diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,36422 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998557067962099,
+  "eval_steps": 500,
+  "global_step": 5197,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00019239093838680199,
+      "grad_norm": 74.76742090036745,
+      "learning_rate": 0.0,
+      "loss": 8.4376,
+      "step": 1
+    },
+    {
+      "epoch": 0.00038478187677360397,
+      "grad_norm": 79.33011477670689,
+      "learning_rate": 1.282051282051282e-07,
+      "loss": 9.0204,
+      "step": 2
+    },
+    {
+      "epoch": 0.0005771728151604059,
+      "grad_norm": 63.57224730851279,
+      "learning_rate": 2.564102564102564e-07,
+      "loss": 8.5677,
+      "step": 3
+    },
+    {
+      "epoch": 0.0007695637535472079,
+      "grad_norm": 67.32116379169396,
+      "learning_rate": 3.846153846153847e-07,
+      "loss": 8.3607,
+      "step": 4
+    },
+    {
+      "epoch": 0.0009619546919340099,
+      "grad_norm": 74.77611101179036,
+      "learning_rate": 5.128205128205128e-07,
+      "loss": 8.8793,
+      "step": 5
+    },
+    {
+      "epoch": 0.0011543456303208118,
+      "grad_norm": 68.88978016179558,
+      "learning_rate": 6.41025641025641e-07,
+      "loss": 8.3854,
+      "step": 6
+    },
+    {
+      "epoch": 0.0013467365687076138,
+      "grad_norm": 68.59001183362076,
+      "learning_rate": 7.692307692307694e-07,
+      "loss": 8.4307,
+      "step": 7
+    },
+    {
+      "epoch": 0.0015391275070944159,
+      "grad_norm": 57.56913548552778,
+      "learning_rate": 8.974358974358975e-07,
+      "loss": 8.0687,
+      "step": 8
+    },
+    {
+      "epoch": 0.001731518445481218,
+      "grad_norm": 69.4893662931534,
+      "learning_rate": 1.0256410256410257e-06,
+      "loss": 8.4788,
+      "step": 9
+    },
+    {
+      "epoch": 0.0019239093838680197,
+      "grad_norm": 62.94225019222865,
+      "learning_rate": 1.153846153846154e-06,
+      "loss": 7.821,
+      "step": 10
+    },
+    {
+      "epoch": 0.002116300322254822,
+      "grad_norm": 49.80883383756083,
+      "learning_rate": 1.282051282051282e-06,
+      "loss": 7.4857,
+      "step": 11
+    },
+    {
+      "epoch": 0.0023086912606416236,
+      "grad_norm": 48.00999095650297,
+      "learning_rate": 1.4102564102564104e-06,
+      "loss": 7.7341,
+      "step": 12
+    },
+    {
+      "epoch": 0.0025010821990284256,
+      "grad_norm": 39.303860305776105,
+      "learning_rate": 1.5384615384615387e-06,
+      "loss": 7.0903,
+      "step": 13
+    },
+    {
+      "epoch": 0.0026934731374152277,
+      "grad_norm": 35.541211246315,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 6.5489,
+      "step": 14
+    },
+    {
+      "epoch": 0.0028858640758020297,
+      "grad_norm": 30.95477189552307,
+      "learning_rate": 1.794871794871795e-06,
+      "loss": 6.587,
+      "step": 15
+    },
+    {
+      "epoch": 0.0030782550141888318,
+      "grad_norm": 28.897465233430832,
+      "learning_rate": 1.9230769230769234e-06,
+      "loss": 6.4831,
+      "step": 16
+    },
+    {
+      "epoch": 0.003270645952575634,
+      "grad_norm": 22.34399685869506,
+      "learning_rate": 2.0512820512820513e-06,
+      "loss": 4.7068,
+      "step": 17
+    },
+    {
+      "epoch": 0.003463036890962436,
+      "grad_norm": 16.908841810500384,
+      "learning_rate": 2.1794871794871797e-06,
+      "loss": 5.5074,
+      "step": 18
+    },
+    {
+      "epoch": 0.0036554278293492374,
+      "grad_norm": 13.551407483374586,
+      "learning_rate": 2.307692307692308e-06,
+      "loss": 5.2183,
+      "step": 19
+    },
+    {
+      "epoch": 0.0038478187677360395,
+      "grad_norm": 14.808699817554038,
+      "learning_rate": 2.435897435897436e-06,
+      "loss": 5.4141,
+      "step": 20
+    },
+    {
+      "epoch": 0.004040209706122842,
+      "grad_norm": 13.655148845423314,
+      "learning_rate": 2.564102564102564e-06,
+      "loss": 5.2218,
+      "step": 21
+    },
+    {
+      "epoch": 0.004232600644509644,
+      "grad_norm": 16.533753561053956,
+      "learning_rate": 2.6923076923076923e-06,
+      "loss": 5.4505,
+      "step": 22
+    },
+    {
+      "epoch": 0.004424991582896445,
+      "grad_norm": 11.385925927528353,
+      "learning_rate": 2.8205128205128207e-06,
+      "loss": 4.6615,
+      "step": 23
+    },
+    {
+      "epoch": 0.004617382521283247,
+      "grad_norm": 10.601999369666686,
+      "learning_rate": 2.948717948717949e-06,
+      "loss": 5.1492,
+      "step": 24
+    },
+    {
+      "epoch": 0.004809773459670049,
+      "grad_norm": 9.870085663339728,
+      "learning_rate": 3.0769230769230774e-06,
+      "loss": 5.2564,
+      "step": 25
+    },
+    {
+      "epoch": 0.005002164398056851,
+      "grad_norm": 8.790296868631707,
+      "learning_rate": 3.205128205128206e-06,
+      "loss": 4.7402,
+      "step": 26
+    },
+    {
+      "epoch": 0.005194555336443653,
+      "grad_norm": 7.967001982873081,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 4.6697,
+      "step": 27
+    },
+    {
+      "epoch": 0.005386946274830455,
+      "grad_norm": 7.900489391123951,
+      "learning_rate": 3.4615384615384617e-06,
+      "loss": 4.6543,
+      "step": 28
+    },
+    {
+      "epoch": 0.005579337213217257,
+      "grad_norm": 7.4170267227330315,
+      "learning_rate": 3.58974358974359e-06,
+      "loss": 4.4941,
+      "step": 29
+    },
+    {
+      "epoch": 0.0057717281516040594,
+      "grad_norm": 7.538454592325176,
+      "learning_rate": 3.7179487179487184e-06,
+      "loss": 4.7992,
+      "step": 30
+    },
+    {
+      "epoch": 0.0059641190899908615,
+      "grad_norm": 7.459940811756082,
+      "learning_rate": 3.846153846153847e-06,
+      "loss": 4.4357,
+      "step": 31
+    },
+    {
+      "epoch": 0.0061565100283776635,
+      "grad_norm": 7.847537412484949,
+      "learning_rate": 3.974358974358974e-06,
+      "loss": 4.5354,
+      "step": 32
+    },
+    {
+      "epoch": 0.0063489009667644656,
+      "grad_norm": 7.24492265939584,
+      "learning_rate": 4.102564102564103e-06,
+      "loss": 4.4586,
+      "step": 33
+    },
+    {
+      "epoch": 0.006541291905151268,
+      "grad_norm": 5.971361058896558,
+      "learning_rate": 4.230769230769231e-06,
+      "loss": 4.2818,
+      "step": 34
+    },
+    {
+      "epoch": 0.00673368284353807,
+      "grad_norm": 5.6771035216833114,
+      "learning_rate": 4.358974358974359e-06,
+      "loss": 4.1985,
+      "step": 35
+    },
+    {
+      "epoch": 0.006926073781924872,
+      "grad_norm": 5.41261569254956,
+      "learning_rate": 4.487179487179488e-06,
+      "loss": 4.2062,
+      "step": 36
+    },
+    {
+      "epoch": 0.007118464720311674,
+      "grad_norm": 5.521039092368365,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 4.4896,
+      "step": 37
+    },
+    {
+      "epoch": 0.007310855658698475,
+      "grad_norm": 5.360452636698551,
+      "learning_rate": 4.743589743589744e-06,
+      "loss": 4.3149,
+      "step": 38
+    },
+    {
+      "epoch": 0.007503246597085277,
+      "grad_norm": 5.368065054492925,
+      "learning_rate": 4.871794871794872e-06,
+      "loss": 4.1896,
+      "step": 39
+    },
+    {
+      "epoch": 0.007695637535472079,
+      "grad_norm": 5.1135235788626865,
+      "learning_rate": 5e-06,
+      "loss": 4.0333,
+      "step": 40
+    },
+    {
+      "epoch": 0.007888028473858881,
+      "grad_norm": 5.5678566409091275,
+      "learning_rate": 5.128205128205128e-06,
+      "loss": 4.1703,
+      "step": 41
+    },
+    {
+      "epoch": 0.008080419412245684,
+      "grad_norm": 4.8545105498335195,
+      "learning_rate": 5.256410256410257e-06,
+      "loss": 4.0248,
+      "step": 42
+    },
+    {
+      "epoch": 0.008272810350632485,
+      "grad_norm": 4.750706695233615,
+      "learning_rate": 5.384615384615385e-06,
+      "loss": 4.2852,
+      "step": 43
+    },
+    {
+      "epoch": 0.008465201289019288,
+      "grad_norm": 4.712280205974809,
+      "learning_rate": 5.512820512820514e-06,
+      "loss": 3.9782,
+      "step": 44
+    },
+    {
+      "epoch": 0.00865759222740609,
+      "grad_norm": 5.024882246853174,
+      "learning_rate": 5.641025641025641e-06,
+      "loss": 4.0336,
+      "step": 45
+    },
+    {
+      "epoch": 0.00884998316579289,
+      "grad_norm": 5.021316765685685,
+      "learning_rate": 5.769230769230769e-06,
+      "loss": 4.0744,
+      "step": 46
+    },
+    {
+      "epoch": 0.009042374104179693,
+      "grad_norm": 5.201702241256377,
+      "learning_rate": 5.897435897435898e-06,
+      "loss": 4.4529,
+      "step": 47
+    },
+    {
+      "epoch": 0.009234765042566494,
+      "grad_norm": 5.001138606189981,
+      "learning_rate": 6.025641025641026e-06,
+      "loss": 4.0311,
+      "step": 48
+    },
+    {
+      "epoch": 0.009427155980953297,
+      "grad_norm": 4.9844879724064795,
+      "learning_rate": 6.153846153846155e-06,
+      "loss": 4.2627,
+      "step": 49
+    },
+    {
+      "epoch": 0.009619546919340098,
+      "grad_norm": 4.731628559545773,
+      "learning_rate": 6.282051282051282e-06,
+      "loss": 4.1636,
+      "step": 50
+    },
+    {
+      "epoch": 0.009811937857726901,
+      "grad_norm": 4.5633120357250805,
+      "learning_rate": 6.410256410256412e-06,
+      "loss": 4.0436,
+      "step": 51
+    },
+    {
+      "epoch": 0.010004328796113703,
+      "grad_norm": 4.695012144968728,
+      "learning_rate": 6.538461538461539e-06,
+      "loss": 4.0528,
+      "step": 52
+    },
+    {
