    },
+    {
+      "epoch": 1.5689805402265466,
+      "grad_norm": 0.7741358280181885,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 13505
+    },
+    {
+      "epoch": 1.5690967179785071,
+      "grad_norm": 0.5788688659667969,
+      "learning_rate": 0.0001,
+      "loss": 1.397,
+      "step": 13506
+    },
+    {
+      "epoch": 1.5692128957304676,
+      "grad_norm": 0.6431933045387268,
+      "learning_rate": 0.0001,
+      "loss": 1.5296,
+      "step": 13507
+    },
+    {
+      "epoch": 1.5693290734824281,
+      "grad_norm": 0.6005233526229858,
+      "learning_rate": 0.0001,
+      "loss": 1.5396,
+      "step": 13508
+    },
+    {
+      "epoch": 1.5694452512343886,
+      "grad_norm": 0.6058092713356018,
+      "learning_rate": 0.0001,
+      "loss": 1.4367,
+      "step": 13509
+    },
+    {
+      "epoch": 1.569561428986349,
+      "grad_norm": 0.6385200023651123,
+      "learning_rate": 0.0001,
+      "loss": 1.6869,
+      "step": 13510
+    },
+    {
+      "epoch": 1.5696776067383096,
+      "grad_norm": 0.5717422366142273,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 13511
+    },
+    {
+      "epoch": 1.56979378449027,
+      "grad_norm": 0.5632378458976746,
+      "learning_rate": 0.0001,
+      "loss": 1.324,
+      "step": 13512
+    },
+    {
+      "epoch": 1.5699099622422306,
+      "grad_norm": 0.5892371535301208,
+      "learning_rate": 0.0001,
+      "loss": 1.4671,
+      "step": 13513
+    },
+    {
+      "epoch": 1.5700261399941913,
+      "grad_norm": 0.5963519811630249,
+      "learning_rate": 0.0001,
+      "loss": 1.4721,
+      "step": 13514
+    },
+    {
+      "epoch": 1.5701423177461518,
+      "grad_norm": 0.6520595550537109,
+      "learning_rate": 0.0001,
+      "loss": 1.6544,
+      "step": 13515
+    },
+    {
+      "epoch": 1.5702584954981123,
+      "grad_norm": 0.6093313097953796,
+      "learning_rate": 0.0001,
+      "loss": 1.3183,
+      "step": 13516
+    },
+    {
+      "epoch": 1.5703746732500727,
+      "grad_norm": 0.6415348052978516,
+      "learning_rate": 0.0001,
+      "loss": 1.5086,
+      "step": 13517
+    },
+    {
+      "epoch": 1.5704908510020332,
+      "grad_norm": 0.6013089418411255,
+      "learning_rate": 0.0001,
+      "loss": 1.4176,
+      "step": 13518
+    },
+    {
+      "epoch": 1.5706070287539937,
+      "grad_norm": 0.6108344197273254,
+      "learning_rate": 0.0001,
+      "loss": 1.4794,
+      "step": 13519
+    },
+    {
+      "epoch": 1.5707232065059542,
+      "grad_norm": 0.6556594371795654,
+      "learning_rate": 0.0001,
+      "loss": 1.6093,
+      "step": 13520
+    },
+    {
+      "epoch": 1.5708393842579147,
+      "grad_norm": 0.6048537492752075,
+      "learning_rate": 0.0001,
+      "loss": 1.4716,
+      "step": 13521
+    },
+    {
+      "epoch": 1.5709555620098752,
+      "grad_norm": 0.5792707800865173,
+      "learning_rate": 0.0001,
+      "loss": 1.3146,
+      "step": 13522
+    },
+    {
+      "epoch": 1.5710717397618357,
+      "grad_norm": 0.6412919163703918,
+      "learning_rate": 0.0001,
+      "loss": 1.4789,
+      "step": 13523
+    },
+    {
+      "epoch": 1.5711879175137962,
+      "grad_norm": 0.6530932188034058,
+      "learning_rate": 0.0001,
+      "loss": 1.4638,
+      "step": 13524
+    },
+    {
+      "epoch": 1.5713040952657567,
+      "grad_norm": 0.6409628391265869,
+      "learning_rate": 0.0001,
+      "loss": 1.4587,
+      "step": 13525
+    },
+    {
+      "epoch": 1.5714202730177171,
+      "grad_norm": 0.6507309675216675,
+      "learning_rate": 0.0001,
+      "loss": 1.6477,
+      "step": 13526
+    },
+    {
+      "epoch": 1.5715364507696776,
+      "grad_norm": 0.6563456654548645,
+      "learning_rate": 0.0001,
+      "loss": 1.3904,
+      "step": 13527
+    },
+    {
+      "epoch": 1.5716526285216381,
+      "grad_norm": 0.6711177825927734,
+      "learning_rate": 0.0001,
+      "loss": 1.4924,
+      "step": 13528
+    },
+    {
+      "epoch": 1.5717688062735986,
+      "grad_norm": 0.6010395884513855,
+      "learning_rate": 0.0001,
+      "loss": 1.4131,
+      "step": 13529
+    },
+    {
+      "epoch": 1.571884984025559,
+      "grad_norm": 0.609821617603302,
+      "learning_rate": 0.0001,
+      "loss": 1.3046,
+      "step": 13530
+    },
+    {
+      "epoch": 1.5720011617775196,
+      "grad_norm": 0.6318057179450989,
+      "learning_rate": 0.0001,
+      "loss": 1.4899,
+      "step": 13531
+    },
+    {
+      "epoch": 1.57211733952948,
+      "grad_norm": 0.6473124027252197,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 13532
+    },
+    {
+      "epoch": 1.5722335172814406,
+      "grad_norm": 0.6723019480705261,
+      "learning_rate": 0.0001,
+      "loss": 1.6904,
+      "step": 13533
+    },
+    {
+      "epoch": 1.572349695033401,
+      "grad_norm": 0.6439656019210815,
+      "learning_rate": 0.0001,
+      "loss": 1.4742,
+      "step": 13534
+    },
+    {
+      "epoch": 1.5724658727853615,
+      "grad_norm": 0.5960809588432312,
+      "learning_rate": 0.0001,
+      "loss": 1.4667,
+      "step": 13535
+    },
+    {
+      "epoch": 1.572582050537322,
+      "grad_norm": 0.6096289753913879,
+      "learning_rate": 0.0001,
+      "loss": 1.4864,
+      "step": 13536
+    },
+    {
+      "epoch": 1.5726982282892825,
+      "grad_norm": 0.5937155485153198,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 13537
+    },
+    {
+      "epoch": 1.572814406041243,
+      "grad_norm": 0.6248606443405151,
+      "learning_rate": 0.0001,
+      "loss": 1.4783,
+      "step": 13538
+    },
+    {
+      "epoch": 1.5729305837932035,
+      "grad_norm": 0.6048620343208313,
+      "learning_rate": 0.0001,
+      "loss": 1.5025,
+      "step": 13539
+    },
+    {
+      "epoch": 1.573046761545164,
+      "grad_norm": 0.5833286643028259,
+      "learning_rate": 0.0001,
+      "loss": 1.4685,
+      "step": 13540
+    },
+    {
+      "epoch": 1.5731629392971245,
+      "grad_norm": 0.6231339573860168,
+      "learning_rate": 0.0001,
+      "loss": 1.5243,
+      "step": 13541
+    },
+    {
+      "epoch": 1.573279117049085,
+      "grad_norm": 0.641281247138977,
+      "learning_rate": 0.0001,
+      "loss": 1.6371,
+      "step": 13542
+    },
+    {
+      "epoch": 1.5733952948010455,
+      "grad_norm": 0.6249711513519287,
+      "learning_rate": 0.0001,
+      "loss": 1.573,
+      "step": 13543
+    },
+    {
+      "epoch": 1.573511472553006,
+      "grad_norm": 0.5779685378074646,
+      "learning_rate": 0.0001,
+      "loss": 1.3116,
+      "step": 13544
+    },
+    {
+      "epoch": 1.5736276503049667,
+      "grad_norm": 0.6345327496528625,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 13545
+    },
+    {
+      "epoch": 1.5737438280569271,
+      "grad_norm": 0.6199609637260437,
+      "learning_rate": 0.0001,
+      "loss": 1.2796,
+      "step": 13546
+    },
+    {
+      "epoch": 1.5738600058088876,
+      "grad_norm": 0.5937657952308655,
+      "learning_rate": 0.0001,
+      "loss": 1.408,
+      "step": 13547
+    },
+    {
+      "epoch": 1.5739761835608481,
+      "grad_norm": 0.6459031701087952,
+      "learning_rate": 0.0001,
+      "loss": 1.4237,
+      "step": 13548
+    },
+    {
+      "epoch": 1.5740923613128086,
+      "grad_norm": 0.6150230169296265,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 13549
+    },
+    {
+      "epoch": 1.574208539064769,
+      "grad_norm": 0.6067149043083191,
+      "learning_rate": 0.0001,
+      "loss": 1.4518,
+      "step": 13550
+    },
+    {
+      "epoch": 1.5743247168167296,
+      "grad_norm": 0.5723419785499573,
+      "learning_rate": 0.0001,
+      "loss": 1.3533,
+      "step": 13551
+    },
+    {
+      "epoch": 1.57444089456869,
+      "grad_norm": 0.5707582831382751,
+      "learning_rate": 0.0001,
+      "loss": 1.255,
+      "step": 13552
+    },
+    {
+      "epoch": 1.5745570723206506,
+      "grad_norm": 0.6069430112838745,
+      "learning_rate": 0.0001,
+      "loss": 1.3843,
+      "step": 13553
+    },
+    {
+      "epoch": 1.574673250072611,
+      "grad_norm": 0.6458742022514343,
+      "learning_rate": 0.0001,
+      "loss": 1.4396,
+      "step": 13554
+    },
+    {
+      "epoch": 1.5747894278245715,
+      "grad_norm": 0.6966086030006409,
+      "learning_rate": 0.0001,
+      "loss": 1.5593,
+      "step": 13555
+    },
+    {
+      "epoch": 1.5749056055765323,
+      "grad_norm": 0.6121847033500671,
+      "learning_rate": 0.0001,
+      "loss": 1.5018,
+      "step": 13556
+    },
+    {
+      "epoch": 1.5750217833284927,
+      "grad_norm": 0.6447102427482605,
+      "learning_rate": 0.0001,
+      "loss": 1.5895,
+      "step": 13557
+    },
+    {
+      "epoch": 1.5751379610804532,
+      "grad_norm": 0.6316473484039307,
+      "learning_rate": 0.0001,
+      "loss": 1.4707,
+      "step": 13558
+    },
+    {
+      "epoch": 1.5752541388324137,
+      "grad_norm": 0.623142659664154,
+      "learning_rate": 0.0001,
+      "loss": 1.4751,
+      "step": 13559
+    },
+    {
+      "epoch": 1.5753703165843742,
+      "grad_norm": 0.5784343481063843,
+      "learning_rate": 0.0001,
+      "loss": 1.3394,
+      "step": 13560
+    },
+    {
+      "epoch": 1.5754864943363347,
+      "grad_norm": 0.5949280261993408,
+      "learning_rate": 0.0001,
+      "loss": 1.3922,
+      "step": 13561
+    },
+    {
+      "epoch": 1.5756026720882952,
+      "grad_norm": 0.5954093933105469,
+      "learning_rate": 0.0001,
+      "loss": 1.519,
+      "step": 13562
+    },
+    {
+      "epoch": 1.5757188498402557,
+      "grad_norm": 0.6469232439994812,
+      "learning_rate": 0.0001,
+      "loss": 1.6181,
+      "step": 13563
+    },
+    {
+      "epoch": 1.5758350275922162,
+      "grad_norm": 0.5921434164047241,
+      "learning_rate": 0.0001,
+      "loss": 1.302,
+      "step": 13564
+    },
+    {
+      "epoch": 1.5759512053441767,
+      "grad_norm": 0.5979151725769043,
+      "learning_rate": 0.0001,
+      "loss": 1.4764,
+      "step": 13565
+    },
+    {
+      "epoch": 1.5760673830961371,
+      "grad_norm": 0.5948922634124756,
+      "learning_rate": 0.0001,
+      "loss": 1.367,
+      "step": 13566
+    },
+    {
+      "epoch": 1.5761835608480976,