+      "epoch": 0.010196719734500505,
+      "grad_norm": 4.53092054564012,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 3.9257,
+      "step": 53
+    },
+    {
+      "epoch": 0.010389110672887307,
+      "grad_norm": 4.337377653716561,
+      "learning_rate": 6.794871794871796e-06,
+      "loss": 4.0441,
+      "step": 54
+    },
+    {
+      "epoch": 0.01058150161127411,
+      "grad_norm": 4.945712311306427,
+      "learning_rate": 6.923076923076923e-06,
+      "loss": 4.2816,
+      "step": 55
+    },
+    {
+      "epoch": 0.01077389254966091,
+      "grad_norm": 4.369511362858024,
+      "learning_rate": 7.051282051282053e-06,
+      "loss": 4.3336,
+      "step": 56
+    },
+    {
+      "epoch": 0.010966283488047714,
+      "grad_norm": 4.367257360959113,
+      "learning_rate": 7.17948717948718e-06,
+      "loss": 4.0273,
+      "step": 57
+    },
+    {
+      "epoch": 0.011158674426434515,
+      "grad_norm": 4.268713174566179,
+      "learning_rate": 7.307692307692308e-06,
+      "loss": 4.1142,
+      "step": 58
+    },
+    {
+      "epoch": 0.011351065364821318,
+      "grad_norm": 4.576801917465012,
+      "learning_rate": 7.435897435897437e-06,
+      "loss": 4.1424,
+      "step": 59
+    },
+    {
+      "epoch": 0.011543456303208119,
+      "grad_norm": 4.485499742938407,
+      "learning_rate": 7.564102564102564e-06,
+      "loss": 3.9728,
+      "step": 60
+    },
+    {
+      "epoch": 0.01173584724159492,
+      "grad_norm": 4.343928381095612,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 4.0,
+      "step": 61
+    },
+    {
+      "epoch": 0.011928238179981723,
+      "grad_norm": 4.1722267079993545,
+      "learning_rate": 7.820512820512822e-06,
+      "loss": 4.2398,
+      "step": 62
+    },
+    {
+      "epoch": 0.012120629118368524,
+      "grad_norm": 4.772547537322177,
+      "learning_rate": 7.948717948717949e-06,
+      "loss": 4.1154,
+      "step": 63
+    },
+    {
+      "epoch": 0.012313020056755327,
+      "grad_norm": 4.611962856854712,
+      "learning_rate": 8.076923076923077e-06,
+      "loss": 3.7417,
+      "step": 64
+    },
+    {
+      "epoch": 0.012505410995142128,
+      "grad_norm": 4.146541278246039,
+      "learning_rate": 8.205128205128205e-06,
+      "loss": 3.7415,
+      "step": 65
+    },
+    {
+      "epoch": 0.012697801933528931,
+      "grad_norm": 4.914474908131988,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 3.862,
+      "step": 66
+    },
+    {
+      "epoch": 0.012890192871915732,
+      "grad_norm": 4.6517223494835545,
+      "learning_rate": 8.461538461538462e-06,
+      "loss": 3.7619,
+      "step": 67
+    },
+    {
+      "epoch": 0.013082583810302535,
+      "grad_norm": 4.555397929838847,
+      "learning_rate": 8.58974358974359e-06,
+      "loss": 4.0253,
+      "step": 68
+    },
+    {
+      "epoch": 0.013274974748689336,
+      "grad_norm": 4.622087027632129,
+      "learning_rate": 8.717948717948719e-06,
+      "loss": 4.0351,
+      "step": 69
+    },
+    {
+      "epoch": 0.01346736568707614,
+      "grad_norm": 4.395045054859037,
+      "learning_rate": 8.846153846153847e-06,
+      "loss": 3.82,
+      "step": 70
+    },
+    {
+      "epoch": 0.01365975662546294,
+      "grad_norm": 4.6167319022377,
+      "learning_rate": 8.974358974358976e-06,
+      "loss": 3.7657,
+      "step": 71
+    },
+    {
+      "epoch": 0.013852147563849743,
+      "grad_norm": 4.5910472985388076,
+      "learning_rate": 9.102564102564104e-06,
+      "loss": 3.9611,
+      "step": 72
+    },
+    {
+      "epoch": 0.014044538502236545,
+      "grad_norm": 4.374055043524445,
+      "learning_rate": 9.230769230769232e-06,
+      "loss": 3.7861,
+      "step": 73
+    },
+    {
+      "epoch": 0.014236929440623347,
+      "grad_norm": 4.570611540896282,
+      "learning_rate": 9.358974358974359e-06,
+      "loss": 3.8689,
+      "step": 74
+    },
+    {
+      "epoch": 0.014429320379010149,
+      "grad_norm": 4.625849517509607,
+      "learning_rate": 9.487179487179487e-06,
+      "loss": 3.7863,
+      "step": 75
+    },
+    {
+      "epoch": 0.01462171131739695,
+      "grad_norm": 4.664546305909135,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 3.8722,
+      "step": 76
+    },
+    {
+      "epoch": 0.014814102255783753,
+      "grad_norm": 4.43683803501588,
+      "learning_rate": 9.743589743589744e-06,
+      "loss": 3.7627,
+      "step": 77
+    },
+    {
+      "epoch": 0.015006493194170554,
+      "grad_norm": 5.167329819621188,
+      "learning_rate": 9.871794871794872e-06,
+      "loss": 4.0601,
+      "step": 78
+    },
+    {
+      "epoch": 0.015198884132557357,
+      "grad_norm": 4.456847868444916,
+      "learning_rate": 1e-05,
+      "loss": 3.9479,
+      "step": 79
+    },
+    {
+      "epoch": 0.015391275070944158,
+      "grad_norm": 4.668342257384097,
+      "learning_rate": 1.012820512820513e-05,
+      "loss": 3.6622,
+      "step": 80
+    },
+    {
+      "epoch": 0.01558366600933096,
+      "grad_norm": 4.303406381730383,
+      "learning_rate": 1.0256410256410256e-05,
+      "loss": 3.87,
+      "step": 81
+    },
+    {
+      "epoch": 0.015776056947717762,
+      "grad_norm": 4.169660944394991,
+      "learning_rate": 1.0384615384615386e-05,
+      "loss": 3.6529,
+      "step": 82
+    },
+    {
+      "epoch": 0.015968447886104565,
+      "grad_norm": 4.273342756304039,
+      "learning_rate": 1.0512820512820514e-05,
+      "loss": 3.6119,
+      "step": 83
+    },
+    {
+      "epoch": 0.016160838824491368,
+      "grad_norm": 4.669850813879593,
+      "learning_rate": 1.0641025641025643e-05,
+      "loss": 3.8179,
+      "step": 84
+    },
+    {
+      "epoch": 0.016353229762878167,
+      "grad_norm": 7.331436537355904,
+      "learning_rate": 1.076923076923077e-05,
+      "loss": 3.5223,
+      "step": 85
+    },
+    {
+      "epoch": 0.01654562070126497,
+      "grad_norm": 6.2429861497816015,
+      "learning_rate": 1.0897435897435898e-05,
+      "loss": 3.7631,
+      "step": 86
+    },
+    {
+      "epoch": 0.016738011639651773,
+      "grad_norm": 4.378790876951514,
+      "learning_rate": 1.1025641025641028e-05,
+      "loss": 3.8734,
+      "step": 87
+    },
+    {
+      "epoch": 0.016930402578038576,
+      "grad_norm": 4.317700499064961,
+      "learning_rate": 1.1153846153846154e-05,
+      "loss": 3.8355,
+      "step": 88
+    },
+    {
+      "epoch": 0.017122793516425375,
+      "grad_norm": 4.362909653909211,
+      "learning_rate": 1.1282051282051283e-05,
+      "loss": 3.9413,
+      "step": 89
+    },
+    {
+      "epoch": 0.01731518445481218,
+      "grad_norm": 4.47721260976014,
+      "learning_rate": 1.1410256410256411e-05,
+      "loss": 3.7691,
+      "step": 90
+    },
+    {
+      "epoch": 0.01750757539319898,
+      "grad_norm": 4.394535849262481,
+      "learning_rate": 1.1538461538461538e-05,
+      "loss": 3.9311,
+      "step": 91
+    },
+    {
+      "epoch": 0.01769996633158578,
+      "grad_norm": 4.591108623108764,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 3.8362,
+      "step": 92
+    },
+    {
+      "epoch": 0.017892357269972584,
+      "grad_norm": 4.758534192270426,
+      "learning_rate": 1.1794871794871796e-05,
+      "loss": 3.8103,
+      "step": 93
+    },
+    {
+      "epoch": 0.018084748208359386,
+      "grad_norm": 4.660587373865889,
+      "learning_rate": 1.1923076923076925e-05,
+      "loss": 3.6185,
+      "step": 94
+    },
+    {
+      "epoch": 0.01827713914674619,
+      "grad_norm": 4.7074581831244355,
+      "learning_rate": 1.2051282051282051e-05,
+      "loss": 3.7277,
+      "step": 95
+    },
+    {
+      "epoch": 0.01846953008513299,
+      "grad_norm": 4.5824011572643055,
+      "learning_rate": 1.217948717948718e-05,
+      "loss": 3.7698,
+      "step": 96
+    },
+    {
+      "epoch": 0.01866192102351979,
+      "grad_norm": 4.276925037875346,
+      "learning_rate": 1.230769230769231e-05,
+      "loss": 3.7242,
+      "step": 97
+    },
+    {
+      "epoch": 0.018854311961906595,
+      "grad_norm": 4.4304881204098665,
+      "learning_rate": 1.2435897435897436e-05,
+      "loss": 3.7302,
+      "step": 98
+    },
+    {
+      "epoch": 0.019046702900293398,
+      "grad_norm": 4.214108995827399,
+      "learning_rate": 1.2564102564102565e-05,
+      "loss": 3.6608,
+      "step": 99
+    },
+    {
+      "epoch": 0.019239093838680197,
+      "grad_norm": 4.322398943427801,
+      "learning_rate": 1.2692307692307693e-05,
+      "loss": 3.8477,
+      "step": 100
+    },
+    {
+      "epoch": 0.019431484777067,
+      "grad_norm": 4.780236732037893,
+      "learning_rate": 1.2820512820512823e-05,
+      "loss": 4.0026,
+      "step": 101
+    },
+    {
+      "epoch": 0.019623875715453803,
+      "grad_norm": 4.658841173270714,
+      "learning_rate": 1.294871794871795e-05,
+      "loss": 3.7324,
+      "step": 102
+    },
+    {
+      "epoch": 0.019816266653840606,
+      "grad_norm": 4.55725597046229,
+      "learning_rate": 1.3076923076923078e-05,
+      "loss": 3.7755,
+      "step": 103
+    },
+    {
+      "epoch": 0.020008657592227405,
+      "grad_norm": 4.182911981287969,
+      "learning_rate": 1.3205128205128207e-05,
+      "loss": 3.5456,
+      "step": 104
+    },
+    {
+      "epoch": 0.020201048530614208,
+      "grad_norm": 4.3938535238520755,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 3.9335,
+      "step": 105
+    },
+    {
+      "epoch": 0.02039343946900101,
+      "grad_norm": 4.350597904084007,