+      "grad_norm": 0.6355971097946167,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 13567
+    },
+    {
+      "epoch": 1.5762997386000581,
+      "grad_norm": 0.654316782951355,
+      "learning_rate": 0.0001,
+      "loss": 1.5292,
+      "step": 13568
+    },
+    {
+      "epoch": 1.5764159163520186,
+      "grad_norm": 0.6081578731536865,
+      "learning_rate": 0.0001,
+      "loss": 1.4942,
+      "step": 13569
+    },
+    {
+      "epoch": 1.576532094103979,
+      "grad_norm": 0.6357074975967407,
+      "learning_rate": 0.0001,
+      "loss": 1.3794,
+      "step": 13570
+    },
+    {
+      "epoch": 1.5766482718559396,
+      "grad_norm": 0.6234276294708252,
+      "learning_rate": 0.0001,
+      "loss": 1.4295,
+      "step": 13571
+    },
+    {
+      "epoch": 1.5767644496079,
+      "grad_norm": 0.6558229327201843,
+      "learning_rate": 0.0001,
+      "loss": 1.5862,
+      "step": 13572
+    },
+    {
+      "epoch": 1.5768806273598606,
+      "grad_norm": 0.6135905981063843,
+      "learning_rate": 0.0001,
+      "loss": 1.5521,
+      "step": 13573
+    },
+    {
+      "epoch": 1.576996805111821,
+      "grad_norm": 0.5907683372497559,
+      "learning_rate": 0.0001,
+      "loss": 1.3218,
+      "step": 13574
+    },
+    {
+      "epoch": 1.5771129828637815,
+      "grad_norm": 0.5935102701187134,
+      "learning_rate": 0.0001,
+      "loss": 1.4558,
+      "step": 13575
+    },
+    {
+      "epoch": 1.577229160615742,
+      "grad_norm": 0.5960171222686768,
+      "learning_rate": 0.0001,
+      "loss": 1.3446,
+      "step": 13576
+    },
+    {
+      "epoch": 1.5773453383677025,
+      "grad_norm": 0.617599606513977,
+      "learning_rate": 0.0001,
+      "loss": 1.4829,
+      "step": 13577
+    },
+    {
+      "epoch": 1.577461516119663,
+      "grad_norm": 0.611723005771637,
+      "learning_rate": 0.0001,
+      "loss": 1.5532,
+      "step": 13578
+    },
+    {
+      "epoch": 1.5775776938716235,
+      "grad_norm": 0.6049669981002808,
+      "learning_rate": 0.0001,
+      "loss": 1.3626,
+      "step": 13579
+    },
+    {
+      "epoch": 1.577693871623584,
+      "grad_norm": 0.6668212413787842,
+      "learning_rate": 0.0001,
+      "loss": 1.5093,
+      "step": 13580
+    },
+    {
+      "epoch": 1.5778100493755445,
+      "grad_norm": 0.5572206377983093,
+      "learning_rate": 0.0001,
+      "loss": 1.1698,
+      "step": 13581
+    },
+    {
+      "epoch": 1.577926227127505,
+      "grad_norm": 0.6364170908927917,
+      "learning_rate": 0.0001,
+      "loss": 1.3022,
+      "step": 13582
+    },
+    {
+      "epoch": 1.5780424048794655,
+      "grad_norm": 0.6507368087768555,
+      "learning_rate": 0.0001,
+      "loss": 1.5794,
+      "step": 13583
+    },
+    {
+      "epoch": 1.578158582631426,
+      "grad_norm": 0.6161065697669983,
+      "learning_rate": 0.0001,
+      "loss": 1.54,
+      "step": 13584
+    },
+    {
+      "epoch": 1.5782747603833864,
+      "grad_norm": 0.6118278503417969,
+      "learning_rate": 0.0001,
+      "loss": 1.5709,
+      "step": 13585
+    },
+    {
+      "epoch": 1.578390938135347,
+      "grad_norm": 0.6194911599159241,
+      "learning_rate": 0.0001,
+      "loss": 1.5626,
+      "step": 13586
+    },
+    {
+      "epoch": 1.5785071158873076,
+      "grad_norm": 0.6314056515693665,
+      "learning_rate": 0.0001,
+      "loss": 1.4473,
+      "step": 13587
+    },
+    {
+      "epoch": 1.5786232936392681,
+      "grad_norm": 0.6135434508323669,
+      "learning_rate": 0.0001,
+      "loss": 1.3764,
+      "step": 13588
+    },
+    {
+      "epoch": 1.5787394713912286,
+      "grad_norm": 0.6144410371780396,
+      "learning_rate": 0.0001,
+      "loss": 1.5922,
+      "step": 13589
+    },
+    {
+      "epoch": 1.578855649143189,
+      "grad_norm": 0.6232594847679138,
+      "learning_rate": 0.0001,
+      "loss": 1.5534,
+      "step": 13590
+    },
+    {
+      "epoch": 1.5789718268951496,
+      "grad_norm": 0.6463111639022827,
+      "learning_rate": 0.0001,
+      "loss": 1.5343,
+      "step": 13591
+    },
+    {
+      "epoch": 1.57908800464711,
+      "grad_norm": 0.6009104251861572,
+      "learning_rate": 0.0001,
+      "loss": 1.4659,
+      "step": 13592
+    },
+    {
+      "epoch": 1.5792041823990706,
+      "grad_norm": 0.6553391218185425,
+      "learning_rate": 0.0001,
+      "loss": 1.5536,
+      "step": 13593
+    },
+    {
+      "epoch": 1.579320360151031,
+      "grad_norm": 0.5963552594184875,
+      "learning_rate": 0.0001,
+      "loss": 1.221,
+      "step": 13594
+    },
+    {
+      "epoch": 1.5794365379029915,
+      "grad_norm": 0.5866657495498657,
+      "learning_rate": 0.0001,
+      "loss": 1.3527,
+      "step": 13595
+    },
+    {
+      "epoch": 1.579552715654952,
+      "grad_norm": 0.6316760182380676,
+      "learning_rate": 0.0001,
+      "loss": 1.3974,
+      "step": 13596
+    },
+    {
+      "epoch": 1.5796688934069125,
+      "grad_norm": 0.5814557671546936,
+      "learning_rate": 0.0001,
+      "loss": 1.257,
+      "step": 13597
+    },
+    {
+      "epoch": 1.5797850711588732,
+      "grad_norm": 0.5807773470878601,
+      "learning_rate": 0.0001,
+      "loss": 1.29,
+      "step": 13598
+    },
+    {
+      "epoch": 1.5799012489108337,
+      "grad_norm": 0.6384397745132446,
+      "learning_rate": 0.0001,
+      "loss": 1.5735,
+      "step": 13599
+    },
+    {
+      "epoch": 1.5800174266627942,
+      "grad_norm": 0.5958534479141235,
+      "learning_rate": 0.0001,
+      "loss": 1.378,
+      "step": 13600
+    },
+    {
+      "epoch": 1.5801336044147547,
+      "grad_norm": 0.633190929889679,
+      "learning_rate": 0.0001,
+      "loss": 1.5265,
+      "step": 13601
+    },
+    {
+      "epoch": 1.5802497821667152,
+      "grad_norm": 0.6206573843955994,
+      "learning_rate": 0.0001,
+      "loss": 1.4828,
+      "step": 13602
+    },
+    {
+      "epoch": 1.5803659599186757,
+      "grad_norm": 0.6325139999389648,
+      "learning_rate": 0.0001,
+      "loss": 1.3249,
+      "step": 13603
+    },
+    {
+      "epoch": 1.5804821376706362,
+      "grad_norm": 0.618319571018219,
+      "learning_rate": 0.0001,
+      "loss": 1.6177,
+      "step": 13604
+    },
+    {
+      "epoch": 1.5805983154225967,
+      "grad_norm": 0.6615894436836243,
+      "learning_rate": 0.0001,
+      "loss": 1.4182,
+      "step": 13605
+    },
+    {
+      "epoch": 1.5807144931745571,
+      "grad_norm": 0.632692813873291,
+      "learning_rate": 0.0001,
+      "loss": 1.4902,
+      "step": 13606
+    },
+    {
+      "epoch": 1.5808306709265176,
+      "grad_norm": 0.6061463356018066,
+      "learning_rate": 0.0001,
+      "loss": 1.3473,
+      "step": 13607
+    },
+    {
+      "epoch": 1.5809468486784781,
+      "grad_norm": 0.6018291711807251,
+      "learning_rate": 0.0001,
+      "loss": 1.4807,
+      "step": 13608
+    },
+    {
+      "epoch": 1.5810630264304386,
+      "grad_norm": 0.6714172959327698,
+      "learning_rate": 0.0001,
+      "loss": 1.4028,
+      "step": 13609
+    },
+    {
+      "epoch": 1.581179204182399,
+      "grad_norm": 0.5730470418930054,
+      "learning_rate": 0.0001,
+      "loss": 1.3864,
+      "step": 13610
+    },
+    {
+      "epoch": 1.5812953819343596,
+      "grad_norm": 0.5907166004180908,
+      "learning_rate": 0.0001,
+      "loss": 1.4745,
+      "step": 13611
+    },
+    {
+      "epoch": 1.58141155968632,
+      "grad_norm": 0.6612714529037476,
+      "learning_rate": 0.0001,
+      "loss": 1.4357,
+      "step": 13612
+    },
+    {
+      "epoch": 1.5815277374382806,
+      "grad_norm": 0.6622298359870911,
+      "learning_rate": 0.0001,
+      "loss": 1.5119,
+      "step": 13613
+    },
+    {
+      "epoch": 1.581643915190241,
+      "grad_norm": 0.6330438852310181,
+      "learning_rate": 0.0001,
+      "loss": 1.5316,
+      "step": 13614
+    },
+    {
+      "epoch": 1.5817600929422015,
+      "grad_norm": 0.6070156693458557,
+      "learning_rate": 0.0001,
+      "loss": 1.4341,
+      "step": 13615
+    },
+    {
+      "epoch": 1.581876270694162,
+      "grad_norm": 0.6398180723190308,
+      "learning_rate": 0.0001,
+      "loss": 1.3261,
+      "step": 13616
+    },
+    {
+      "epoch": 1.5819924484461225,
+      "grad_norm": 0.6727312803268433,
+      "learning_rate": 0.0001,
+      "loss": 1.6721,
+      "step": 13617
+    },
+    {
+      "epoch": 1.582108626198083,
+      "grad_norm": 0.5788384079933167,
+      "learning_rate": 0.0001,
+      "loss": 1.3515,
+      "step": 13618
+    },
+    {
+      "epoch": 1.5822248039500435,
+      "grad_norm": 0.5994517803192139,
+      "learning_rate": 0.0001,
+      "loss": 1.5195,
+      "step": 13619
+    },
+    {
+      "epoch": 1.582340981702004,
+      "grad_norm": 0.661238431930542,
+      "learning_rate": 0.0001,
+      "loss": 1.5039,
+      "step": 13620
+    },
+    {
+      "epoch": 1.5824571594539645,
+      "grad_norm": 0.6708731055259705,
+      "learning_rate": 0.0001,
+      "loss": 1.423,
+      "step": 13621
+    },
+    {
+      "epoch": 1.582573337205925,
+      "grad_norm": 0.6710119843482971,
+      "learning_rate": 0.0001,
+      "loss": 1.6503,
+      "step": 13622
+    },
+    {
+      "epoch": 1.5826895149578855,
+      "grad_norm": 0.632266640663147,
+      "learning_rate": 0.0001,
+      "loss": 1.7291,
+      "step": 13623
+    },
+    {
+      "epoch": 1.582805692709846,
+      "grad_norm": 0.6530042290687561,
+      "learning_rate": 0.0001,
+      "loss": 1.4614,
+      "step": 13624
+    },
+    {
+      "epoch": 1.5829218704618064,
+      "grad_norm": 0.5772076845169067,
+      "learning_rate": 0.0001,
+      "loss": 1.4532,
+      "step": 13625
+    },
+    {
+      "epoch": 1.583038048213767,
+      "grad_norm": 0.6056697368621826,
+      "learning_rate": 0.0001,
+      "loss": 1.5996,
+      "step": 13626
+    },
+    {
+      "epoch": 1.5831542259657274,
+      "grad_norm": 0.6514211893081665,
+      "learning_rate": 0.0001,
+      "loss": 1.6922,
+      "step": 13627
+    },
+    {
+      "epoch": 1.583270403717688,
+      "grad_norm": 0.5848867297172546,
+      "learning_rate": 0.0001,
+      "loss": 1.2466,
+      "step": 13628
+    },
+    {
+      "epoch": 1.5833865814696486,
+      "grad_norm": 0.5798253417015076,
+      "learning_rate": 0.0001,
+      "loss": 1.334,
+      "step": 13629
+    },
+    {
+      "epoch": 1.583502759221609,
+      "grad_norm": 0.6590627431869507,
+      "learning_rate": 0.0001,
+      "loss": 1.6691,
+      "step": 13630
+    },
+    {
+      "epoch": 1.5836189369735696,
+      "grad_norm": 0.5774937272071838,
+      "learning_rate": 0.0001,
+      "loss": 1.4867,
+      "step": 13631
+    },
+    {
+      "epoch": 1.58373511472553,
+      "grad_norm": 0.6019846200942993,
+      "learning_rate": 0.0001,
+      "loss": 1.3748,
+      "step": 13632
+    },
+    {
+      "epoch": 1.5838512924774906,
+      "grad_norm": 0.581792414188385,
+      "learning_rate": 0.0001,
+      "loss": 1.3973,
+      "step": 13633
+    },
+    {
+      "epoch": 1.583967470229451,
+      "grad_norm": 0.6952236890792847,
+      "learning_rate": 0.0001,
+      "loss": 1.7609,
+      "step": 13634
+    },
+    {
+      "epoch": 1.5840836479814115,
+      "grad_norm": 0.5895935297012329,
+      "learning_rate": 0.0001,
+      "loss": 1.4348,
+      "step": 13635
+    },
+    {
+      "epoch": 1.584199825733372,
+      "grad_norm": 0.6527734994888306,
+      "learning_rate": 0.0001,
+      "loss": 1.4164,
+      "step": 13636
+    },
+    {
+      "epoch": 1.5843160034853325,
+      "grad_norm": 0.5872225761413574,
+      "learning_rate": 0.0001,
+      "loss": 1.4422,
+      "step": 13637
+    },
+    {
+      "epoch": 1.584432181237293,
+      "grad_norm": 0.5942081212997437,
+      "learning_rate": 0.0001,
+      "loss": 1.3844,
+      "step": 13638
+    },
+    {
+      "epoch": 1.5845483589892535,
+      "grad_norm": 0.5980642437934875,
+      "learning_rate": 0.0001,
+      "loss": 1.717,
+      "step": 13639
+    },
+    {
+      "epoch": 1.5846645367412142,
+      "grad_norm": 0.5560755729675293,
+      "learning_rate": 0.0001,
+      "loss": 1.2702,
+      "step": 13640
+    },
+    {
+      "epoch": 1.5847807144931747,
+      "grad_norm": 0.6350087523460388,
+      "learning_rate": 0.0001,
+      "loss": 1.555,
+      "step": 13641
+    },
+    {
+      "epoch": 1.5848968922451352,
+      "grad_norm": 0.6073043346405029,
+      "learning_rate": 0.0001,
+      "loss": 1.5905,
+      "step": 13642
+    },
+    {
+      "epoch": 1.5850130699970957,
+      "grad_norm": 0.6224149465560913,
+      "learning_rate": 0.0001,
+      "loss": 1.3986,
+      "step": 13643
+    },
+    {
+      "epoch": 1.5851292477490562,
+      "grad_norm": 0.6422401070594788,
+      "learning_rate": 0.0001,
+      "loss": 1.5798,
+      "step": 13644
+    },
+    {
+      "epoch": 1.5852454255010167,
+      "grad_norm": 0.59529048204422,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 13645
+    },
+    {
+      "epoch": 1.5853616032529771,
+      "grad_norm": 0.6929322481155396,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 13646
+    },
+    {
+      "epoch": 1.5854777810049376,
+      "grad_norm": 0.5858911871910095,
+      "learning_rate": 0.0001,
+      "loss": 1.4672,
+      "step": 13647
+    },
+    {
+      "epoch": 1.5855939587568981,
+      "grad_norm": 0.6204400658607483,
+      "learning_rate": 0.0001,
+      "loss": 1.3553,
+      "step": 13648
+    },
+    {
+      "epoch": 1.5857101365088586,
+      "grad_norm": 0.5746094584465027,
+      "learning_rate": 0.0001,
+      "loss": 1.2365,
+      "step": 13649
+    },
+    {
+      "epoch": 1.585826314260819,
+      "grad_norm": 0.6758047938346863,
+      "learning_rate": 0.0001,
+      "loss": 1.5019,
+      "step": 13650
+    },
+    {
+      "epoch": 1.5859424920127796,
+      "grad_norm": 0.6562144756317139,
+      "learning_rate": 0.0001,
+      "loss": 1.4617,
+      "step": 13651
+    },
+    {
+      "epoch": 1.58605866976474,
+      "grad_norm": 0.6426905989646912,
+      "learning_rate": 0.0001,
+      "loss": 1.4973,
+      "step": 13652
+    },
+    {
+      "epoch": 1.5861748475167006,
+      "grad_norm": 0.5699962973594666,
+      "learning_rate": 0.0001,
+      "loss": 1.3537,
+      "step": 13653
+    },
+    {
+      "epoch": 1.586291025268661,
+      "grad_norm": 0.5969005227088928,
+      "learning_rate": 0.0001,
+      "loss": 1.3264,
+      "step": 13654
+    },
+    {
+      "epoch": 1.5864072030206215,
+      "grad_norm": 0.6157596111297607,
+      "learning_rate": 0.0001,
+      "loss": 1.4763,
+      "step": 13655
+    },
+    {
+      "epoch": 1.586523380772582,
+      "grad_norm": 0.5986966490745544,
+      "learning_rate": 0.0001,
+      "loss": 1.4321,
+      "step": 13656
+    },
+    {
+      "epoch": 1.5866395585245425,
+      "grad_norm": 0.6195567846298218,
+      "learning_rate": 0.0001,
+      "loss": 1.5363,
+      "step": 13657
+    },
+    {
+      "epoch": 1.586755736276503,
+      "grad_norm": 0.6283451914787292,
+      "learning_rate": 0.0001,
+      "loss": 1.3878,
+      "step": 13658
+    },
+    {
+      "epoch": 1.5868719140284635,
+      "grad_norm": 0.6001632809638977,
+      "learning_rate": 0.0001,
+      "loss": 1.4915,
+      "step": 13659
+    },
+    {
+      "epoch": 1.586988091780424,
+      "grad_norm": 0.6226193308830261,
+      "learning_rate": 0.0001,
+      "loss": 1.5789,
+      "step": 13660
+    },
+    {
+      "epoch": 1.5871042695323845,
+      "grad_norm": 0.6030400991439819,
+      "learning_rate": 0.0001,
+      "loss": 1.4806,
+      "step": 13661
+    },
+    {
+      "epoch": 1.587220447284345,
+      "grad_norm": 0.5935747623443604,
+      "learning_rate": 0.0001,
+      "loss": 1.337,
+      "step": 13662
+    },
+    {
+      "epoch": 1.5873366250363055,
+      "grad_norm": 0.6464823484420776,
+      "learning_rate": 0.0001,
+      "loss": 1.612,
+      "step": 13663
+    },
+    {
+      "epoch": 1.587452802788266,
+      "grad_norm": 0.6798169612884521,
+      "learning_rate": 0.0001,
+      "loss": 1.277,
+      "step": 13664
+    },
+    {
+      "epoch": 1.5875689805402264,
+      "grad_norm": 0.5947486162185669,
+      "learning_rate": 0.0001,
+      "loss": 1.3965,
+      "step": 13665
+    },
+    {
+      "epoch": 1.587685158292187,
+      "grad_norm": 0.6533806324005127,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 13666
+    },
+    {
+      "epoch": 1.5878013360441474,
+      "grad_norm": 0.6578819155693054,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 13667
+    },
+    {
+      "epoch": 1.587917513796108,
+      "grad_norm": 0.6176968216896057,
+      "learning_rate": 0.0001,
+      "loss": 1.413,
+      "step": 13668
+    },
+    {
+      "epoch": 1.5880336915480684,
+      "grad_norm": 0.6410322785377502,
+      "learning_rate": 0.0001,
+      "loss": 1.546,
+      "step": 13669
+    },
+    {
+      "epoch": 1.5881498693000289,
+      "grad_norm": 0.6015816330909729,
+      "learning_rate": 0.0001,
+      "loss": 1.464,
+      "step": 13670
+    },
+    {
+      "epoch": 1.5882660470519896,
+      "grad_norm": 0.6452057957649231,
+      "learning_rate": 0.0001,
+      "loss": 1.5803,
+      "step": 13671
+    },
+    {
+      "epoch": 1.58838222480395,
+      "grad_norm": 0.6403473019599915,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 13672
+    },
+    {
+      "epoch": 1.5884984025559106,
+      "grad_norm": 0.5736103653907776,
+      "learning_rate": 0.0001,
+      "loss": 1.3283,
+      "step": 13673
+    },
+    {
+      "epoch": 1.588614580307871,
+      "grad_norm": 0.5940598845481873,
+      "learning_rate": 0.0001,
+      "loss": 1.588,
+      "step": 13674
+    },
+    {
+      "epoch": 1.5887307580598315,
+      "grad_norm": 0.6314423084259033,
+      "learning_rate": 0.0001,
+      "loss": 1.4163,
+      "step": 13675
+    },
+    {
+      "epoch": 1.588846935811792,
+      "grad_norm": 0.5856815576553345,
+      "learning_rate": 0.0001,
+      "loss": 1.4124,
+      "step": 13676
+    },
+    {
+      "epoch": 1.5889631135637525,
+      "grad_norm": 0.6191335916519165,
+      "learning_rate": 0.0001,
+      "loss": 1.5927,
+      "step": 13677
+    },
+    {
+      "epoch": 1.589079291315713,
+      "grad_norm": 0.5804166793823242,
+      "learning_rate": 0.0001,
+      "loss": 1.3609,
+      "step": 13678
+    },
+    {
+      "epoch": 1.5891954690676735,
+      "grad_norm": 0.6077446937561035,
+      "learning_rate": 0.0001,
+      "loss": 1.4106,
+      "step": 13679
+    },
+    {
+      "epoch": 1.589311646819634,
+      "grad_norm": 0.668630838394165,
+      "learning_rate": 0.0001,
+      "loss": 1.3604,
+      "step": 13680
+    },
+    {
+      "epoch": 1.5894278245715947,
+      "grad_norm": 0.6589301824569702,
+      "learning_rate": 0.0001,
+      "loss": 1.7169,
+      "step": 13681
+    },
+    {
+      "epoch": 1.5895440023235552,
+      "grad_norm": 0.62638920545578,
+      "learning_rate": 0.0001,
+      "loss": 1.5183,
+      "step": 13682
+    },
+    {
+      "epoch": 1.5896601800755157,
+      "grad_norm": 0.626444935798645,
+      "learning_rate": 0.0001,
+      "loss": 1.5418,
+      "step": 13683
+    },
+    {
+      "epoch": 1.5897763578274762,
+      "grad_norm": 0.6382818818092346,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 13684
+    },
+    {
+      "epoch": 1.5898925355794367,
+      "grad_norm": 0.6530130505561829,
+      "learning_rate": 0.0001,
+      "loss": 1.534,
+      "step": 13685
+    },
+    {
+      "epoch": 1.5900087133313971,
+      "grad_norm": 0.5917280316352844,
+      "learning_rate": 0.0001,
+      "loss": 1.2945,
+      "step": 13686
+    },
+    {
+      "epoch": 1.5901248910833576,
+      "grad_norm": 0.601091206073761,
+      "learning_rate": 0.0001,
+      "loss": 1.4798,
+      "step": 13687
+    },
+    {
+      "epoch": 1.5902410688353181,
+      "grad_norm": 0.6319622993469238,
+      "learning_rate": 0.0001,
+      "loss": 1.4709,
+      "step": 13688
+    },
+    {
+      "epoch": 1.5903572465872786,
+      "grad_norm": 0.5939752459526062,
+      "learning_rate": 0.0001,
+      "loss": 1.3826,
+      "step": 13689
+    },
+    {
+      "epoch": 1.590473424339239,
+      "grad_norm": 0.6311963200569153,
+      "learning_rate": 0.0001,
+      "loss": 1.3094,
+      "step": 13690
+    },
+    {
+      "epoch": 1.5905896020911996,
+      "grad_norm": 0.6844353079795837,
+      "learning_rate": 0.0001,
+      "loss": 1.6273,