+      "learning_rate": 1.3461538461538463e-05,
+      "loss": 3.7623,
+      "step": 106
+    },
+    {
+      "epoch": 0.02058583040738781,
+      "grad_norm": 4.237115801611297,
+      "learning_rate": 1.3589743589743592e-05,
+      "loss": 3.8867,
+      "step": 107
+    },
+    {
+      "epoch": 0.020778221345774613,
+      "grad_norm": 4.354987167906702,
+      "learning_rate": 1.3717948717948718e-05,
+      "loss": 3.5333,
+      "step": 108
+    },
+    {
+      "epoch": 0.020970612284161416,
+      "grad_norm": 4.452174731549598,
+      "learning_rate": 1.3846153846153847e-05,
+      "loss": 3.6496,
+      "step": 109
+    },
+    {
+      "epoch": 0.02116300322254822,
+      "grad_norm": 4.4495674993059495,
+      "learning_rate": 1.3974358974358975e-05,
+      "loss": 3.7172,
+      "step": 110
+    },
+    {
+      "epoch": 0.02135539416093502,
+      "grad_norm": 4.543646395267219,
+      "learning_rate": 1.4102564102564105e-05,
+      "loss": 3.7605,
+      "step": 111
+    },
+    {
+      "epoch": 0.02154778509932182,
+      "grad_norm": 4.052395949971428,
+      "learning_rate": 1.4230769230769232e-05,
+      "loss": 3.8176,
+      "step": 112
+    },
+    {
+      "epoch": 0.021740176037708624,
+      "grad_norm": 4.401048841096906,
+      "learning_rate": 1.435897435897436e-05,
+      "loss": 3.4555,
+      "step": 113
+    },
+    {
+      "epoch": 0.021932566976095427,
+      "grad_norm": 4.625123514353994,
+      "learning_rate": 1.4487179487179489e-05,
+      "loss": 3.8474,
+      "step": 114
+    },
+    {
+      "epoch": 0.022124957914482227,
+      "grad_norm": 4.514859363495827,
+      "learning_rate": 1.4615384615384615e-05,
+      "loss": 3.6824,
+      "step": 115
+    },
+    {
+      "epoch": 0.02231734885286903,
+      "grad_norm": 4.420569985170252,
+      "learning_rate": 1.4743589743589745e-05,
+      "loss": 3.6996,
+      "step": 116
+    },
+    {
+      "epoch": 0.022509739791255833,
+      "grad_norm": 4.540988228996743,
+      "learning_rate": 1.4871794871794874e-05,
+      "loss": 3.5873,
+      "step": 117
+    },
+    {
+      "epoch": 0.022702130729642635,
+      "grad_norm": 4.5783834011783515,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 3.582,
+      "step": 118
+    },
+    {
+      "epoch": 0.022894521668029435,
+      "grad_norm": 4.455441823233252,
+      "learning_rate": 1.5128205128205129e-05,
+      "loss": 3.8111,
+      "step": 119
+    },
+    {
+      "epoch": 0.023086912606416238,
+      "grad_norm": 4.49858045666079,
+      "learning_rate": 1.5256410256410257e-05,
+      "loss": 3.79,
+      "step": 120
+    },
+    {
+      "epoch": 0.02327930354480304,
+      "grad_norm": 4.469000446944302,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 3.7944,
+      "step": 121
+    },
+    {
+      "epoch": 0.02347169448318984,
+      "grad_norm": 4.533830407946132,
+      "learning_rate": 1.5512820512820516e-05,
+      "loss": 3.5317,
+      "step": 122
+    },
+    {
+      "epoch": 0.023664085421576643,
+      "grad_norm": 4.92721168346582,
+      "learning_rate": 1.5641025641025644e-05,
+      "loss": 3.7331,
+      "step": 123
+    },
+    {
+      "epoch": 0.023856476359963446,
+      "grad_norm": 4.487648525802084,
+      "learning_rate": 1.576923076923077e-05,
+      "loss": 3.8542,
+      "step": 124
+    },
+    {
+      "epoch": 0.02404886729835025,
+      "grad_norm": 4.925984376529946,
+      "learning_rate": 1.5897435897435897e-05,
+      "loss": 3.6392,
+      "step": 125
+    },
+    {
+      "epoch": 0.02424125823673705,
+      "grad_norm": 4.4980669790303605,
+      "learning_rate": 1.602564102564103e-05,
+      "loss": 3.8276,
+      "step": 126
+    },
+    {
+      "epoch": 0.02443364917512385,
+      "grad_norm": 4.434622791923132,
+      "learning_rate": 1.6153846153846154e-05,
+      "loss": 3.6931,
+      "step": 127
+    },
+    {
+      "epoch": 0.024626040113510654,
+      "grad_norm": 4.277116677222338,
+      "learning_rate": 1.6282051282051282e-05,
+      "loss": 3.5876,
+      "step": 128
+    },
+    {
+      "epoch": 0.024818431051897457,
+      "grad_norm": 4.645797354114801,
+      "learning_rate": 1.641025641025641e-05,
+      "loss": 3.6233,
+      "step": 129
+    },
+    {
+      "epoch": 0.025010821990284256,
+      "grad_norm": 4.6306254124571025,
+      "learning_rate": 1.653846153846154e-05,
+      "loss": 3.8146,
+      "step": 130
+    },
+    {
+      "epoch": 0.02520321292867106,
+      "grad_norm": 4.277750837777625,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 3.4728,
+      "step": 131
+    },
+    {
+      "epoch": 0.025395603867057862,
+      "grad_norm": 4.3879414914856465,
+      "learning_rate": 1.6794871794871796e-05,
+      "loss": 3.843,
+      "step": 132
+    },
+    {
+      "epoch": 0.025587994805444665,
+      "grad_norm": 4.380093351628392,
+      "learning_rate": 1.6923076923076924e-05,
+      "loss": 3.6971,
+      "step": 133
+    },
+    {
+      "epoch": 0.025780385743831465,
+      "grad_norm": 4.406323569276621,
+      "learning_rate": 1.7051282051282053e-05,
+      "loss": 3.6803,
+      "step": 134
+    },
+    {
+      "epoch": 0.025972776682218267,
+      "grad_norm": 4.731237200203369,
+      "learning_rate": 1.717948717948718e-05,
+      "loss": 3.8734,
+      "step": 135
+    },
+    {
+      "epoch": 0.02616516762060507,
+      "grad_norm": 4.9137956497562705,
+      "learning_rate": 1.730769230769231e-05,
+      "loss": 3.4023,
+      "step": 136
+    },
+    {
+      "epoch": 0.02635755855899187,
+      "grad_norm": 4.506766566825677,
+      "learning_rate": 1.7435897435897438e-05,
+      "loss": 3.3433,
+      "step": 137
+    },
+    {
+      "epoch": 0.026549949497378673,
+      "grad_norm": 4.118078699585032,
+      "learning_rate": 1.7564102564102566e-05,
+      "loss": 3.8904,
+      "step": 138
+    },
+    {
+      "epoch": 0.026742340435765476,
+      "grad_norm": 4.658207005464315,
+      "learning_rate": 1.7692307692307694e-05,
+      "loss": 3.8714,
+      "step": 139
+    },
+    {
+      "epoch": 0.02693473137415228,
+      "grad_norm": 4.403956100022367,
+      "learning_rate": 1.7820512820512823e-05,
+      "loss": 3.8094,
+      "step": 140
+    },
+    {
+      "epoch": 0.027127122312539078,
+      "grad_norm": 4.06941786894682,
+      "learning_rate": 1.794871794871795e-05,
+      "loss": 3.6068,
+      "step": 141
+    },
+    {
+      "epoch": 0.02731951325092588,
+      "grad_norm": 4.387282928649993,
+      "learning_rate": 1.807692307692308e-05,
+      "loss": 3.7082,
+      "step": 142
+    },
+    {
+      "epoch": 0.027511904189312684,
+      "grad_norm": 4.24591923328563,
+      "learning_rate": 1.8205128205128208e-05,
+      "loss": 3.4015,
+      "step": 143
+    },
+    {
+      "epoch": 0.027704295127699487,
+      "grad_norm": 4.526362132095499,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 3.5368,
+      "step": 144
+    },
+    {
+      "epoch": 0.027896686066086286,
+      "grad_norm": 4.415493777098275,
+      "learning_rate": 1.8461538461538465e-05,
+      "loss": 3.3588,
+      "step": 145
+    },
+    {
+      "epoch": 0.02808907700447309,
+      "grad_norm": 4.119864824146889,
+      "learning_rate": 1.8589743589743593e-05,
+      "loss": 3.4158,
+      "step": 146
+    },
+    {
+      "epoch": 0.028281467942859892,
+      "grad_norm": 4.04973312615369,
+      "learning_rate": 1.8717948717948718e-05,
+      "loss": 3.7722,
+      "step": 147
+    },
+    {
+      "epoch": 0.028473858881246695,
+      "grad_norm": 4.423236787705223,
+      "learning_rate": 1.8846153846153846e-05,
+      "loss": 3.3338,
+      "step": 148
+    },
+    {
+      "epoch": 0.028666249819633494,
+      "grad_norm": 4.474091155535545,
+      "learning_rate": 1.8974358974358975e-05,
+      "loss": 3.8837,
+      "step": 149
+    },
+    {
+      "epoch": 0.028858640758020297,
+      "grad_norm": 4.873216439823803,
+      "learning_rate": 1.9102564102564106e-05,
+      "loss": 3.7152,
+      "step": 150
+    },
+    {
+      "epoch": 0.0290510316964071,
+      "grad_norm": 4.423889220633257,
+      "learning_rate": 1.923076923076923e-05,
+      "loss": 3.4435,
+      "step": 151
+    },
+    {
+      "epoch": 0.0292434226347939,
+      "grad_norm": 4.149734530944072,
+      "learning_rate": 1.935897435897436e-05,
+      "loss": 3.5883,
+      "step": 152
+    },
+    {
+      "epoch": 0.029435813573180702,
+      "grad_norm": 4.187223930183034,
+      "learning_rate": 1.9487179487179488e-05,
+      "loss": 3.6978,
+      "step": 153
+    },
+    {
+      "epoch": 0.029628204511567505,
+      "grad_norm": 4.290701328402348,
+      "learning_rate": 1.9615384615384617e-05,
+      "loss": 3.735,
+      "step": 154
+    },
+    {
+      "epoch": 0.02982059544995431,
+      "grad_norm": 4.352087140369863,
+      "learning_rate": 1.9743589743589745e-05,
+      "loss": 3.6623,
+      "step": 155
+    },
+    {
+      "epoch": 0.030012986388341108,
+      "grad_norm": 4.73365696619401,
+      "learning_rate": 1.9871794871794873e-05,
+      "loss": 3.6984,
+      "step": 156
+    },
+    {
+      "epoch": 0.03020537732672791,
+      "grad_norm": 4.246830209627491,
+      "learning_rate": 2e-05,
+      "loss": 3.6604,
+      "step": 157
+    },
+    {
+      "epoch": 0.030397768265114714,
+      "grad_norm": 4.5749515377826,
+      "learning_rate": 1.9999998058057616e-05,
+      "loss": 3.5818,
+      "step": 158
+    },
+    {
+      "epoch": 0.030590159203501516,
+      "grad_norm": 4.653971955282082,
+      "learning_rate": 1.9999992232231216e-05,
+      "loss": 3.4215,
+      "step": 159
+    },
+    {
+      "epoch": 0.030782550141888316,
+      "grad_norm": 4.340101874798806,
+      "learning_rate": 1.999998252252306e-05,
+      "loss": 3.7264,
+      "step": 160
+    },
+    {
+      "epoch": 0.03097494108027512,
+      "grad_norm": 4.010438788037505,
+      "learning_rate": 1.9999968928936924e-05,
+      "loss": 3.438,
+      "step": 161
+    },
+    {
+      "epoch": 0.03116733201866192,
+      "grad_norm": 4.179721697710951,
+      "learning_rate": 1.999995145147809e-05,
+      "loss": 3.3307,
+      "step": 162
+    },
+    {
+      "epoch": 0.031359722957048725,
+      "grad_norm": 4.689340678877607,
+      "learning_rate": 1.9999930090153335e-05,
+      "loss": 3.5852,
+      "step": 163
+    },
+    {
+      "epoch": 0.031552113895435524,
+      "grad_norm": 4.576443780815137,
+      "learning_rate": 1.9999904844970963e-05,
+      "loss": 3.6416,
+      "step": 164
+    },
+    {
+      "epoch": 0.03174450483382233,
+      "grad_norm": 4.21518375873765,
+      "learning_rate": 1.999987571594078e-05,
+      "loss": 3.7872,
+      "step": 165
+    },
+    {
+      "epoch": 0.03193689577220913,
+      "grad_norm": 4.472621611693934,
+      "learning_rate": 1.99998427030741e-05,
+      "loss": 3.4556,
+      "step": 166
+    },
+    {
+      "epoch": 0.03212928671059593,
+      "grad_norm": 8.050356181912623,
+      "learning_rate": 1.999980580638374e-05,
+      "loss": 3.723,
+      "step": 167
+    },
+    {
+      "epoch": 0.032321677648982736,
+      "grad_norm": 4.6617719140808,
+      "learning_rate": 1.999976502588403e-05,
+      "loss": 3.7401,
+      "step": 168
+    },
+    {
+      "epoch": 0.032514068587369535,
+      "grad_norm": 3.3101421702679312,
+      "learning_rate": 1.9999720361590812e-05,
+      "loss": 3.2813,
+      "step": 169
+    },
+    {
+      "epoch": 0.032706459525756335,
+      "grad_norm": 4.6107348464636475,
+      "learning_rate": 1.9999671813521435e-05,
+      "loss": 3.5936,
+      "step": 170
+    },
+    {
+      "epoch": 0.03289885046414314,
+      "grad_norm": 4.63765907941053,
+      "learning_rate": 1.999961938169475e-05,
+      "loss": 3.5854,
+      "step": 171
+    },
+    {
+      "epoch": 0.03309124140252994,
+      "grad_norm": 4.485074960251747,
+      "learning_rate": 1.9999563066131124e-05,
+      "loss": 3.649,
+      "step": 172
+    },
+    {
+      "epoch": 0.03328363234091674,
+      "grad_norm": 4.243284880801974,
+      "learning_rate": 1.9999502866852427e-05,
+      "loss": 3.506,
+      "step": 173
+    },
+    {
+      "epoch": 0.033476023279303546,
+      "grad_norm": 4.586582617434041,
+      "learning_rate": 1.999943878388204e-05,
+      "loss": 3.7544,
+      "step": 174
+    },
+    {
+      "epoch": 0.033668414217690346,
+      "grad_norm": 4.417672824030158,
+      "learning_rate": 1.9999370817244853e-05,
+      "loss": 3.4676,
+      "step": 175
+    },
+    {
+      "epoch": 0.03386080515607715,
+      "grad_norm": 4.230917399317115,
+      "learning_rate": 1.9999298966967264e-05,
+      "loss": 3.681,
+      "step": 176
+    },
+    {
+      "epoch": 0.03405319609446395,
+      "grad_norm": 4.113882991809835,
+      "learning_rate": 1.9999223233077178e-05,
+      "loss": 3.4992,
+      "step": 177
+    },
+    {
+      "epoch": 0.03424558703285075,
+      "grad_norm": 4.250758811829217,
+      "learning_rate": 1.999914361560401e-05,
+      "loss": 3.7114,
+      "step": 178
+    },
+    {
+      "epoch": 0.03443797797123756,
+      "grad_norm": 4.102502621742442,
+      "learning_rate": 1.9999060114578682e-05,
+      "loss": 3.6501,
+      "step": 179
+    },
+    {
+      "epoch": 0.03463036890962436,
+      "grad_norm": 4.3034734534936,
+      "learning_rate": 1.9998972730033624e-05,
+      "loss": 3.6393,
+      "step": 180
+    },
+    {
+      "epoch": 0.034822759848011156,
+      "grad_norm": 4.059597498288871,
+      "learning_rate": 1.9998881462002778e-05,
+      "loss": 3.5942,
+      "step": 181
+    },
+    {
+      "epoch": 0.03501515078639796,
+      "grad_norm": 4.414222472162948,
+      "learning_rate": 1.9998786310521585e-05,
+      "loss": 3.4541,
+      "step": 182
+    },
+    {
+      "epoch": 0.03520754172478476,
+      "grad_norm": 4.377744025191428,
+      "learning_rate": 1.9998687275627008e-05,
+      "loss": 3.6702,
+      "step": 183
+    },
+    {
+      "epoch": 0.03539993266317156,
+      "grad_norm": 4.4169255280591235,
+      "learning_rate": 1.9998584357357503e-05,
+      "loss": 3.994,
+      "step": 184
+    },
+    {
+      "epoch": 0.03559232360155837,
+      "grad_norm": 4.106314995619256,
+      "learning_rate": 1.9998477555753054e-05,
+      "loss": 3.4964,
+      "step": 185
+    },
+    {
+      "epoch": 0.03578471453994517,
+      "grad_norm": 4.489960479563426,
+      "learning_rate": 1.9998366870855134e-05,
+      "loss": 3.5013,
+      "step": 186
+    },
+    {
+      "epoch": 0.035977105478331974,
+      "grad_norm": 4.2336701564164185,
+      "learning_rate": 1.999825230270673e-05,
+      "loss": 3.7058,
+      "step": 187
+    },
+    {
+      "epoch": 0.03616949641671877,
+      "grad_norm": 4.525571271166487,
+      "learning_rate": 1.9998133851352342e-05,
+      "loss": 3.5142,
+      "step": 188
+    },
+    {
+      "epoch": 0.03636188735510557,
+      "grad_norm": 4.737322716933247,
+      "learning_rate": 1.9998011516837974e-05,
+      "loss": 3.7005,
+      "step": 189
+    },
+    {
+      "epoch": 0.03655427829349238,
+      "grad_norm": 4.457219929570615,
+      "learning_rate": 1.999788529921114e-05,
+      "loss": 3.6005,
+      "step": 190
+    },
+    {
+      "epoch": 0.03674666923187918,
+      "grad_norm": 3.9482984645727015,
+      "learning_rate": 1.999775519852086e-05,
+      "loss": 3.4932,
+      "step": 191
+    },
+    {
+      "epoch": 0.03693906017026598,
+      "grad_norm": 4.183258862402467,
+      "learning_rate": 1.999762121481767e-05,
+      "loss": 3.629,
+      "step": 192
+    },
+    {
+      "epoch": 0.037131451108652784,
+      "grad_norm": 4.192376436293008,
+      "learning_rate": 1.99974833481536e-05,
+      "loss": 3.5087,
+      "step": 193
+    },
+    {
+      "epoch": 0.03732384204703958,
+      "grad_norm": 4.503868244360812,
+      "learning_rate": 1.9997341598582197e-05,
+      "loss": 3.4329,
+      "step": 194
+    },
+    {
+      "epoch": 0.03751623298542639,
+      "grad_norm": 3.9412644035284368,
+      "learning_rate": 1.9997195966158518e-05,
+      "loss": 3.6971,
+      "step": 195
+    },
+    {
+      "epoch": 0.03770862392381319,
+      "grad_norm": 4.380854194846852,
+      "learning_rate": 1.9997046450939122e-05,
+      "loss": 3.5388,
+      "step": 196
+    },
+    {
+      "epoch": 0.03790101486219999,
+      "grad_norm": 4.104484660214742,
+      "learning_rate": 1.9996893052982083e-05,
+      "loss": 3.2475,
+      "step": 197
+    },
+    {
+      "epoch": 0.038093405800586795,
+      "grad_norm": 4.395666546781883,
+      "learning_rate": 1.9996735772346973e-05,
+      "loss": 3.6052,
+      "step": 198
+    },
+    {
+      "epoch": 0.038285796738973595,
+      "grad_norm": 4.263777084363807,
+      "learning_rate": 1.9996574609094887e-05,
+      "loss": 3.7045,
+      "step": 199
+    },
+    {
+      "epoch": 0.038478187677360394,
+      "grad_norm": 4.2504378505319025,
+      "learning_rate": 1.9996409563288404e-05,
+      "loss": 3.294,
+      "step": 200
+    },
+    {
+      "epoch": 0.0386705786157472,
+      "grad_norm": 4.203287762966517,
+      "learning_rate": 1.9996240634991645e-05,
+      "loss": 3.5734,
+      "step": 201
+    },
+    {
+      "epoch": 0.038862969554134,
+      "grad_norm": 4.261989309814438,
+      "learning_rate": 1.9996067824270204e-05,
+      "loss": 3.6731,
+      "step": 202
+    },
+    {
+      "epoch": 0.0390553604925208,
+      "grad_norm": 4.3295041667486265,
+      "learning_rate": 1.999589113119121e-05,
+      "loss": 3.7544,
+      "step": 203
+    },
+    {
+      "epoch": 0.039247751430907606,
+      "grad_norm": 4.342413631327524,
+      "learning_rate": 1.9995710555823277e-05,
+      "loss": 3.4985,
+      "step": 204
+    },
+    {
+      "epoch": 0.039440142369294405,
+      "grad_norm": 4.351440172988684,
+      "learning_rate": 1.999552609823655e-05,
+      "loss": 3.502,
+      "step": 205
+    },
+    {
+      "epoch": 0.03963253330768121,
+      "grad_norm": 4.1616233683705675,
+      "learning_rate": 1.999533775850266e-05,
+      "loss": 3.6287,
+      "step": 206
+    },
+    {
+      "epoch": 0.03982492424606801,
+      "grad_norm": 3.2077268909193846,
+      "learning_rate": 1.9995145536694764e-05,
+      "loss": 3.1276,
+      "step": 207
+    },
+    {
+      "epoch": 0.04001731518445481,
+      "grad_norm": 4.198249141744509,
+      "learning_rate": 1.9994949432887512e-05,
+      "loss": 3.4165,
+      "step": 208
+    },
+    {
+      "epoch": 0.04020970612284162,
+      "grad_norm": 4.366089416602252,
+      "learning_rate": 1.999474944715708e-05,
+      "loss": 3.5885,
+      "step": 209
+    },
+    {
+      "epoch": 0.040402097061228416,
+      "grad_norm": 4.207057899586103,
+      "learning_rate": 1.9994545579581125e-05,
+      "loss": 3.7235,
+      "step": 210
+    },
+    {
+      "epoch": 0.040594487999615216,
+      "grad_norm": 4.472402126231881,
+      "learning_rate": 1.9994337830238836e-05,
+      "loss": 3.7793,
+      "step": 211
+    },
+    {
+      "epoch": 0.04078687893800202,