+      "step": 13691
+    },
+    {
+      "epoch": 1.59070577984316,
+      "grad_norm": 0.5649846196174622,
+      "learning_rate": 0.0001,
+      "loss": 1.4596,
+      "step": 13692
+    },
+    {
+      "epoch": 1.5908219575951206,
+      "grad_norm": 0.6137500405311584,
+      "learning_rate": 0.0001,
+      "loss": 1.4631,
+      "step": 13693
+    },
+    {
+      "epoch": 1.590938135347081,
+      "grad_norm": 0.6590669751167297,
+      "learning_rate": 0.0001,
+      "loss": 1.5914,
+      "step": 13694
+    },
+    {
+      "epoch": 1.5910543130990416,
+      "grad_norm": 0.6377047896385193,
+      "learning_rate": 0.0001,
+      "loss": 1.5466,
+      "step": 13695
+    },
+    {
+      "epoch": 1.591170490851002,
+      "grad_norm": 0.5739707946777344,
+      "learning_rate": 0.0001,
+      "loss": 1.4131,
+      "step": 13696
+    },
+    {
+      "epoch": 1.5912866686029625,
+      "grad_norm": 0.5823712944984436,
+      "learning_rate": 0.0001,
+      "loss": 1.3342,
+      "step": 13697
+    },
+    {
+      "epoch": 1.591402846354923,
+      "grad_norm": 0.6388639807701111,
+      "learning_rate": 0.0001,
+      "loss": 1.6015,
+      "step": 13698
+    },
+    {
+      "epoch": 1.5915190241068835,
+      "grad_norm": 0.5976583361625671,
+      "learning_rate": 0.0001,
+      "loss": 1.4843,
+      "step": 13699
+    },
+    {
+      "epoch": 1.591635201858844,
+      "grad_norm": 0.5898172855377197,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 13700
+    },
+    {
+      "epoch": 1.5917513796108045,
+      "grad_norm": 0.6254405379295349,
+      "learning_rate": 0.0001,
+      "loss": 1.4318,
+      "step": 13701
+    },
+    {
+      "epoch": 1.591867557362765,
+      "grad_norm": 0.6013739109039307,
+      "learning_rate": 0.0001,
+      "loss": 1.6765,
+      "step": 13702
+    },
+    {
+      "epoch": 1.5919837351147255,
+      "grad_norm": 0.6028892993927002,
+      "learning_rate": 0.0001,
+      "loss": 1.5473,
+      "step": 13703
+    },
+    {
+      "epoch": 1.592099912866686,
+      "grad_norm": 0.5645247101783752,
+      "learning_rate": 0.0001,
+      "loss": 1.3645,
+      "step": 13704
+    },
+    {
+      "epoch": 1.5922160906186464,
+      "grad_norm": 0.5704538822174072,
+      "learning_rate": 0.0001,
+      "loss": 1.4181,
+      "step": 13705
+    },
+    {
+      "epoch": 1.592332268370607,
+      "grad_norm": 0.6046001315116882,
+      "learning_rate": 0.0001,
+      "loss": 1.4436,
+      "step": 13706
+    },
+    {
+      "epoch": 1.5924484461225674,
+      "grad_norm": 0.6235845685005188,
+      "learning_rate": 0.0001,
+      "loss": 1.4548,
+      "step": 13707
+    },
+    {
+      "epoch": 1.592564623874528,
+      "grad_norm": 0.6004418730735779,
+      "learning_rate": 0.0001,
+      "loss": 1.3711,
+      "step": 13708
+    },
+    {
+      "epoch": 1.5926808016264884,
+      "grad_norm": 0.6089489459991455,
+      "learning_rate": 0.0001,
+      "loss": 1.4301,
+      "step": 13709
+    },
+    {
+      "epoch": 1.5927969793784489,
+      "grad_norm": 0.5987474322319031,
+      "learning_rate": 0.0001,
+      "loss": 1.6283,
+      "step": 13710
+    },
+    {
+      "epoch": 1.5929131571304094,
+      "grad_norm": 0.6150535345077515,
+      "learning_rate": 0.0001,
+      "loss": 1.6291,
+      "step": 13711
+    },
+    {
+      "epoch": 1.5930293348823699,
+      "grad_norm": 0.6323658227920532,
+      "learning_rate": 0.0001,
+      "loss": 1.5579,
+      "step": 13712
+    },
+    {
+      "epoch": 1.5931455126343306,
+      "grad_norm": 0.6383227705955505,
+      "learning_rate": 0.0001,
+      "loss": 1.4719,
+      "step": 13713
+    },
+    {
+      "epoch": 1.593261690386291,
+      "grad_norm": 0.6205024719238281,
+      "learning_rate": 0.0001,
+      "loss": 1.5389,
+      "step": 13714
+    },
+    {
+      "epoch": 1.5933778681382516,
+      "grad_norm": 0.6311460733413696,
+      "learning_rate": 0.0001,
+      "loss": 1.5924,
+      "step": 13715
+    },
+    {
+      "epoch": 1.593494045890212,
+      "grad_norm": 0.650108277797699,
+      "learning_rate": 0.0001,
+      "loss": 1.3354,
+      "step": 13716
+    },
+    {
+      "epoch": 1.5936102236421725,
+      "grad_norm": 0.5857402682304382,
+      "learning_rate": 0.0001,
+      "loss": 1.3433,
+      "step": 13717
+    },
+    {
+      "epoch": 1.593726401394133,
+      "grad_norm": 0.5971586108207703,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 13718
+    },
+    {
+      "epoch": 1.5938425791460935,
+      "grad_norm": 0.6388098001480103,
+      "learning_rate": 0.0001,
+      "loss": 1.4883,
+      "step": 13719
+    },
+    {
+      "epoch": 1.593958756898054,
+      "grad_norm": 0.6415640115737915,
+      "learning_rate": 0.0001,
+      "loss": 1.4228,
+      "step": 13720
+    },
+    {
+      "epoch": 1.5940749346500145,
+      "grad_norm": 0.5820626616477966,
+      "learning_rate": 0.0001,
+      "loss": 1.4049,
+      "step": 13721
+    },
+    {
+      "epoch": 1.594191112401975,
+      "grad_norm": 0.569054126739502,
+      "learning_rate": 0.0001,
+      "loss": 1.3599,
+      "step": 13722
+    },
+    {
+      "epoch": 1.5943072901539357,
+      "grad_norm": 0.6303136348724365,
+      "learning_rate": 0.0001,
+      "loss": 1.4897,
+      "step": 13723
+    },
+    {
+      "epoch": 1.5944234679058962,
+      "grad_norm": 0.5948778986930847,
+      "learning_rate": 0.0001,
+      "loss": 1.5478,
+      "step": 13724
+    },
+    {
+      "epoch": 1.5945396456578567,
+      "grad_norm": 0.6140435338020325,
+      "learning_rate": 0.0001,
+      "loss": 1.4991,
+      "step": 13725
+    },
+    {
+      "epoch": 1.5946558234098172,
+      "grad_norm": 0.5883561372756958,
+      "learning_rate": 0.0001,
+      "loss": 1.5103,
+      "step": 13726
+    },
+    {
+      "epoch": 1.5947720011617776,
+      "grad_norm": 0.5896515846252441,
+      "learning_rate": 0.0001,
+      "loss": 1.4454,
+      "step": 13727
+    },
+    {
+      "epoch": 1.5948881789137381,
+      "grad_norm": 0.5697752833366394,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 13728
+    },
+    {
+      "epoch": 1.5950043566656986,
+      "grad_norm": 0.6240254640579224,
+      "learning_rate": 0.0001,
+      "loss": 1.525,
+      "step": 13729
+    },
+    {
+      "epoch": 1.595120534417659,
+      "grad_norm": 0.6201429963111877,
+      "learning_rate": 0.0001,
+      "loss": 1.5679,
+      "step": 13730
+    },
+    {
+      "epoch": 1.5952367121696196,
+      "grad_norm": 0.590234637260437,
+      "learning_rate": 0.0001,
+      "loss": 1.3998,
+      "step": 13731
+    },
+    {
+      "epoch": 1.59535288992158,
+      "grad_norm": 0.5969529747962952,
+      "learning_rate": 0.0001,
+      "loss": 1.3655,
+      "step": 13732
+    },
+    {
+      "epoch": 1.5954690676735406,
+      "grad_norm": 0.6133432984352112,
+      "learning_rate": 0.0001,
+      "loss": 1.1459,
+      "step": 13733
+    },
+    {
+      "epoch": 1.595585245425501,
+      "grad_norm": 0.6062743067741394,
+      "learning_rate": 0.0001,
+      "loss": 1.4189,
+      "step": 13734
+    },
+    {
+      "epoch": 1.5957014231774616,
+      "grad_norm": 0.658184826374054,
+      "learning_rate": 0.0001,
+      "loss": 1.6716,
+      "step": 13735
+    },
+    {
+      "epoch": 1.595817600929422,
+      "grad_norm": 0.599326491355896,
+      "learning_rate": 0.0001,
+      "loss": 1.3884,
+      "step": 13736
+    },
+    {
+      "epoch": 1.5959337786813825,
+      "grad_norm": 0.6448878049850464,
+      "learning_rate": 0.0001,
+      "loss": 1.4524,
+      "step": 13737
+    },
+    {
+      "epoch": 1.596049956433343,
+      "grad_norm": 0.6201961636543274,
+      "learning_rate": 0.0001,
+      "loss": 1.3097,
+      "step": 13738
+    },
+    {
+      "epoch": 1.5961661341853035,
+      "grad_norm": 0.6305376291275024,
+      "learning_rate": 0.0001,
+      "loss": 1.4291,
+      "step": 13739
+    },
+    {
+      "epoch": 1.596282311937264,
+      "grad_norm": 0.6596425175666809,
+      "learning_rate": 0.0001,
+      "loss": 1.557,
+      "step": 13740
+    },
+    {
+      "epoch": 1.5963984896892245,
+      "grad_norm": 0.6207994818687439,
+      "learning_rate": 0.0001,
+      "loss": 1.4566,
+      "step": 13741
+    },
+    {
+      "epoch": 1.596514667441185,
+      "grad_norm": 0.6085087060928345,
+      "learning_rate": 0.0001,
+      "loss": 1.5105,
+      "step": 13742
+    },
+    {
+      "epoch": 1.5966308451931455,
+      "grad_norm": 0.6169493794441223,
+      "learning_rate": 0.0001,
+      "loss": 1.5793,
+      "step": 13743
+    },
+    {
+      "epoch": 1.596747022945106,
+      "grad_norm": 0.6025657653808594,
+      "learning_rate": 0.0001,
+      "loss": 1.28,
+      "step": 13744
+    },
+    {
+      "epoch": 1.5968632006970664,
+      "grad_norm": 0.6102682948112488,
+      "learning_rate": 0.0001,
+      "loss": 1.6037,
+      "step": 13745
+    },
+    {
+      "epoch": 1.596979378449027,
+      "grad_norm": 0.5944964289665222,
+      "learning_rate": 0.0001,
+      "loss": 1.6677,
+      "step": 13746
+    },
+    {
+      "epoch": 1.5970955562009874,
+      "grad_norm": 0.6884152889251709,
+      "learning_rate": 0.0001,
+      "loss": 1.6509,
+      "step": 13747
+    },
+    {
+      "epoch": 1.597211733952948,
+      "grad_norm": 0.600307285785675,
+      "learning_rate": 0.0001,
+      "loss": 1.4579,
+      "step": 13748
+    },
+    {
+      "epoch": 1.5973279117049084,
+      "grad_norm": 0.5742710828781128,
+      "learning_rate": 0.0001,
+      "loss": 1.2788,
+      "step": 13749
+    },
+    {
+      "epoch": 1.5974440894568689,
+      "grad_norm": 0.5742681622505188,
+      "learning_rate": 0.0001,
+      "loss": 1.4084,
+      "step": 13750
+    },
+    {
+      "epoch": 1.5975602672088294,
+      "grad_norm": 0.5895428657531738,
+      "learning_rate": 0.0001,
+      "loss": 1.5114,
+      "step": 13751
+    },
+    {
+      "epoch": 1.5976764449607899,
+      "grad_norm": 0.6041052937507629,
+      "learning_rate": 0.0001,
+      "loss": 1.4984,
+      "step": 13752
+    },
+    {
+      "epoch": 1.5977926227127504,
+      "grad_norm": 0.6442006826400757,
+      "learning_rate": 0.0001,
+      "loss": 1.6977,
+      "step": 13753