+      "grad_norm": 4.204895165936287,
+      "learning_rate": 1.9994126199210897e-05,
+      "loss": 3.497,
+      "step": 212
+    },
+    {
+      "epoch": 0.04097926987638882,
+      "grad_norm": 4.243799893261459,
+      "learning_rate": 1.999391068657951e-05,
+      "loss": 3.2779,
+      "step": 213
+    },
+    {
+      "epoch": 0.04117166081477562,
+      "grad_norm": 4.084975458362756,
+      "learning_rate": 1.9993691292428364e-05,
+      "loss": 3.6899,
+      "step": 214
+    },
+    {
+      "epoch": 0.04136405175316243,
+      "grad_norm": 4.239548947658577,
+      "learning_rate": 1.9993468016842684e-05,
+      "loss": 3.7086,
+      "step": 215
+    },
+    {
+      "epoch": 0.04155644269154923,
+      "grad_norm": 4.009983862130015,
+      "learning_rate": 1.999324085990918e-05,
+      "loss": 3.4471,
+      "step": 216
+    },
+    {
+      "epoch": 0.04174883362993603,
+      "grad_norm": 4.2390933713953745,
+      "learning_rate": 1.9993009821716076e-05,
+      "loss": 3.4456,
+      "step": 217
+    },
+    {
+      "epoch": 0.04194122456832283,
+      "grad_norm": 4.303694888436622,
+      "learning_rate": 1.9992774902353104e-05,
+      "loss": 3.8316,
+      "step": 218
+    },
+    {
+      "epoch": 0.04213361550670963,
+      "grad_norm": 4.674828621173306,
+      "learning_rate": 1.999253610191151e-05,
+      "loss": 3.5896,
+      "step": 219
+    },
+    {
+      "epoch": 0.04232600644509644,
+      "grad_norm": 4.194770734159858,
+      "learning_rate": 1.999229342048404e-05,
+      "loss": 3.484,
+      "step": 220
+    },
+    {
+      "epoch": 0.04251839738348324,
+      "grad_norm": 3.024659470540638,
+      "learning_rate": 1.9992046858164942e-05,
+      "loss": 3.0603,
+      "step": 221
+    },
+    {
+      "epoch": 0.04271078832187004,
+      "grad_norm": 4.278236819490771,
+      "learning_rate": 1.999179641504999e-05,
+      "loss": 3.5671,
+      "step": 222
+    },
+    {
+      "epoch": 0.042903179260256843,
+      "grad_norm": 4.332874277938807,
+      "learning_rate": 1.9991542091236438e-05,
+      "loss": 3.9978,
+      "step": 223
+    },
+    {
+      "epoch": 0.04309557019864364,
+      "grad_norm": 4.402005574682734,
+      "learning_rate": 1.9991283886823075e-05,
+      "loss": 3.4355,
+      "step": 224
+    },
+    {
+      "epoch": 0.04328796113703045,
+      "grad_norm": 3.9961918734928257,
+      "learning_rate": 1.9991021801910177e-05,
+      "loss": 3.5492,
+      "step": 225
+    },
+    {
+      "epoch": 0.04348035207541725,
+      "grad_norm": 3.927993459691407,
+      "learning_rate": 1.999075583659954e-05,
+      "loss": 3.5564,
+      "step": 226
+    },
+    {
+      "epoch": 0.04367274301380405,
+      "grad_norm": 4.112421969573917,
+      "learning_rate": 1.999048599099446e-05,
+      "loss": 3.5671,
+      "step": 227
+    },
+    {
+      "epoch": 0.043865133952190855,
+      "grad_norm": 4.2345901318581465,
+      "learning_rate": 1.9990212265199738e-05,
+      "loss": 3.7642,
+      "step": 228
+    },
+    {
+      "epoch": 0.044057524890577654,
+      "grad_norm": 4.169654127476836,
+      "learning_rate": 1.998993465932169e-05,
+      "loss": 3.496,
+      "step": 229
+    },
+    {
+      "epoch": 0.04424991582896445,
+      "grad_norm": 4.524255103591184,
+      "learning_rate": 1.9989653173468137e-05,
+      "loss": 3.5811,
+      "step": 230
+    },
+    {
+      "epoch": 0.04444230676735126,
+      "grad_norm": 4.420486032866371,
+      "learning_rate": 1.99893678077484e-05,
+      "loss": 3.6351,
+      "step": 231
+    },
+    {
+      "epoch": 0.04463469770573806,
+      "grad_norm": 4.18650496827198,
+      "learning_rate": 1.9989078562273313e-05,
+      "loss": 3.2907,
+      "step": 232
+    },
+    {
+      "epoch": 0.04482708864412486,
+      "grad_norm": 4.13774044044454,
+      "learning_rate": 1.9988785437155222e-05,
+      "loss": 3.3827,
+      "step": 233
+    },
+    {
+      "epoch": 0.045019479582511665,
+      "grad_norm": 4.1265229003669095,
+      "learning_rate": 1.9988488432507963e-05,
+      "loss": 3.2617,
+      "step": 234
+    },
+    {
+      "epoch": 0.045211870520898464,
+      "grad_norm": 4.321087667504246,
+      "learning_rate": 1.9988187548446895e-05,
+      "loss": 3.7033,
+      "step": 235
+    },
+    {
+      "epoch": 0.04540426145928527,
+      "grad_norm": 4.4390264098485,
+      "learning_rate": 1.998788278508888e-05,
+      "loss": 3.4489,
+      "step": 236
+    },
+    {
+      "epoch": 0.04559665239767207,
+      "grad_norm": 4.134783183540298,
+      "learning_rate": 1.9987574142552274e-05,
+      "loss": 3.5893,
+      "step": 237
+    },
+    {
+      "epoch": 0.04578904333605887,
+      "grad_norm": 4.049310442261922,
+      "learning_rate": 1.9987261620956964e-05,
+      "loss": 3.6126,
+      "step": 238
+    },
+    {
+      "epoch": 0.045981434274445676,
+      "grad_norm": 4.274940410593232,
+      "learning_rate": 1.9986945220424326e-05,
+      "loss": 3.4693,
+      "step": 239
+    },
+    {
+      "epoch": 0.046173825212832476,
+      "grad_norm": 4.3661840704022605,
+      "learning_rate": 1.998662494107724e-05,
+      "loss": 3.4833,
+      "step": 240
+    },
+    {
+      "epoch": 0.046366216151219275,
+      "grad_norm": 4.047921646320505,
+      "learning_rate": 1.99863007830401e-05,
+      "loss": 3.3906,
+      "step": 241
+    },
+    {
+      "epoch": 0.04655860708960608,
+      "grad_norm": 4.275174498163874,
+      "learning_rate": 1.9985972746438815e-05,
+      "loss": 3.779,
+      "step": 242
+    },
+    {
+      "epoch": 0.04675099802799288,
+      "grad_norm": 4.042290540650862,
+      "learning_rate": 1.9985640831400778e-05,
+      "loss": 3.4847,
+      "step": 243
+    },
+    {
+      "epoch": 0.04694338896637968,
+      "grad_norm": 4.1186453967384775,
+      "learning_rate": 1.998530503805491e-05,
+      "loss": 3.6872,
+      "step": 244
+    },
+    {
+      "epoch": 0.04713577990476649,
+      "grad_norm": 4.064253866309314,
+      "learning_rate": 1.9984965366531624e-05,
+      "loss": 3.6104,
+      "step": 245
+    },
+    {
+      "epoch": 0.047328170843153286,
+      "grad_norm": 4.438741446863669,
+      "learning_rate": 1.9984621816962843e-05,
+      "loss": 3.5947,
+      "step": 246
+    },
+    {
+      "epoch": 0.04752056178154009,
+      "grad_norm": 4.077813999683359,
+      "learning_rate": 1.9984274389482005e-05,
+      "loss": 3.2386,
+      "step": 247
+    },
+    {
+      "epoch": 0.04771295271992689,
+      "grad_norm": 2.7240188366728044,
+      "learning_rate": 1.9983923084224047e-05,
+      "loss": 3.0275,
+      "step": 248
+    },
+    {
+      "epoch": 0.04790534365831369,
+      "grad_norm": 4.037961146238751,
+      "learning_rate": 1.9983567901325404e-05,
+      "loss": 3.5935,
+      "step": 249
+    },
+    {
+      "epoch": 0.0480977345967005,
+      "grad_norm": 4.002894723619208,
+      "learning_rate": 1.9983208840924028e-05,
+      "loss": 3.7338,
+      "step": 250
+    },
+    {
+      "epoch": 0.0482901255350873,
+      "grad_norm": 4.166838230969168,
+      "learning_rate": 1.998284590315937e-05,
+      "loss": 3.6228,
+      "step": 251
+    },
+    {
+      "epoch": 0.0484825164734741,
+      "grad_norm": 3.8731563734198486,
+      "learning_rate": 1.9982479088172403e-05,
+      "loss": 3.4162,
+      "step": 252
+    },
+    {
+      "epoch": 0.0486749074118609,
+      "grad_norm": 2.781721926243559,
+      "learning_rate": 1.9982108396105584e-05,
+      "loss": 3.1438,
+      "step": 253
+    },
+    {
+      "epoch": 0.0488672983502477,
+      "grad_norm": 4.123085605203631,
+      "learning_rate": 1.9981733827102884e-05,
+      "loss": 3.6474,
+      "step": 254
+    },
+    {
+      "epoch": 0.04905968928863451,
+      "grad_norm": 4.113696433885685,
+      "learning_rate": 1.998135538130979e-05,
+      "loss": 3.761,
+      "step": 255
+    },
+    {
+      "epoch": 0.04925208022702131,
+      "grad_norm": 2.751216498161506,
+      "learning_rate": 1.998097305887328e-05,
+      "loss": 3.0778,
+      "step": 256
+    },
+    {
+      "epoch": 0.04944447116540811,
+      "grad_norm": 3.8881351707154588,
+      "learning_rate": 1.9980586859941846e-05,
+      "loss": 3.0616,
+      "step": 257
+    },
+    {
+      "epoch": 0.049636862103794914,
+      "grad_norm": 2.5645914515465718,
+      "learning_rate": 1.998019678466548e-05,
+      "loss": 2.7314,
+      "step": 258
+    },
+    {
+      "epoch": 0.04982925304218171,
+      "grad_norm": 4.177974292304787,
+      "learning_rate": 1.997980283319568e-05,
+      "loss": 3.4044,
+      "step": 259
+    },
+    {
+      "epoch": 0.05002164398056851,
+      "grad_norm": 4.1969485423507855,
+      "learning_rate": 1.9979405005685466e-05,
+      "loss": 3.6449,
+      "step": 260
+    },
+    {
+      "epoch": 0.05021403491895532,
+      "grad_norm": 4.054246025657358,
+      "learning_rate": 1.9979003302289336e-05,
+      "loss": 3.5704,
+      "step": 261
+    },
+    {
+      "epoch": 0.05040642585734212,
+      "grad_norm": 4.232759967962379,
+      "learning_rate": 1.997859772316331e-05,
+      "loss": 3.4643,
+      "step": 262
+    },
+    {
+      "epoch": 0.05059881679572892,
+      "grad_norm": 4.442221549650081,
+      "learning_rate": 1.9978188268464912e-05,
+      "loss": 3.4338,
+      "step": 263
+    },
+    {
+      "epoch": 0.050791207734115724,
+      "grad_norm": 3.75698587325541,
+      "learning_rate": 1.997777493835317e-05,
+      "loss": 3.3792,
+      "step": 264
+    },
+    {
+      "epoch": 0.050983598672502524,
+      "grad_norm": 3.91284700522455,
+      "learning_rate": 1.9977357732988616e-05,
+      "loss": 3.6092,
+      "step": 265
+    },
+    {
+      "epoch": 0.05117598961088933,
+      "grad_norm": 4.133101080666413,
+      "learning_rate": 1.9976936652533288e-05,
+      "loss": 3.3001,
+      "step": 266
+    },
+    {
+      "epoch": 0.05136838054927613,
+      "grad_norm": 4.162148394371124,
+      "learning_rate": 1.997651169715073e-05,
+      "loss": 3.6693,
+      "step": 267
+    },
+    {
+      "epoch": 0.05156077148766293,
+      "grad_norm": 3.9853521772366287,
+      "learning_rate": 1.9976082867005985e-05,
+      "loss": 3.1807,
+      "step": 268
+    },
+    {
+      "epoch": 0.051753162426049736,
+      "grad_norm": 4.640624405928715,
+      "learning_rate": 1.997565016226561e-05,
+      "loss": 3.4972,
+      "step": 269
+    },
+    {
+      "epoch": 0.051945553364436535,
+      "grad_norm": 4.078188539657672,
+      "learning_rate": 1.997521358309766e-05,
+      "loss": 3.5764,
+      "step": 270
+    },
+    {
+      "epoch": 0.052137944302823334,
+      "grad_norm": 3.8394182013907203,
+      "learning_rate": 1.99747731296717e-05,
+      "loss": 3.3129,
+      "step": 271
+    },
+    {
+      "epoch": 0.05233033524121014,
+      "grad_norm": 4.377678009856555,
+      "learning_rate": 1.9974328802158798e-05,
+      "loss": 3.3386,
+      "step": 272
+    },
+    {
+      "epoch": 0.05252272617959694,
+      "grad_norm": 3.908903147228433,
+      "learning_rate": 1.997388060073152e-05,
+      "loss": 3.7049,
+      "step": 273
+    },
+    {
+      "epoch": 0.05271511711798374,
+      "grad_norm": 3.886618161517831,
+      "learning_rate": 1.9973428525563948e-05,
+      "loss": 3.5823,
+      "step": 274
+    },
+    {
+      "epoch": 0.052907508056370546,
+      "grad_norm": 4.301495601300363,
+      "learning_rate": 1.9972972576831656e-05,
+      "loss": 3.3992,
+      "step": 275
+    },
+    {
+      "epoch": 0.053099898994757345,
+      "grad_norm": 4.156942906438688,
+      "learning_rate": 1.9972512754711738e-05,
+      "loss": 3.6127,
+      "step": 276
+    },
+    {
+      "epoch": 0.05329228993314415,
+      "grad_norm": 3.7431194411552333,
+      "learning_rate": 1.997204905938278e-05,
+      "loss": 3.5118,
+      "step": 277
+    },
+    {
+      "epoch": 0.05348468087153095,
+      "grad_norm": 4.160217891371853,
+      "learning_rate": 1.9971581491024873e-05,
+      "loss": 3.4062,
+      "step": 278
+    },
+    {
+      "epoch": 0.05367707180991775,
+      "grad_norm": 4.2660099905990245,
+      "learning_rate": 1.997111004981962e-05,
+      "loss": 3.3616,
+      "step": 279
+    },
+    {
+      "epoch": 0.05386946274830456,
+      "grad_norm": 4.115648425635137,
+      "learning_rate": 1.9970634735950117e-05,
+      "loss": 3.6158,
+      "step": 280
+    },
+    {
+      "epoch": 0.05406185368669136,
+      "grad_norm": 3.78170537196026,
+      "learning_rate": 1.9970155549600978e-05,
+      "loss": 3.2782,
+      "step": 281
+    },
+    {
+      "epoch": 0.054254244625078156,
+      "grad_norm": 4.245489890058941,
+      "learning_rate": 1.9969672490958304e-05,
+      "loss": 3.5282,
+      "step": 282
+    },
+    {
+      "epoch": 0.05444663556346496,
+      "grad_norm": 4.236593144831113,
+      "learning_rate": 1.996918556020972e-05,
+      "loss": 3.5694,
+      "step": 283
+    },
+    {
+      "epoch": 0.05463902650185176,
+      "grad_norm": 4.042203541344946,
+      "learning_rate": 1.996869475754434e-05,
+      "loss": 3.3191,
+      "step": 284
+    },
+    {
+      "epoch": 0.05483141744023857,
+      "grad_norm": 3.906122373625533,
+      "learning_rate": 1.9968200083152784e-05,
+      "loss": 3.3541,
+      "step": 285
+    },
+    {
+      "epoch": 0.05502380837862537,
+      "grad_norm": 4.030550217309951,
+      "learning_rate": 1.9967701537227175e-05,
+      "loss": 3.5112,
+      "step": 286
+    },
+    {
+      "epoch": 0.05521619931701217,
+      "grad_norm": 2.6254955706152447,
+      "learning_rate": 1.996719911996115e-05,
+      "loss": 2.6904,
+      "step": 287
+    },
+    {
+      "epoch": 0.05540859025539897,
+      "grad_norm": 4.1587734720334755,
+      "learning_rate": 1.996669283154984e-05,
+      "loss": 3.6566,
+      "step": 288
+    },
+    {
+      "epoch": 0.05560098119378577,
+      "grad_norm": 2.7029014850496327,
+      "learning_rate": 1.996618267218988e-05,
+      "loss": 2.9971,
+      "step": 289
+    },
+    {
+      "epoch": 0.05579337213217257,
+      "grad_norm": 4.084588047553907,
+      "learning_rate": 1.996566864207941e-05,
+      "loss": 3.4672,
+      "step": 290
+    },
+    {
+      "epoch": 0.05598576307055938,
+      "grad_norm": 4.134957288700149,
+      "learning_rate": 1.9965150741418072e-05,
+      "loss": 3.6236,
+      "step": 291
+    },
+    {
+      "epoch": 0.05617815400894618,
+      "grad_norm": 4.320327148963083,
+      "learning_rate": 1.9964628970407018e-05,
+      "loss": 3.5149,
+      "step": 292
+    },
+    {
+      "epoch": 0.05637054494733298,
+      "grad_norm": 3.918341635143802,
+      "learning_rate": 1.9964103329248892e-05,
+      "loss": 3.7625,
+      "step": 293
+    },
+    {
+      "epoch": 0.056562935885719784,
+      "grad_norm": 4.220818022867905,
+      "learning_rate": 1.996357381814785e-05,
+      "loss": 3.7018,
+      "step": 294
+    },
+    {
+      "epoch": 0.05675532682410658,
+      "grad_norm": 3.7615150301996754,
+      "learning_rate": 1.996304043730955e-05,
+      "loss": 3.5287,
+      "step": 295
+    },
+    {
+      "epoch": 0.05694771776249339,
+      "grad_norm": 3.9650574093313047,
+      "learning_rate": 1.9962503186941143e-05,
+      "loss": 3.2487,
+      "step": 296
+    },
+    {
+      "epoch": 0.05714010870088019,
+      "grad_norm": 4.139879385677748,
+      "learning_rate": 1.9961962067251298e-05,
+      "loss": 3.4934,
+      "step": 297
+    },
+    {
+      "epoch": 0.05733249963926699,
+      "grad_norm": 4.004431350466355,
+      "learning_rate": 1.9961417078450177e-05,
+      "loss": 3.4066,
+      "step": 298
+    },
+    {
+      "epoch": 0.057524890577653795,
+      "grad_norm": 3.8125305548471866,
+      "learning_rate": 1.996086822074945e-05,
+      "loss": 3.5998,
+      "step": 299
+    },
+    {
+      "epoch": 0.057717281516040594,
+      "grad_norm": 3.5682261495121006,
+      "learning_rate": 1.9960315494362286e-05,
+      "loss": 3.3609,
+      "step": 300
+    },
+    {
+      "epoch": 0.057909672454427394,
+      "grad_norm": 4.024799745330692,
+      "learning_rate": 1.9959758899503355e-05,
+      "loss": 3.3995,
+      "step": 301
+    },
+    {
+      "epoch": 0.0581020633928142,
+      "grad_norm": 4.144508158539394,
+      "learning_rate": 1.995919843638883e-05,
+      "loss": 3.5204,
+      "step": 302
+    },
+    {
+      "epoch": 0.058294454331201,
+      "grad_norm": 4.5038594089058055,
+      "learning_rate": 1.9958634105236395e-05,
+      "loss": 3.665,
+      "step": 303
+    },
+    {
+      "epoch": 0.0584868452695878,
+      "grad_norm": 4.069450134473167,
+      "learning_rate": 1.9958065906265228e-05,
+      "loss": 3.4292,
+      "step": 304
+    },
+    {
+      "epoch": 0.058679236207974605,
+      "grad_norm": 4.03680654159872,
+      "learning_rate": 1.9957493839696013e-05,
+      "loss": 3.5665,
+      "step": 305
+    },
+    {
+      "epoch": 0.058871627146361405,
+      "grad_norm": 4.181964173035944,
+      "learning_rate": 1.9956917905750926e-05,
+      "loss": 3.5384,
+      "step": 306
+    },
+    {
+      "epoch": 0.05906401808474821,
+      "grad_norm": 4.2733946928240805,
+      "learning_rate": 1.995633810465366e-05,
+      "loss": 3.5495,
+      "step": 307
+    },
+    {
+      "epoch": 0.05925640902313501,
+      "grad_norm": 3.7000718184435053,
+      "learning_rate": 1.99557544366294e-05,
+      "loss": 3.5097,
+      "step": 308
+    },
+    {
+      "epoch": 0.05944879996152181,
+      "grad_norm": 4.060647555139414,
+      "learning_rate": 1.9955166901904838e-05,
+      "loss": 3.4916,
+      "step": 309
+    },
+    {
+      "epoch": 0.05964119089990862,
+      "grad_norm": 4.193088006983839,
+      "learning_rate": 1.9954575500708164e-05,
+      "loss": 3.8344,
+      "step": 310
+    },
+    {
+      "epoch": 0.059833581838295416,
+      "grad_norm": 4.1875860765633,
+      "learning_rate": 1.995398023326907e-05,
+      "loss": 3.7057,
+      "step": 311
+    },
+    {
+      "epoch": 0.060025972776682215,
+      "grad_norm": 2.777495685086682,
+      "learning_rate": 1.9953381099818756e-05,
+      "loss": 2.9028,
+      "step": 312
+    },
+    {
+      "epoch": 0.06021836371506902,
+      "grad_norm": 3.771548179568147,
+      "learning_rate": 1.9952778100589912e-05,
+      "loss": 3.2566,
+      "step": 313
+    },
+    {
+      "epoch": 0.06041075465345582,
+      "grad_norm": 4.576210937708079,
+      "learning_rate": 1.9952171235816747e-05,
+      "loss": 3.509,
+      "step": 314
+    },
+    {
+      "epoch": 0.06060314559184263,
+      "grad_norm": 3.99022528439649,
+      "learning_rate": 1.9951560505734948e-05,
+      "loss": 3.5081,
+      "step": 315
+    },
+    {
+      "epoch": 0.06079553653022943,
+      "grad_norm": 4.043638337785179,
+      "learning_rate": 1.9950945910581718e-05,
+      "loss": 3.4782,
+      "step": 316
+    },
+    {
+      "epoch": 0.060987927468616226,
+      "grad_norm": 3.7996631406497388,
+      "learning_rate": 1.9950327450595766e-05,