+    },
+    {
+      "epoch": 1.5979088004647108,
+      "grad_norm": 0.5974199175834656,
+      "learning_rate": 0.0001,
+      "loss": 1.3967,
+      "step": 13754
+    },
+    {
+      "epoch": 1.5980249782166716,
+      "grad_norm": 0.6257326602935791,
+      "learning_rate": 0.0001,
+      "loss": 1.5538,
+      "step": 13755
+    },
+    {
+      "epoch": 1.598141155968632,
+      "grad_norm": 0.6209074258804321,
+      "learning_rate": 0.0001,
+      "loss": 1.4854,
+      "step": 13756
+    },
+    {
+      "epoch": 1.5982573337205925,
+      "grad_norm": 0.6362159252166748,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 13757
+    },
+    {
+      "epoch": 1.598373511472553,
+      "grad_norm": 0.611994743347168,
+      "learning_rate": 0.0001,
+      "loss": 1.4347,
+      "step": 13758
+    },
+    {
+      "epoch": 1.5984896892245135,
+      "grad_norm": 0.6070793867111206,
+      "learning_rate": 0.0001,
+      "loss": 1.3342,
+      "step": 13759
+    },
+    {
+      "epoch": 1.598605866976474,
+      "grad_norm": 0.6359820365905762,
+      "learning_rate": 0.0001,
+      "loss": 1.4627,
+      "step": 13760
+    },
+    {
+      "epoch": 1.5987220447284345,
+      "grad_norm": 0.6544227004051208,
+      "learning_rate": 0.0001,
+      "loss": 1.4379,
+      "step": 13761
+    },
+    {
+      "epoch": 1.598838222480395,
+      "grad_norm": 0.669750452041626,
+      "learning_rate": 0.0001,
+      "loss": 1.3736,
+      "step": 13762
+    },
+    {
+      "epoch": 1.5989544002323555,
+      "grad_norm": 0.6785888075828552,
+      "learning_rate": 0.0001,
+      "loss": 1.6779,
+      "step": 13763
+    },
+    {
+      "epoch": 1.599070577984316,
+      "grad_norm": 0.6282570958137512,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 13764
+    },
+    {
+      "epoch": 1.5991867557362767,
+      "grad_norm": 0.5843489170074463,
+      "learning_rate": 0.0001,
+      "loss": 1.4549,
+      "step": 13765
+    },
+    {
+      "epoch": 1.5993029334882372,
+      "grad_norm": 0.6315165758132935,
+      "learning_rate": 0.0001,
+      "loss": 1.4352,
+      "step": 13766
+    },
+    {
+      "epoch": 1.5994191112401976,
+      "grad_norm": 0.6136773824691772,
+      "learning_rate": 0.0001,
+      "loss": 1.5576,
+      "step": 13767
+    },
+    {
+      "epoch": 1.5995352889921581,
+      "grad_norm": 0.6138652563095093,
+      "learning_rate": 0.0001,
+      "loss": 1.5366,
+      "step": 13768
+    },
+    {
+      "epoch": 1.5996514667441186,
+      "grad_norm": 0.6372418999671936,
+      "learning_rate": 0.0001,
+      "loss": 1.6065,
+      "step": 13769
+    },
+    {
+      "epoch": 1.599767644496079,
+      "grad_norm": 0.6061596870422363,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 13770
+    },
+    {
+      "epoch": 1.5998838222480396,
+      "grad_norm": 0.6062629818916321,
+      "learning_rate": 0.0001,
+      "loss": 1.2875,
+      "step": 13771
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.6032814979553223,
+      "learning_rate": 0.0001,
+      "loss": 1.4574,
+      "step": 13772
+    },
+    {
+      "epoch": 1.6001161777519606,
+      "grad_norm": 0.6498891115188599,
+      "learning_rate": 0.0001,
+      "loss": 1.5472,
+      "step": 13773
+    },
+    {
+      "epoch": 1.600232355503921,
+      "grad_norm": 0.6103582978248596,
+      "learning_rate": 0.0001,
+      "loss": 1.4621,
+      "step": 13774
+    },
+    {
+      "epoch": 1.6003485332558816,
+      "grad_norm": 0.5915868282318115,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 13775
+    },
+    {
+      "epoch": 1.600464711007842,
+      "grad_norm": 0.5753716826438904,
+      "learning_rate": 0.0001,
+      "loss": 1.2423,
+      "step": 13776
+    },
+    {
+      "epoch": 1.6005808887598025,
+      "grad_norm": 0.5940613746643066,
+      "learning_rate": 0.0001,
+      "loss": 1.5206,
+      "step": 13777
+    },
+    {
+      "epoch": 1.600697066511763,
+      "grad_norm": 0.5621521472930908,
+      "learning_rate": 0.0001,
+      "loss": 1.3927,
+      "step": 13778
+    },
+    {
+      "epoch": 1.6008132442637235,
+      "grad_norm": 0.6004639267921448,
+      "learning_rate": 0.0001,
+      "loss": 1.3794,
+      "step": 13779
+    },
+    {
+      "epoch": 1.600929422015684,
+      "grad_norm": 0.6216353178024292,
+      "learning_rate": 0.0001,
+      "loss": 1.559,
+      "step": 13780
+    },
+    {
+      "epoch": 1.6010455997676445,
+      "grad_norm": 0.5929364562034607,
+      "learning_rate": 0.0001,
+      "loss": 1.5074,
+      "step": 13781
+    },
+    {
+      "epoch": 1.601161777519605,
+      "grad_norm": 0.6302357316017151,
+      "learning_rate": 0.0001,
+      "loss": 1.3182,
+      "step": 13782
+    },
+    {
+      "epoch": 1.6012779552715655,
+      "grad_norm": 0.6794602274894714,
+      "learning_rate": 0.0001,
+      "loss": 1.6703,
+      "step": 13783
+    },
+    {
+      "epoch": 1.601394133023526,
+      "grad_norm": 0.5782144069671631,
+      "learning_rate": 0.0001,
+      "loss": 1.36,
+      "step": 13784
+    },
+    {
+      "epoch": 1.6015103107754864,
+      "grad_norm": 0.659773588180542,
+      "learning_rate": 0.0001,
+      "loss": 1.4868,
+      "step": 13785
+    },
+    {
+      "epoch": 1.601626488527447,
+      "grad_norm": 0.6725043058395386,
+      "learning_rate": 0.0001,
+      "loss": 1.4797,
+      "step": 13786
+    },
+    {
+      "epoch": 1.6017426662794074,
+      "grad_norm": 0.6366990208625793,
+      "learning_rate": 0.0001,
+      "loss": 1.4486,
+      "step": 13787
+    },
+    {
+      "epoch": 1.601858844031368,
+      "grad_norm": 0.6560510993003845,
+      "learning_rate": 0.0001,
+      "loss": 1.4995,
+      "step": 13788
+    },
+    {
+      "epoch": 1.6019750217833284,
+      "grad_norm": 0.6620891690254211,
+      "learning_rate": 0.0001,
+      "loss": 1.3972,
+      "step": 13789
+    },
+    {
+      "epoch": 1.602091199535289,
+      "grad_norm": 0.6221482157707214,
+      "learning_rate": 0.0001,
+      "loss": 1.5887,
+      "step": 13790
+    },
+    {
+      "epoch": 1.6022073772872494,
+      "grad_norm": 0.6068333983421326,
+      "learning_rate": 0.0001,
+      "loss": 1.4633,
+      "step": 13791
+    },
+    {
+      "epoch": 1.6023235550392099,
+      "grad_norm": 0.6149522066116333,
+      "learning_rate": 0.0001,
+      "loss": 1.5298,
+      "step": 13792
+    },
+    {
+      "epoch": 1.6024397327911704,
+      "grad_norm": 0.6315502524375916,
+      "learning_rate": 0.0001,
+      "loss": 1.3288,
+      "step": 13793
+    },
+    {
+      "epoch": 1.6025559105431308,
+      "grad_norm": 0.6193204522132874,
+      "learning_rate": 0.0001,
+      "loss": 1.5412,
+      "step": 13794
+    },
+    {
+      "epoch": 1.6026720882950913,
+      "grad_norm": 0.6118963956832886,
+      "learning_rate": 0.0001,
+      "loss": 1.6071,
+      "step": 13795
+    },
+    {
+      "epoch": 1.6027882660470518,
+      "grad_norm": 0.5986089110374451,
+      "learning_rate": 0.0001,
+      "loss": 1.4186,
+      "step": 13796
+    },
+    {
+      "epoch": 1.6029044437990125,
+      "grad_norm": 0.5849127769470215,
+      "learning_rate": 0.0001,
+      "loss": 1.6019,
+      "step": 13797
+    },
+    {
+      "epoch": 1.603020621550973,
+      "grad_norm": 0.6488816738128662,
+      "learning_rate": 0.0001,
+      "loss": 1.592,
+      "step": 13798
+    },
+    {
+      "epoch": 1.6031367993029335,
+      "grad_norm": 0.5963447690010071,
+      "learning_rate": 0.0001,
+      "loss": 1.4233,
+      "step": 13799
+    },
+    {
+      "epoch": 1.603252977054894,
+      "grad_norm": 0.5969744920730591,
+      "learning_rate": 0.0001,
+      "loss": 1.4197,
+      "step": 13800
+    },
+    {
+      "epoch": 1.6033691548068545,
+      "grad_norm": 0.5776385068893433,
+      "learning_rate": 0.0001,
+      "loss": 1.4639,
+      "step": 13801
+    },
+    {
+      "epoch": 1.603485332558815,
+      "grad_norm": 0.568734347820282,
+      "learning_rate": 0.0001,
+      "loss": 1.384,
+      "step": 13802
+    },
+    {
+      "epoch": 1.6036015103107755,
+      "grad_norm": 0.6355451345443726,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 13803
+    },
+    {
+      "epoch": 1.603717688062736,
+      "grad_norm": 0.5923163294792175,
+      "learning_rate": 0.0001,
+      "loss": 1.4938,
+      "step": 13804
+    },
+    {
+      "epoch": 1.6038338658146964,
+      "grad_norm": 0.6059244275093079,
+      "learning_rate": 0.0001,
+      "loss": 1.4256,
+      "step": 13805
+    },
+    {
+      "epoch": 1.603950043566657,
+      "grad_norm": 0.5898007750511169,
+      "learning_rate": 0.0001,
+      "loss": 1.4724,
+      "step": 13806
+    },
+    {
+      "epoch": 1.6040662213186176,
+      "grad_norm": 0.5829443335533142,
+      "learning_rate": 0.0001,
+      "loss": 1.3609,
+      "step": 13807
+    },
+    {
+      "epoch": 1.6041823990705781,
+      "grad_norm": 0.6079580783843994,
+      "learning_rate": 0.0001,
+      "loss": 1.3347,
+      "step": 13808
+    },
+    {
+      "epoch": 1.6042985768225386,
+      "grad_norm": 0.6549359560012817,
+      "learning_rate": 0.0001,
+      "loss": 1.5499,
+      "step": 13809
+    },
+    {
+      "epoch": 1.6044147545744991,
+      "grad_norm": 0.5998852252960205,
+      "learning_rate": 0.0001,
+      "loss": 1.3346,
+      "step": 13810
+    },
+    {
+      "epoch": 1.6045309323264596,
+      "grad_norm": 0.6005247831344604,
+      "learning_rate": 0.0001,
+      "loss": 1.4291,
+      "step": 13811
+    },
+    {
+      "epoch": 1.60464711007842,
+      "grad_norm": 0.6408963799476624,
+      "learning_rate": 0.0001,
+      "loss": 1.5221,
+      "step": 13812
+    },
+    {
+      "epoch": 1.6047632878303806,
+      "grad_norm": 0.589667022228241,
+      "learning_rate": 0.0001,
+      "loss": 1.334,
+      "step": 13813
+    },
+    {
+      "epoch": 1.604879465582341,
+      "grad_norm": 0.5643407702445984,
+      "learning_rate": 0.0001,
+      "loss": 1.3689,
+      "step": 13814
+    },
+    {
+      "epoch": 1.6049956433343016,
+      "grad_norm": 0.659737229347229,
+      "learning_rate": 0.0001,
+      "loss": 1.4327,
+      "step": 13815
+    },
+    {
+      "epoch": 1.605111821086262,