+      "loss": 2.984,
+      "step": 317
+    },
+    {
+      "epoch": 0.06118031840700303,
+      "grad_norm": 3.6639461180919968,
+      "learning_rate": 1.9949705126017286e-05,
+      "loss": 3.4034,
+      "step": 318
+    },
+    {
+      "epoch": 0.06137270934538983,
+      "grad_norm": 4.316787496596526,
+      "learning_rate": 1.9949078937087988e-05,
+      "loss": 3.6822,
+      "step": 319
+    },
+    {
+      "epoch": 0.06156510028377663,
+      "grad_norm": 4.442925687680224,
+      "learning_rate": 1.994844888405107e-05,
+      "loss": 3.6517,
+      "step": 320
+    },
+    {
+      "epoch": 0.06175749122216344,
+      "grad_norm": 4.267793422577924,
+      "learning_rate": 1.9947814967151246e-05,
+      "loss": 3.6106,
+      "step": 321
+    },
+    {
+      "epoch": 0.06194988216055024,
+      "grad_norm": 3.9243382252149033,
+      "learning_rate": 1.9947177186634716e-05,
+      "loss": 3.2789,
+      "step": 322
+    },
+    {
+      "epoch": 0.06214227309893704,
+      "grad_norm": 4.113970340685003,
+      "learning_rate": 1.9946535542749187e-05,
+      "loss": 3.5643,
+      "step": 323
+    },
+    {
+      "epoch": 0.06233466403732384,
+      "grad_norm": 2.818273329109669,
+      "learning_rate": 1.9945890035743866e-05,
+      "loss": 3.1084,
+      "step": 324
+    },
+    {
+      "epoch": 0.06252705497571065,
+      "grad_norm": 2.4288960933920323,
+      "learning_rate": 1.9945240665869465e-05,
+      "loss": 2.9438,
+      "step": 325
+    },
+    {
+      "epoch": 0.06271944591409745,
+      "grad_norm": 4.197447804708069,
+      "learning_rate": 1.9944587433378187e-05,
+      "loss": 3.3359,
+      "step": 326
+    },
+    {
+      "epoch": 0.06291183685248425,
+      "grad_norm": 3.795804619025396,
+      "learning_rate": 1.994393033852374e-05,
+      "loss": 3.1017,
+      "step": 327
+    },
+    {
+      "epoch": 0.06310422779087105,
+      "grad_norm": 3.9791280295535607,
+      "learning_rate": 1.9943269381561334e-05,
+      "loss": 3.6698,
+      "step": 328
+    },
+    {
+      "epoch": 0.06329661872925785,
+      "grad_norm": 4.120995378659135,
+      "learning_rate": 1.994260456274768e-05,
+      "loss": 3.5482,
+      "step": 329
+    },
+    {
+      "epoch": 0.06348900966764466,
+      "grad_norm": 4.198244825977875,
+      "learning_rate": 1.9941935882340976e-05,
+      "loss": 3.5511,
+      "step": 330
+    },
+    {
+      "epoch": 0.06368140060603146,
+      "grad_norm": 4.183336401311049,
+      "learning_rate": 1.994126334060094e-05,
+      "loss": 3.2305,
+      "step": 331
+    },
+    {
+      "epoch": 0.06387379154441826,
+      "grad_norm": 4.3067715686446375,
+      "learning_rate": 1.994058693778878e-05,
+      "loss": 3.5361,
+      "step": 332
+    },
+    {
+      "epoch": 0.06406618248280506,
+      "grad_norm": 4.247818635590897,
+      "learning_rate": 1.9939906674167192e-05,
+      "loss": 3.5425,
+      "step": 333
+    },
+    {
+      "epoch": 0.06425857342119186,
+      "grad_norm": 3.9945701926905235,
+      "learning_rate": 1.993922255000039e-05,
+      "loss": 3.3971,
+      "step": 334
+    },
+    {
+      "epoch": 0.06445096435957866,
+      "grad_norm": 3.775280527558848,
+      "learning_rate": 1.993853456555408e-05,
+      "loss": 3.2704,
+      "step": 335
+    },
+    {
+      "epoch": 0.06464335529796547,
+      "grad_norm": 3.966092894655815,
+      "learning_rate": 1.9937842721095468e-05,
+      "loss": 3.59,
+      "step": 336
+    },
+    {
+      "epoch": 0.06483574623635227,
+      "grad_norm": 2.475907114144814,
+      "learning_rate": 1.9937147016893257e-05,
+      "loss": 2.9548,
+      "step": 337
+    },
+    {
+      "epoch": 0.06502813717473907,
+      "grad_norm": 4.27737332442169,
+      "learning_rate": 1.9936447453217646e-05,
+      "loss": 3.4278,
+      "step": 338
+    },
+    {
+      "epoch": 0.06522052811312587,
+      "grad_norm": 4.126141183561587,
+      "learning_rate": 1.9935744030340347e-05,
+      "loss": 3.2947,
+      "step": 339
+    },
+    {
+      "epoch": 0.06541291905151267,
+      "grad_norm": 4.247467324190018,
+      "learning_rate": 1.9935036748534555e-05,
+      "loss": 3.7221,
+      "step": 340
+    },
+    {
+      "epoch": 0.06560530998989948,
+      "grad_norm": 4.273861053234695,
+      "learning_rate": 1.993432560807497e-05,
+      "loss": 3.6894,
+      "step": 341
+    },
+    {
+      "epoch": 0.06579770092828628,
+      "grad_norm": 4.014562654430113,
+      "learning_rate": 1.993361060923779e-05,
+      "loss": 3.4778,
+      "step": 342
+    },
+    {
+      "epoch": 0.06599009186667308,
+      "grad_norm": 4.200374466878388,
+      "learning_rate": 1.9932891752300717e-05,
+      "loss": 3.8486,
+      "step": 343
+    },
+    {
+      "epoch": 0.06618248280505988,
+      "grad_norm": 4.04995469435432,
+      "learning_rate": 1.9932169037542947e-05,
+      "loss": 3.5626,
+      "step": 344
+    },
+    {
+      "epoch": 0.06637487374344668,
+      "grad_norm": 3.92772398020147,
+      "learning_rate": 1.9931442465245164e-05,
+      "loss": 3.6307,
+      "step": 345
+    },
+    {
+      "epoch": 0.06656726468183348,
+      "grad_norm": 3.9156243232556642,
+      "learning_rate": 1.9930712035689576e-05,
+      "loss": 3.5048,
+      "step": 346
+    },
+    {
+      "epoch": 0.06675965562022029,
+      "grad_norm": 4.13954807263478,
+      "learning_rate": 1.992997774915986e-05,
+      "loss": 3.3753,
+      "step": 347
+    },
+    {
+      "epoch": 0.06695204655860709,
+      "grad_norm": 2.5546218608255105,
+      "learning_rate": 1.992923960594121e-05,
+      "loss": 2.9452,
+      "step": 348
+    },
+    {
+      "epoch": 0.06714443749699389,
+      "grad_norm": 2.451887454350923,
+      "learning_rate": 1.9928497606320308e-05,
+      "loss": 2.8187,
+      "step": 349
+    },
+    {
+      "epoch": 0.06733682843538069,
+      "grad_norm": 3.960704736773125,
+      "learning_rate": 1.992775175058535e-05,
+      "loss": 3.3344,
+      "step": 350
+    },
+    {
+      "epoch": 0.06752921937376749,
+      "grad_norm": 4.349302276285121,
+      "learning_rate": 1.9927002039026002e-05,
+      "loss": 3.2852,
+      "step": 351
+    },
+    {
+      "epoch": 0.0677216103121543,
+      "grad_norm": 4.305626604842038,
+      "learning_rate": 1.9926248471933453e-05,
+      "loss": 3.6201,
+      "step": 352
+    },
+    {
+      "epoch": 0.0679140012505411,
+      "grad_norm": 4.489115405890942,
+      "learning_rate": 1.9925491049600382e-05,
+      "loss": 3.6227,
+      "step": 353
+    },
+    {
+      "epoch": 0.0681063921889279,
+      "grad_norm": 3.6523577585345937,
+      "learning_rate": 1.9924729772320953e-05,
+      "loss": 3.4387,
+      "step": 354
+    },
+    {
+      "epoch": 0.0682987831273147,
+      "grad_norm": 4.376569036142267,
+      "learning_rate": 1.9923964640390846e-05,
+      "loss": 3.6149,
+      "step": 355
+    },
+    {
+      "epoch": 0.0684911740657015,
+      "grad_norm": 3.8284334722448063,
+      "learning_rate": 1.9923195654107227e-05,
+      "loss": 3.6052,
+      "step": 356
+    },
+    {
+      "epoch": 0.0686835650040883,
+      "grad_norm": 4.210527122974296,
+      "learning_rate": 1.992242281376876e-05,
+      "loss": 3.6576,
+      "step": 357
+    },
+    {
+      "epoch": 0.06887595594247511,
+      "grad_norm": 4.071881225242726,
+      "learning_rate": 1.9921646119675606e-05,
+      "loss": 3.5367,
+      "step": 358
+    },
+    {
+      "epoch": 0.06906834688086191,
+      "grad_norm": 4.065318453341959,
+      "learning_rate": 1.9920865572129426e-05,
+      "loss": 3.5265,
+      "step": 359
+    },
+    {
+      "epoch": 0.06926073781924871,
+      "grad_norm": 4.260198819236625,
+      "learning_rate": 1.9920081171433377e-05,
+      "loss": 3.7023,
+      "step": 360
+    },
+    {
+      "epoch": 0.06945312875763551,
+      "grad_norm": 3.8047302990893845,
+      "learning_rate": 1.991929291789211e-05,
+      "loss": 3.3524,
+      "step": 361
+    },
+    {
+      "epoch": 0.06964551969602231,
+      "grad_norm": 3.917793510212044,
+      "learning_rate": 1.9918500811811778e-05,
+      "loss": 3.2886,
+      "step": 362
+    },
+    {
+      "epoch": 0.06983791063440913,
+      "grad_norm": 4.256154927720927,
+      "learning_rate": 1.991770485350002e-05,
+      "loss": 3.6384,
+      "step": 363
+    },
+    {
+      "epoch": 0.07003030157279592,
+      "grad_norm": 2.511506206055887,
+      "learning_rate": 1.991690504326597e-05,
+      "loss": 2.9049,
+      "step": 364
+    },
+    {
+      "epoch": 0.07022269251118272,
+      "grad_norm": 4.097120078786226,
+      "learning_rate": 1.9916101381420285e-05,
+      "loss": 3.35,
+      "step": 365
+    },
+    {
+      "epoch": 0.07041508344956952,
+      "grad_norm": 4.435319115251024,
+      "learning_rate": 1.9915293868275083e-05,
+      "loss": 3.1834,
+      "step": 366
+    },
+    {
+      "epoch": 0.07060747438795632,
+      "grad_norm": 2.5320760769012742,
+      "learning_rate": 1.9914482504143996e-05,
+      "loss": 2.9228,
+      "step": 367
+    },
+    {
+      "epoch": 0.07079986532634312,
+      "grad_norm": 3.824562482655438,
+      "learning_rate": 1.9913667289342147e-05,
+      "loss": 3.1923,
+      "step": 368
+    },
+    {
+      "epoch": 0.07099225626472994,
+      "grad_norm": 3.989588676707961,
+      "learning_rate": 1.991284822418616e-05,
+      "loss": 3.3852,
+      "step": 369
+    },
+    {
+      "epoch": 0.07118464720311674,
+      "grad_norm": 4.019970743198744,