+      "grad_norm": 0.6060417890548706,
+      "learning_rate": 0.0001,
+      "loss": 1.4205,
+      "step": 13816
+    },
+    {
+      "epoch": 1.6052279988382225,
+      "grad_norm": 0.6752500534057617,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 13817
+    },
+    {
+      "epoch": 1.605344176590183,
+      "grad_norm": 0.6170288324356079,
+      "learning_rate": 0.0001,
+      "loss": 1.2098,
+      "step": 13818
+    },
+    {
+      "epoch": 1.6054603543421435,
+      "grad_norm": 0.6077424883842468,
+      "learning_rate": 0.0001,
+      "loss": 1.5053,
+      "step": 13819
+    },
+    {
+      "epoch": 1.605576532094104,
+      "grad_norm": 0.5606176257133484,
+      "learning_rate": 0.0001,
+      "loss": 1.3151,
+      "step": 13820
+    },
+    {
+      "epoch": 1.6056927098460645,
+      "grad_norm": 0.6182757019996643,
+      "learning_rate": 0.0001,
+      "loss": 1.503,
+      "step": 13821
+    },
+    {
+      "epoch": 1.605808887598025,
+      "grad_norm": 0.6296382546424866,
+      "learning_rate": 0.0001,
+      "loss": 1.5476,
+      "step": 13822
+    },
+    {
+      "epoch": 1.6059250653499855,
+      "grad_norm": 0.6217150688171387,
+      "learning_rate": 0.0001,
+      "loss": 1.3125,
+      "step": 13823
+    },
+    {
+      "epoch": 1.606041243101946,
+      "grad_norm": 0.6524258852005005,
+      "learning_rate": 0.0001,
+      "loss": 1.5611,
+      "step": 13824
+    },
+    {
+      "epoch": 1.6061574208539064,
+      "grad_norm": 0.6134025454521179,
+      "learning_rate": 0.0001,
+      "loss": 1.4981,
+      "step": 13825
+    },
+    {
+      "epoch": 1.606273598605867,
+      "grad_norm": 0.6530399918556213,
+      "learning_rate": 0.0001,
+      "loss": 1.4917,
+      "step": 13826
+    },
+    {
+      "epoch": 1.6063897763578274,
+      "grad_norm": 0.5791024565696716,
+      "learning_rate": 0.0001,
+      "loss": 1.2731,
+      "step": 13827
+    },
+    {
+      "epoch": 1.606505954109788,
+      "grad_norm": 0.6186858415603638,
+      "learning_rate": 0.0001,
+      "loss": 1.3552,
+      "step": 13828
+    },
+    {
+      "epoch": 1.6066221318617484,
+      "grad_norm": 0.5937260985374451,
+      "learning_rate": 0.0001,
+      "loss": 1.3172,
+      "step": 13829
+    },
+    {
+      "epoch": 1.606738309613709,
+      "grad_norm": 0.6377929449081421,
+      "learning_rate": 0.0001,
+      "loss": 1.5637,
+      "step": 13830
+    },
+    {
+      "epoch": 1.6068544873656694,
+      "grad_norm": 0.5948269367218018,
+      "learning_rate": 0.0001,
+      "loss": 1.3598,
+      "step": 13831
+    },
+    {
+      "epoch": 1.6069706651176299,
+      "grad_norm": 0.5641445517539978,
+      "learning_rate": 0.0001,
+      "loss": 1.2045,
+      "step": 13832
+    },
+    {
+      "epoch": 1.6070868428695904,
+      "grad_norm": 0.638268232345581,
+      "learning_rate": 0.0001,
+      "loss": 1.4597,
+      "step": 13833
+    },
+    {
+      "epoch": 1.6072030206215508,
+      "grad_norm": 0.5887224674224854,
+      "learning_rate": 0.0001,
+      "loss": 1.4827,
+      "step": 13834
+    },
+    {
+      "epoch": 1.6073191983735113,
+      "grad_norm": 0.6423467397689819,
+      "learning_rate": 0.0001,
+      "loss": 1.7015,
+      "step": 13835
+    },
+    {
+      "epoch": 1.6074353761254718,
+      "grad_norm": 0.6366564631462097,
+      "learning_rate": 0.0001,
+      "loss": 1.6689,
+      "step": 13836
+    },
+    {
+      "epoch": 1.6075515538774323,
+      "grad_norm": 0.5624208450317383,
+      "learning_rate": 0.0001,
+      "loss": 1.3148,
+      "step": 13837
+    },
+    {
+      "epoch": 1.607667731629393,
+      "grad_norm": 0.5844243168830872,
+      "learning_rate": 0.0001,
+      "loss": 1.4741,
+      "step": 13838
+    },
+    {
+      "epoch": 1.6077839093813535,
+      "grad_norm": 0.6397031545639038,
+      "learning_rate": 0.0001,
+      "loss": 1.6976,
+      "step": 13839
+    },
+    {
+      "epoch": 1.607900087133314,
+      "grad_norm": 0.6005331873893738,
+      "learning_rate": 0.0001,
+      "loss": 1.3997,
+      "step": 13840
+    },
+    {
+      "epoch": 1.6080162648852745,
+      "grad_norm": 0.6438605189323425,
+      "learning_rate": 0.0001,
+      "loss": 1.6067,
+      "step": 13841
+    },
+    {
+      "epoch": 1.608132442637235,
+      "grad_norm": 0.5851279497146606,
+      "learning_rate": 0.0001,
+      "loss": 1.2065,
+      "step": 13842
+    },
+    {
+      "epoch": 1.6082486203891955,
+      "grad_norm": 0.567348301410675,
+      "learning_rate": 0.0001,
+      "loss": 1.3517,
+      "step": 13843
+    },
+    {
+      "epoch": 1.608364798141156,
+      "grad_norm": 0.6039325594902039,
+      "learning_rate": 0.0001,
+      "loss": 1.4571,
+      "step": 13844
+    },
+    {
+      "epoch": 1.6084809758931164,
+      "grad_norm": 0.5768235921859741,
+      "learning_rate": 0.0001,
+      "loss": 1.3531,
+      "step": 13845
+    },
+    {
+      "epoch": 1.608597153645077,
+      "grad_norm": 0.5796616077423096,
+      "learning_rate": 0.0001,
+      "loss": 1.3958,
+      "step": 13846
+    },
+    {
+      "epoch": 1.6087133313970374,
+      "grad_norm": 0.603740930557251,
+      "learning_rate": 0.0001,
+      "loss": 1.4034,
+      "step": 13847
+    },
+    {
+      "epoch": 1.608829509148998,
+      "grad_norm": 0.6364991068840027,
+      "learning_rate": 0.0001,
+      "loss": 1.5609,
+      "step": 13848
+    },
+    {
+      "epoch": 1.6089456869009586,
+      "grad_norm": 0.6351603269577026,
+      "learning_rate": 0.0001,
+      "loss": 1.564,
+      "step": 13849
+    },
+    {
+      "epoch": 1.6090618646529191,
+      "grad_norm": 0.6068862676620483,
+      "learning_rate": 0.0001,
+      "loss": 1.4097,
+      "step": 13850
+    },
+    {
+      "epoch": 1.6091780424048796,
+      "grad_norm": 0.642614483833313,
+      "learning_rate": 0.0001,
+      "loss": 1.4754,
+      "step": 13851
+    },
+    {
+      "epoch": 1.60929422015684,
+      "grad_norm": 0.6429895162582397,
+      "learning_rate": 0.0001,
+      "loss": 1.6018,
+      "step": 13852
+    },
+    {
+      "epoch": 1.6094103979088006,
+      "grad_norm": 0.5926845073699951,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 13853
+    },
+    {
+      "epoch": 1.609526575660761,
+      "grad_norm": 0.658845067024231,
+      "learning_rate": 0.0001,
+      "loss": 1.3906,
+      "step": 13854
+    },
+    {
+      "epoch": 1.6096427534127216,
+      "grad_norm": 0.5958040952682495,
+      "learning_rate": 0.0001,
+      "loss": 1.3695,
+      "step": 13855
+    },
+    {
+      "epoch": 1.609758931164682,
+      "grad_norm": 0.6442962288856506,
+      "learning_rate": 0.0001,
+      "loss": 1.6451,
+      "step": 13856
+    },
+    {
+      "epoch": 1.6098751089166425,
+      "grad_norm": 0.5802430510520935,
+      "learning_rate": 0.0001,
+      "loss": 1.4404,
+      "step": 13857
+    },
+    {
+      "epoch": 1.609991286668603,
+      "grad_norm": 0.6123525500297546,
+      "learning_rate": 0.0001,
+      "loss": 1.3439,
+      "step": 13858
+    },
+    {
+      "epoch": 1.6101074644205635,
+      "grad_norm": 0.6517568230628967,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 13859
+    },
+    {
+      "epoch": 1.610223642172524,
+      "grad_norm": 0.634818434715271,
+      "learning_rate": 0.0001,
+      "loss": 1.6394,
+      "step": 13860
+    },
+    {
+      "epoch": 1.6103398199244845,
+      "grad_norm": 0.5639699697494507,
+      "learning_rate": 0.0001,
+      "loss": 1.5309,
+      "step": 13861
+    },
+    {
+      "epoch": 1.610455997676445,
+      "grad_norm": 0.5527640581130981,
+      "learning_rate": 0.0001,
+      "loss": 1.39,
+      "step": 13862
+    },
+    {
+      "epoch": 1.6105721754284055,
+      "grad_norm": 0.6287981271743774,
+      "learning_rate": 0.0001,
+      "loss": 1.4817,
+      "step": 13863
+    },
+    {
+      "epoch": 1.610688353180366,
+      "grad_norm": 0.5936501622200012,
+      "learning_rate": 0.0001,
+      "loss": 1.5035,
+      "step": 13864
+    },
+    {
+      "epoch": 1.6108045309323265,
+      "grad_norm": 0.6397016644477844,
+      "learning_rate": 0.0001,
+      "loss": 1.3037,
+      "step": 13865
+    },
+    {
+      "epoch": 1.610920708684287,
+      "grad_norm": 0.6394860744476318,
+      "learning_rate": 0.0001,
+      "loss": 1.4609,
+      "step": 13866
+    },
+    {
+      "epoch": 1.6110368864362474,
+      "grad_norm": 0.6565226316452026,
+      "learning_rate": 0.0001,
+      "loss": 1.5441,
+      "step": 13867
+    },
+    {
+      "epoch": 1.611153064188208,
+      "grad_norm": 0.6310607194900513,
+      "learning_rate": 0.0001,
+      "loss": 1.4312,
+      "step": 13868
+    },
+    {
+      "epoch": 1.6112692419401684,
+      "grad_norm": 0.6256000995635986,
+      "learning_rate": 0.0001,
+      "loss": 1.6485,
+      "step": 13869
+    },
+    {
+      "epoch": 1.611385419692129,
+      "grad_norm": 0.6283453106880188,
+      "learning_rate": 0.0001,
+      "loss": 1.4648,
+      "step": 13870
+    },
+    {
+      "epoch": 1.6115015974440894,
+      "grad_norm": 0.6144470572471619,
+      "learning_rate": 0.0001,
+      "loss": 1.3494,
+      "step": 13871
+    },
+    {
+      "epoch": 1.6116177751960499,
+      "grad_norm": 0.6374574303627014,
+      "learning_rate": 0.0001,
+      "loss": 1.6428,
+      "step": 13872
+    },
+    {
+      "epoch": 1.6117339529480104,
+      "grad_norm": 0.6335989236831665,
+      "learning_rate": 0.0001,
+      "loss": 1.5362,
+      "step": 13873
+    },
+    {
+      "epoch": 1.6118501306999709,
+      "grad_norm": 0.6003391146659851,
+      "learning_rate": 0.0001,
+      "loss": 1.4554,
+      "step": 13874
+    },
+    {
+      "epoch": 1.6119663084519313,
+      "grad_norm": 0.6658709645271301,
+      "learning_rate": 0.0001,
+      "loss": 1.6036,
+      "step": 13875
+    },
+    {
+      "epoch": 1.6120824862038918,
+      "grad_norm": 0.6419057250022888,
+      "learning_rate": 0.0001,
+      "loss": 1.4394,
+      "step": 13876
+    },
+    {
+      "epoch": 1.6121986639558523,
+      "grad_norm": 0.737707793712616,
+      "learning_rate": 0.0001,
+      "loss": 1.3447,
+      "step": 13877
+    },
+    {
+      "epoch": 1.6123148417078128,
+      "grad_norm": 0.6373167634010315,
+      "learning_rate": 0.0001,
+      "loss": 1.5011,
+      "step": 13878
+    },
+    {
+      "epoch": 1.6124310194597733,
+      "grad_norm": 0.6646495461463928,
+      "learning_rate": 0.0001,
+      "loss": 1.6556,
+      "step": 13879
+    },
+    {
+      "epoch": 1.612547197211734,
+      "grad_norm": 0.6386552453041077,
+      "learning_rate": 0.0001,
+      "loss": 1.5203,
+      "step": 13880
+    },
+    {
+      "epoch": 1.6126633749636945,
+      "grad_norm": 0.5788254141807556,
+      "learning_rate": 0.0001,
+      "loss": 1.5247,
+      "step": 13881
+    },
+    {
+      "epoch": 1.612779552715655,
+      "grad_norm": 0.6018937826156616,
+      "learning_rate": 0.0001,
+      "loss": 1.4407,
+      "step": 13882
+    },
+    {
+      "epoch": 1.6128957304676155,
+      "grad_norm": 0.6025969386100769,
+      "learning_rate": 0.0001,
+      "loss": 1.3986,
+      "step": 13883
+    },
+    {
+      "epoch": 1.613011908219576,
+      "grad_norm": 0.5800029635429382,
+      "learning_rate": 0.0001,
+      "loss": 1.4766,
+      "step": 13884
+    },
+    {
+      "epoch": 1.6131280859715365,
+      "grad_norm": 0.5896215438842773,
+      "learning_rate": 0.0001,
+      "loss": 1.3036,
+      "step": 13885
+    },
+    {
+      "epoch": 1.613244263723497,
+      "grad_norm": 0.6163521409034729,
+      "learning_rate": 0.0001,
+      "loss": 1.4066,
+      "step": 13886
+    },
+    {
+      "epoch": 1.6133604414754574,
+      "grad_norm": 0.6672067642211914,
+      "learning_rate": 0.0001,
+      "loss": 1.5426,
+      "step": 13887
+    },
+    {
+      "epoch": 1.613476619227418,
+      "grad_norm": 0.5820779204368591,
+      "learning_rate": 0.0001,
+      "loss": 1.3377,
+      "step": 13888
+    },
+    {
+      "epoch": 1.6135927969793784,
+      "grad_norm": 0.6226430535316467,
+      "learning_rate": 0.0001,
+      "loss": 1.4441,
+      "step": 13889
+    },
+    {
+      "epoch": 1.613708974731339,
+      "grad_norm": 0.6409672498703003,
+      "learning_rate": 0.0001,
+      "loss": 1.5232,
+      "step": 13890
+    },
+    {
+      "epoch": 1.6138251524832996,
+      "grad_norm": 0.6150544881820679,
+      "learning_rate": 0.0001,
+      "loss": 1.4207,
+      "step": 13891
+    },
+    {
+      "epoch": 1.61394133023526,
+      "grad_norm": 0.6505727767944336,
+      "learning_rate": 0.0001,
+      "loss": 1.5336,
+      "step": 13892
+    },
+    {
+      "epoch": 1.6140575079872206,
+      "grad_norm": 0.6071352958679199,
+      "learning_rate": 0.0001,
+      "loss": 1.5639,
+      "step": 13893
+    },
+    {
+      "epoch": 1.614173685739181,
+      "grad_norm": 0.5638280510902405,
+      "learning_rate": 0.0001,
+      "loss": 1.4476,
+      "step": 13894
+    },
+    {
+      "epoch": 1.6142898634911416,
+      "grad_norm": 0.6587737798690796,
+      "learning_rate": 0.0001,
+      "loss": 1.6427,
+      "step": 13895
+    },
+    {
+      "epoch": 1.614406041243102,
+      "grad_norm": 0.6675108075141907,
+      "learning_rate": 0.0001,
+      "loss": 1.5899,
+      "step": 13896
+    },
+    {
+      "epoch": 1.6145222189950625,
+      "grad_norm": 0.6368352770805359,
+      "learning_rate": 0.0001,
+      "loss": 1.6209,
+      "step": 13897
+    },
+    {
+      "epoch": 1.614638396747023,
+      "grad_norm": 0.6211996674537659,
+      "learning_rate": 0.0001,
+      "loss": 1.416,
+      "step": 13898
+    },
+    {
+      "epoch": 1.6147545744989835,
+      "grad_norm": 0.6183685660362244,
+      "learning_rate": 0.0001,
+      "loss": 1.4282,
+      "step": 13899
+    },
+    {
+      "epoch": 1.614870752250944,
+      "grad_norm": 0.615385115146637,
+      "learning_rate": 0.0001,
+      "loss": 1.4632,
+      "step": 13900
+    },
+    {
+      "epoch": 1.6149869300029045,
+      "grad_norm": 0.6182390451431274,
+      "learning_rate": 0.0001,
+      "loss": 1.5878,
+      "step": 13901
+    },
+    {
+      "epoch": 1.615103107754865,
+      "grad_norm": 0.6461263298988342,
+      "learning_rate": 0.0001,
+      "loss": 1.4021,
+      "step": 13902
+    },
+    {
+      "epoch": 1.6152192855068255,
+      "grad_norm": 0.5660906434059143,
+      "learning_rate": 0.0001,
+      "loss": 1.2598,
+      "step": 13903
+    },
+    {
+      "epoch": 1.615335463258786,
+      "grad_norm": 0.6474366188049316,
+      "learning_rate": 0.0001,
+      "loss": 1.4905,
+      "step": 13904
+    },
+    {
+      "epoch": 1.6154516410107465,
+      "grad_norm": 0.6442524790763855,
+      "learning_rate": 0.0001,
+      "loss": 1.4119,
+      "step": 13905
+    },
+    {
+      "epoch": 1.615567818762707,
+      "grad_norm": 0.6138772964477539,
+      "learning_rate": 0.0001,
+      "loss": 1.4201,
+      "step": 13906
+    },
+    {
+      "epoch": 1.6156839965146674,
+      "grad_norm": 0.6902275085449219,
+      "learning_rate": 0.0001,
+      "loss": 1.7217,
+      "step": 13907
+    },
+    {
+      "epoch": 1.615800174266628,
+      "grad_norm": 0.5896263122558594,
+      "learning_rate": 0.0001,
+      "loss": 1.3702,
+      "step": 13908
+    },
+    {
+      "epoch": 1.6159163520185884,
+      "grad_norm": 0.6327207684516907,
+      "learning_rate": 0.0001,
+      "loss": 1.5585,
+      "step": 13909
+    },
+    {
+      "epoch": 1.616032529770549,
+      "grad_norm": 0.6263841986656189,
+      "learning_rate": 0.0001,
+      "loss": 1.371,
+      "step": 13910
+    },
+    {
+      "epoch": 1.6161487075225094,
+      "grad_norm": 0.6256563067436218,
+      "learning_rate": 0.0001,
+      "loss": 1.3766,
+      "step": 13911
+    },
+    {
+      "epoch": 1.6162648852744699,
+      "grad_norm": 0.6188286542892456,
+      "learning_rate": 0.0001,
+      "loss": 1.3662,
+      "step": 13912
+    },
+    {
+      "epoch": 1.6163810630264304,
+      "grad_norm": 0.5888686180114746,
+      "learning_rate": 0.0001,
+      "loss": 1.328,
+      "step": 13913
+    },
+    {
+      "epoch": 1.6164972407783909,
+      "grad_norm": 0.5912812352180481,
+      "learning_rate": 0.0001,
+      "loss": 1.3469,
+      "step": 13914
+    },
+    {
+      "epoch": 1.6166134185303513,
+      "grad_norm": 0.58259516954422,
+      "learning_rate": 0.0001,
+      "loss": 1.3371,
+      "step": 13915
+    },
+    {
+      "epoch": 1.6167295962823118,
+      "grad_norm": 0.636556088924408,
+      "learning_rate": 0.0001,
+      "loss": 1.5573,
+      "step": 13916
+    },
+    {
+      "epoch": 1.6168457740342723,
+      "grad_norm": 0.5752967596054077,
+      "learning_rate": 0.0001,
+      "loss": 1.3001,
+      "step": 13917
+    },
+    {
+      "epoch": 1.6169619517862328,
+      "grad_norm": 0.6357734203338623,
+      "learning_rate": 0.0001,
+      "loss": 1.5934,
+      "step": 13918
+    },
+    {
+      "epoch": 1.6170781295381933,
+      "grad_norm": 0.6322953701019287,
+      "learning_rate": 0.0001,
+      "loss": 1.5612,
+      "step": 13919
+    },
+    {
+      "epoch": 1.6171943072901538,
+      "grad_norm": 0.6408698558807373,
+      "learning_rate": 0.0001,
+      "loss": 1.5034,
+      "step": 13920
+    },
+    {
+      "epoch": 1.6173104850421143,
+      "grad_norm": 0.6438761949539185,
+      "learning_rate": 0.0001,
+      "loss": 1.5634,
+      "step": 13921
+    },
+    {
+      "epoch": 1.617426662794075,
+      "grad_norm": 0.6261228322982788,
+      "learning_rate": 0.0001,
+      "loss": 1.4357,
+      "step": 13922
+    },
+    {
+      "epoch": 1.6175428405460355,
+      "grad_norm": 0.6390764713287354,
+      "learning_rate": 0.0001,
+      "loss": 1.458,
+      "step": 13923
+    },
+    {
+      "epoch": 1.617659018297996,
+      "grad_norm": 0.5800904631614685,
+      "learning_rate": 0.0001,
+      "loss": 1.2348,
+      "step": 13924
+    },
+    {
+      "epoch": 1.6177751960499565,
+      "grad_norm": 0.5801419615745544,
+      "learning_rate": 0.0001,
+      "loss": 1.4536,
+      "step": 13925
+    },
+    {
+      "epoch": 1.617891373801917,
+      "grad_norm": 0.6229808926582336,
+      "learning_rate": 0.0001,
+      "loss": 1.4334,
+      "step": 13926
+    },
+    {
+      "epoch": 1.6180075515538774,
+      "grad_norm": 0.5984506011009216,
+      "learning_rate": 0.0001,
+      "loss": 1.3765,
+      "step": 13927
+    },
+    {
+      "epoch": 1.618123729305838,
+      "grad_norm": 0.6117129325866699,
+      "learning_rate": 0.0001,
+      "loss": 1.4251,
+      "step": 13928
+    },
+    {
+      "epoch": 1.6182399070577984,
+      "grad_norm": 0.6036126613616943,
+      "learning_rate": 0.0001,
+      "loss": 1.3457,
+      "step": 13929
+    },
+    {
+      "epoch": 1.618356084809759,
+      "grad_norm": 0.6293709874153137,
+      "learning_rate": 0.0001,
+      "loss": 1.5344,
+      "step": 13930
+    },
+    {
+      "epoch": 1.6184722625617194,
+      "grad_norm": 0.6404412388801575,
+      "learning_rate": 0.0001,
+      "loss": 1.5842,
+      "step": 13931
+    },
+    {
+      "epoch": 1.6185884403136799,
+      "grad_norm": 0.6405262351036072,
+      "learning_rate": 0.0001,
+      "loss": 1.5674,
+      "step": 13932
+    },
+    {
+      "epoch": 1.6187046180656406,
+      "grad_norm": 0.6383447647094727,
+      "learning_rate": 0.0001,
+      "loss": 1.5493,
+      "step": 13933
+    },
+    {
+      "epoch": 1.618820795817601,
+      "grad_norm": 0.6693515181541443,
+      "learning_rate": 0.0001,
+      "loss": 1.6526,
+      "step": 13934
+    },
+    {
+      "epoch": 1.6189369735695616,
+      "grad_norm": 0.6073635220527649,
+      "learning_rate": 0.0001,
+      "loss": 1.5188,
+      "step": 13935
+    },
+    {
+      "epoch": 1.619053151321522,
+      "grad_norm": 0.6150029301643372,
+      "learning_rate": 0.0001,
+      "loss": 1.4316,
+      "step": 13936
+    },
+    {
+      "epoch": 1.6191693290734825,
+      "grad_norm": 0.6156079173088074,
+      "learning_rate": 0.0001,
+      "loss": 1.3389,
+      "step": 13937
+    },
+    {
+      "epoch": 1.619285506825443,
+      "grad_norm": 0.6327203512191772,
+      "learning_rate": 0.0001,
+      "loss": 1.5262,
+      "step": 13938
+    },
+    {
+      "epoch": 1.6194016845774035,
+      "grad_norm": 0.5985980033874512,
+      "learning_rate": 0.0001,
+      "loss": 1.3391,
+      "step": 13939
+    },
+    {
+      "epoch": 1.619517862329364,
+      "grad_norm": 0.5909581780433655,
+      "learning_rate": 0.0001,
+      "loss": 1.4643,
+      "step": 13940
+    },
+    {
+      "epoch": 1.6196340400813245,
+      "grad_norm": 0.6321226358413696,
+      "learning_rate": 0.0001,
+      "loss": 1.5628,
+      "step": 13941
+    },
+    {
+      "epoch": 1.619750217833285,
+      "grad_norm": 0.6735860109329224,
+      "learning_rate": 0.0001,
+      "loss": 1.3241,
+      "step": 13942
+    },
+    {
+      "epoch": 1.6198663955852455,
+      "grad_norm": 0.6376870274543762,
+      "learning_rate": 0.0001,
+      "loss": 1.463,
+      "step": 13943
+    },
+    {
+      "epoch": 1.619982573337206,
+      "grad_norm": 0.6346738338470459,
+      "learning_rate": 0.0001,
+      "loss": 1.4896,
+      "step": 13944
+    },
+    {
+      "epoch": 1.6200987510891665,
+      "grad_norm": 0.6026695966720581,
+      "learning_rate": 0.0001,
+      "loss": 1.3668,
+      "step": 13945
+    },
+    {
+      "epoch": 1.620214928841127,
+      "grad_norm": 0.5611673593521118,
+      "learning_rate": 0.0001,
+      "loss": 1.3694,
+      "step": 13946
+    },
+    {
+      "epoch": 1.6203311065930874,
+      "grad_norm": 0.6367985606193542,
+      "learning_rate": 0.0001,
+      "loss": 1.5537,
+      "step": 13947
+    },
+    {
+      "epoch": 1.620447284345048,
+      "grad_norm": 0.6661987900733948,
+      "learning_rate": 0.0001,
+      "loss": 1.5108,
+      "step": 13948
+    },
+    {
+      "epoch": 1.6205634620970084,
+      "grad_norm": 0.6102688908576965,
+      "learning_rate": 0.0001,
+      "loss": 1.4665,
+      "step": 13949
+    },
+    {
+      "epoch": 1.620679639848969,
+      "grad_norm": 0.5848194360733032,
+      "learning_rate": 0.0001,
+      "loss": 1.4065,
+      "step": 13950
+    },
+    {
+      "epoch": 1.6207958176009294,
+      "grad_norm": 0.5830379128456116,
+      "learning_rate": 0.0001,
+      "loss": 1.513,
+      "step": 13951
+    },
+    {
+      "epoch": 1.6209119953528899,
+      "grad_norm": 0.6410690546035767,
+      "learning_rate": 0.0001,
+      "loss": 1.4485,
+      "step": 13952
+    },
+    {
+      "epoch": 1.6210281731048504,
+      "grad_norm": 0.6042836904525757,
+      "learning_rate": 0.0001,
+      "loss": 1.4449,
+      "step": 13953
+    },
+    {
+      "epoch": 1.6211443508568109,
+      "grad_norm": 0.6346921920776367,
+      "learning_rate": 0.0001,
+      "loss": 1.5349,
+      "step": 13954
+    },
+    {
+      "epoch": 1.6212605286087713,
+      "grad_norm": 0.5908709764480591,
+      "learning_rate": 0.0001,
+      "loss": 1.3898,
+      "step": 13955
+    },
+    {
+      "epoch": 1.6213767063607318,
+      "grad_norm": 0.6253523230552673,
+      "learning_rate": 0.0001,
+      "loss": 1.455,
+      "step": 13956
+    },
+    {
+      "epoch": 1.6214928841126923,
+      "grad_norm": 0.6313303112983704,
+      "learning_rate": 0.0001,
+      "loss": 1.4671,
+      "step": 13957
+    },
+    {
+      "epoch": 1.6216090618646528,
+      "grad_norm": 0.6551414728164673,
+      "learning_rate": 0.0001,
+      "loss": 1.5872,
+      "step": 13958
+    },
+    {
+      "epoch": 1.6217252396166133,
+      "grad_norm": 0.602879524230957,
+      "learning_rate": 0.0001,
+      "loss": 1.4072,
+      "step": 13959
+    },
+    {
+      "epoch": 1.6218414173685738,
+      "grad_norm": 0.6179625988006592,
+      "learning_rate": 0.0001,
+      "loss": 1.4204,
+      "step": 13960
+    },
+    {
+      "epoch": 1.6219575951205343,
+      "grad_norm": 0.5867196321487427,
+      "learning_rate": 0.0001,
+      "loss": 1.3604,
+      "step": 13961
+    },
+    {
+      "epoch": 1.6220737728724948,
+      "grad_norm": 0.6081652045249939,
+      "learning_rate": 0.0001,
+      "loss": 1.5271,
+      "step": 13962
+    },
+    {
+      "epoch": 1.6221899506244553,
+      "grad_norm": 0.5890710353851318,
+      "learning_rate": 0.0001,
+      "loss": 1.3407,
+      "step": 13963
+    },
+    {
+      "epoch": 1.622306128376416,
+      "grad_norm": 0.6529542207717896,
+      "learning_rate": 0.0001,
+      "loss": 1.5192,
+      "step": 13964
+    },
+    {
+      "epoch": 1.6224223061283765,
+      "grad_norm": 0.6106531023979187,
+      "learning_rate": 0.0001,
+      "loss": 1.4651,
+      "step": 13965
+    },
+    {
+      "epoch": 1.622538483880337,
+      "grad_norm": 0.6239684224128723,
+      "learning_rate": 0.0001,
+      "loss": 1.4949,
+      "step": 13966
+    },
+    {
+      "epoch": 1.6226546616322974,
+      "grad_norm": 0.6258800625801086,
+      "learning_rate": 0.0001,
+      "loss": 1.4508,
+      "step": 13967
+    },
+    {
+      "epoch": 1.622770839384258,
+      "grad_norm": 0.5880980491638184,
+      "learning_rate": 0.0001,
+      "loss": 1.4816,
+      "step": 13968
+    },
+    {
+      "epoch": 1.6228870171362184,
+      "grad_norm": 0.588895320892334,
+      "learning_rate": 0.0001,
+      "loss": 1.2776,
+      "step": 13969
+    },
+    {
+      "epoch": 1.623003194888179,
+      "grad_norm": 0.5825314521789551,
+      "learning_rate": 0.0001,
+      "loss": 1.3614,
+      "step": 13970
+    },
+    {
+      "epoch": 1.6231193726401394,
+      "grad_norm": 0.6523398756980896,
+      "learning_rate": 0.0001,
+      "loss": 1.4717,
+      "step": 13971
+    },
+    {
+      "epoch": 1.6232355503920999,
+      "grad_norm": 0.5812602639198303,
+      "learning_rate": 0.0001,
+      "loss": 1.3427,
+      "step": 13972
+    },
+    {
+      "epoch": 1.6233517281440604,
+      "grad_norm": 0.6783197522163391,
+      "learning_rate": 0.0001,
+      "loss": 1.4434,
+      "step": 13973
+    },
+    {
+      "epoch": 1.6234679058960209,
+      "grad_norm": 0.6416046619415283,
+      "learning_rate": 0.0001,
+      "loss": 1.3776,
+      "step": 13974
+    },
+    {
+      "epoch": 1.6235840836479816,
+      "grad_norm": 0.6453458666801453,
+      "learning_rate": 0.0001,
+      "loss": 1.5991,
+      "step": 13975
+    },
+    {
+      "epoch": 1.623700261399942,
+      "grad_norm": 0.6448045969009399,
+      "learning_rate": 0.0001,
+      "loss": 1.4237,
+      "step": 13976
+    },
+    {
+      "epoch": 1.6238164391519025,
+      "grad_norm": 0.6527649760246277,
+      "learning_rate": 0.0001,
+      "loss": 1.431,
+      "step": 13977
+    },
+    {
+      "epoch": 1.623932616903863,
+      "grad_norm": 0.6334816217422485,
+      "learning_rate": 0.0001,
+      "loss": 1.323,
+      "step": 13978
+    },
+    {
+      "epoch": 1.6240487946558235,
+      "grad_norm": 0.6312617659568787,
+      "learning_rate": 0.0001,
+      "loss": 1.5421,
+      "step": 13979
+    },
+    {
+      "epoch": 1.624164972407784,
+      "grad_norm": 0.6112910509109497,
+      "learning_rate": 0.0001,
+      "loss": 1.5255,
+      "step": 13980
+    },
+    {
+      "epoch": 1.6242811501597445,
+      "grad_norm": 0.6557388305664062,
+      "learning_rate": 0.0001,
+      "loss": 1.582,
+      "step": 13981
+    },
+    {
+      "epoch": 1.624397327911705,
+      "grad_norm": 0.6169906854629517,
+      "learning_rate": 0.0001,
+      "loss": 1.4738,
+      "step": 13982
+    },
+    {
+      "epoch": 1.6245135056636655,
+      "grad_norm": 0.6463947296142578,
+      "learning_rate": 0.0001,
+      "loss": 1.4123,
+      "step": 13983
+    },
+    {
+      "epoch": 1.624629683415626,
+      "grad_norm": 0.5652637481689453,
+      "learning_rate": 0.0001,
+      "loss": 1.2282,
+      "step": 13984
+    },
+    {
+      "epoch": 1.6247458611675865,
+      "grad_norm": 0.6067284941673279,
+      "learning_rate": 0.0001,
+      "loss": 1.5313,
+      "step": 13985
+    },
+    {
+      "epoch": 1.624862038919547,
+      "grad_norm": 0.6332645416259766,
+      "learning_rate": 0.0001,
+      "loss": 1.4048,
+      "step": 13986
+    },
+    {
+      "epoch": 1.6249782166715074,
+      "grad_norm": 0.5808968544006348,
+      "learning_rate": 0.0001,
+      "loss": 1.4871,
+      "step": 13987
+    },
+    {
+      "epoch": 1.625094394423468,
+      "grad_norm": 0.5867791771888733,
+      "learning_rate": 0.0001,
+      "loss": 1.44,
+      "step": 13988
+    },
+    {
+      "epoch": 1.6252105721754284,
+      "grad_norm": 0.643679678440094,
+      "learning_rate": 0.0001,
+      "loss": 1.5416,
+      "step": 13989
+    },
+    {
+      "epoch": 1.625326749927389,
+      "grad_norm": 0.6203275322914124,
+      "learning_rate": 0.0001,
+      "loss": 1.4913,
+      "step": 13990
+    },
+    {
+      "epoch": 1.6254429276793494,
+      "grad_norm": 0.5824859738349915,
+      "learning_rate": 0.0001,
+      "loss": 1.3436,
+      "step": 13991
+    },
+    {
+      "epoch": 1.6255591054313099,
+      "grad_norm": 0.6150936484336853,
+      "learning_rate": 0.0001,
+      "loss": 1.3999,
+      "step": 13992
+    },
+    {
+      "epoch": 1.6256752831832704,
+      "grad_norm": 0.579694926738739,
+      "learning_rate": 0.0001,
+      "loss": 1.6301,
+      "step": 13993
+    },
+    {
+      "epoch": 1.6257914609352309,
+      "grad_norm": 0.5742935538291931,
+      "learning_rate": 0.0001,
+      "loss": 1.4251,
+      "step": 13994
+    },
+    {
+      "epoch": 1.6259076386871913,
+      "grad_norm": 0.5996615886688232,
+      "learning_rate": 0.0001,
+      "loss": 1.4744,
+      "step": 13995
+    },
+    {
+      "epoch": 1.6260238164391518,
+      "grad_norm": 0.7094278931617737,
+      "learning_rate": 0.0001,
+      "loss": 1.4994,
+      "step": 13996
+    },
+    {
+      "epoch": 1.6261399941911123,
+      "grad_norm": 0.6233917474746704,
+      "learning_rate": 0.0001,
+      "loss": 1.4866,
+      "step": 13997
+    },
+    {
+      "epoch": 1.6262561719430728,
+      "grad_norm": 0.6166983246803284,
+      "learning_rate": 0.0001,
+      "loss": 1.375,
+      "step": 13998
+    },
+    {
+      "epoch": 1.6263723496950333,
+      "grad_norm": 0.638420581817627,
+      "learning_rate": 0.0001,
+      "loss": 1.4988,
+      "step": 13999
+    },
+    {
+      "epoch": 1.6264885274469938,
+      "grad_norm": 0.6417514681816101,
+      "learning_rate": 0.0001,
+      "loss": 1.4279,
+      "step": 14000
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 17214,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.1119190102441984e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}