+      "learning_rate": 1.9912025308994146e-05,
+      "loss": 3.5934,
+      "step": 370
+    },
+    {
+      "epoch": 0.07137703814150353,
+      "grad_norm": 3.682960052052758,
+      "learning_rate": 1.9911198544085723e-05,
+      "loss": 3.1773,
+      "step": 371
+    },
+    {
+      "epoch": 0.07156942907989033,
+      "grad_norm": 2.5338865198952703,
+      "learning_rate": 1.991036792978199e-05,
+      "loss": 2.8947,
+      "step": 372
+    },
+    {
+      "epoch": 0.07176182001827713,
+      "grad_norm": 3.9909402714350084,
+      "learning_rate": 1.990953346640555e-05,
+      "loss": 3.7834,
+      "step": 373
+    },
+    {
+      "epoch": 0.07195421095666395,
+      "grad_norm": 3.7803375506786208,
+      "learning_rate": 1.9908695154280496e-05,
+      "loss": 3.5038,
+      "step": 374
+    },
+    {
+      "epoch": 0.07214660189505075,
+      "grad_norm": 4.362541473375203,
+      "learning_rate": 1.9907852993732425e-05,
+      "loss": 3.5633,
+      "step": 375
+    },
+    {
+      "epoch": 0.07233899283343755,
+      "grad_norm": 4.0911093221708335,
+      "learning_rate": 1.990700698508842e-05,
+      "loss": 3.6687,
+      "step": 376
+    },
+    {
+      "epoch": 0.07253138377182435,
+      "grad_norm": 3.980854721542477,
+      "learning_rate": 1.990615712867706e-05,
+      "loss": 3.3093,
+      "step": 377
+    },
+    {
+      "epoch": 0.07272377471021114,
+      "grad_norm": 3.7839099725134933,
+      "learning_rate": 1.9905303424828418e-05,
+      "loss": 3.3661,
+      "step": 378
+    },
+    {
+      "epoch": 0.07291616564859794,
+      "grad_norm": 4.136060335127341,
+      "learning_rate": 1.9904445873874068e-05,
+      "loss": 3.7081,
+      "step": 379
+    },
+    {
+      "epoch": 0.07310855658698476,
+      "grad_norm": 4.000534739992824,
+      "learning_rate": 1.9903584476147066e-05,
+      "loss": 3.5411,
+      "step": 380
+    },
+    {
+      "epoch": 0.07330094752537156,
+      "grad_norm": 4.400267969288538,
+      "learning_rate": 1.9902719231981975e-05,
+      "loss": 3.5979,
+      "step": 381
+    },
+    {
+      "epoch": 0.07349333846375836,
+      "grad_norm": 4.3320551177932405,
+      "learning_rate": 1.9901850141714843e-05,
+      "loss": 3.4726,
+      "step": 382
+    },
+    {
+      "epoch": 0.07368572940214516,
+      "grad_norm": 4.184961269877884,
+      "learning_rate": 1.9900977205683213e-05,
+      "loss": 3.6865,
+      "step": 383
+    },
+    {
+      "epoch": 0.07387812034053196,
+      "grad_norm": 2.5818809746868694,
+      "learning_rate": 1.9900100424226124e-05,
+      "loss": 2.8352,
+      "step": 384
+    },
+    {
+      "epoch": 0.07407051127891877,
+      "grad_norm": 3.898568016479355,
+      "learning_rate": 1.9899219797684113e-05,
+      "loss": 3.2721,
+      "step": 385
+    },
+    {
+      "epoch": 0.07426290221730557,
+      "grad_norm": 3.9099796625715837,
+      "learning_rate": 1.98983353263992e-05,
+      "loss": 3.2911,
+      "step": 386
+    },
+    {
+      "epoch": 0.07445529315569237,
+      "grad_norm": 3.910806524792175,
+      "learning_rate": 1.9897447010714905e-05,
+      "loss": 3.649,
+      "step": 387
+    },
+    {
+      "epoch": 0.07464768409407917,
+      "grad_norm": 3.9207300520991297,
+      "learning_rate": 1.989655485097624e-05,
+      "loss": 3.4928,
+      "step": 388
+    },
+    {
+      "epoch": 0.07484007503246597,
+      "grad_norm": 4.078108684587361,
+      "learning_rate": 1.989565884752971e-05,
+      "loss": 3.6028,
+      "step": 389
+    },
+    {
+      "epoch": 0.07503246597085278,
+      "grad_norm": 3.9253285075455175,
+      "learning_rate": 1.9894759000723308e-05,
+      "loss": 3.209,
+      "step": 390
+    },
+    {
+      "epoch": 0.07522485690923958,
+      "grad_norm": 3.690873103470385,
+      "learning_rate": 1.9893855310906526e-05,
+      "loss": 3.3524,
+      "step": 391
+    },
+    {
+      "epoch": 0.07541724784762638,
+      "grad_norm": 4.699138259958471,
+      "learning_rate": 1.9892947778430352e-05,
+      "loss": 3.4169,
+      "step": 392
+    },
+    {
+      "epoch": 0.07560963878601318,
+      "grad_norm": 3.823515218591656,
+      "learning_rate": 1.9892036403647256e-05,
+      "loss": 3.4246,
+      "step": 393
+    },
+    {
+      "epoch": 0.07580202972439998,
+      "grad_norm": 3.898286922231221,
+      "learning_rate": 1.9891121186911207e-05,
+      "loss": 3.4286,
+      "step": 394
+    },
+    {
+      "epoch": 0.07599442066278678,
+      "grad_norm": 3.7144542993671132,
+      "learning_rate": 1.9890202128577664e-05,
+      "loss": 3.5955,
+      "step": 395
+    },
+    {
+      "epoch": 0.07618681160117359,
+      "grad_norm": 3.9836407961354103,
+      "learning_rate": 1.988927922900358e-05,
+      "loss": 3.5311,
+      "step": 396
+    },
+    {
+      "epoch": 0.07637920253956039,
+      "grad_norm": 4.340838670912758,
+      "learning_rate": 1.9888352488547397e-05,
+      "loss": 3.3253,
+      "step": 397
+    },
+    {
+      "epoch": 0.07657159347794719,
+      "grad_norm": 3.7978046137423576,
+      "learning_rate": 1.988742190756905e-05,
+      "loss": 3.6639,
+      "step": 398
+    },
+    {
+      "epoch": 0.07676398441633399,
+      "grad_norm": 3.9014858960985226,
+      "learning_rate": 1.9886487486429966e-05,
+      "loss": 3.6927,
+      "step": 399
+    },
+    {
+      "epoch": 0.07695637535472079,
+      "grad_norm": 4.07872582667308,
+      "learning_rate": 1.9885549225493064e-05,
+      "loss": 3.4303,
+      "step": 400
+    },
+    {
+      "epoch": 0.0771487662931076,
+      "grad_norm": 3.9015856181317337,
+      "learning_rate": 1.9884607125122753e-05,
+      "loss": 3.5125,
+      "step": 401
+    },
+    {
+      "epoch": 0.0773411572314944,
+      "grad_norm": 4.142794192832855,
+      "learning_rate": 1.988366118568494e-05,
+      "loss": 3.4668,
+      "step": 402
+    },
+    {
+      "epoch": 0.0775335481698812,
+      "grad_norm": 3.7797187648839654,
+      "learning_rate": 1.988271140754701e-05,
+      "loss": 3.3679,
+      "step": 403
+    },
+    {
+      "epoch": 0.077725939108268,
+      "grad_norm": 3.9925780783593674,
+      "learning_rate": 1.9881757791077848e-05,
+      "loss": 3.6466,
+      "step": 404
+    },
+    {
+      "epoch": 0.0779183300466548,
+      "grad_norm": 3.8643625506677712,
+      "learning_rate": 1.9880800336647825e-05,
+      "loss": 3.292,
+      "step": 405
+    },
+    {
+      "epoch": 0.0781107209850416,
+      "grad_norm": 4.209963894511985,
+      "learning_rate": 1.987983904462881e-05,
+      "loss": 3.7363,
+      "step": 406
+    },
+    {
+      "epoch": 0.07830311192342841,
+      "grad_norm": 4.117148779616679,
+      "learning_rate": 1.9878873915394154e-05,
+      "loss": 3.5243,
+      "step": 407
+    },
+    {
+      "epoch": 0.07849550286181521,
+      "grad_norm": 3.9209242816258003,
+      "learning_rate": 1.9877904949318704e-05,
+      "loss": 3.1608,
+      "step": 408
+    },
+    {
+      "epoch": 0.07868789380020201,
+      "grad_norm": 3.734700362445911,
+      "learning_rate": 1.9876932146778796e-05,
+      "loss": 3.4566,
+      "step": 409
+    },
+    {
+      "epoch": 0.07888028473858881,
+      "grad_norm": 4.4185830135767,
+      "learning_rate": 1.9875955508152254e-05,
+      "loss": 3.4592,
+      "step": 410
+    },
+    {
+      "epoch": 0.07907267567697561,
+      "grad_norm": 4.035613384222432,
+      "learning_rate": 1.987497503381839e-05,
+      "loss": 3.5922,
+      "step": 411
+    },
+    {
+      "epoch": 0.07926506661536242,
+      "grad_norm": 3.765514322867427,
+      "learning_rate": 1.9873990724158014e-05,
+      "loss": 3.312,
+      "step": 412
+    },
+    {
+      "epoch": 0.07945745755374922,
+      "grad_norm": 4.036714578449483,
+      "learning_rate": 1.987300257955342e-05,
+      "loss": 3.578,
+      "step": 413
+    },
+    {
+      "epoch": 0.07964984849213602,
+      "grad_norm": 4.3090039449158235,
+      "learning_rate": 1.987201060038839e-05,
+      "loss": 3.5415,
+      "step": 414
+    },
+    {
+      "epoch": 0.07984223943052282,
+      "grad_norm": 4.1175483720077395,
+      "learning_rate": 1.9871014787048197e-05,
+      "loss": 3.407,
+      "step": 415
+    },
+    {
+      "epoch": 0.08003463036890962,
+      "grad_norm": 4.027616328485031,
+      "learning_rate": 1.9870015139919606e-05,
+      "loss": 3.5152,
+      "step": 416
+    },
+    {
+      "epoch": 0.08022702130729642,
+      "grad_norm": 4.0830156024143855,
+      "learning_rate": 1.9869011659390866e-05,
+      "loss": 3.4349,
+      "step": 417
+    },
+    {
+      "epoch": 0.08041941224568323,
+      "grad_norm": 3.7637622650504965,
+      "learning_rate": 1.9868004345851716e-05,
+      "loss": 3.3197,
+      "step": 418
+    },
+    {
+      "epoch": 0.08061180318407003,
+      "grad_norm": 2.6570961671620132,
+      "learning_rate": 1.9866993199693393e-05,
+      "loss": 2.8545,
+      "step": 419
+    },
+    {
+      "epoch": 0.08080419412245683,
+      "grad_norm": 4.61377392718083,
+      "learning_rate": 1.98659782213086e-05,
+      "loss": 3.7128,
+      "step": 420
+    },
+    {
+      "epoch": 0.08099658506084363,
+      "grad_norm": 4.185706741557558,
+      "learning_rate": 1.986495941109156e-05,
+      "loss": 3.4406,
+      "step": 421
+    },
+    {
+      "epoch": 0.08118897599923043,
+      "grad_norm": 4.118502039432723,
+      "learning_rate": 1.9863936769437956e-05,
+      "loss": 3.5509,
+      "step": 422
+    },
+    {
+      "epoch": 0.08138136693761724,
+      "grad_norm": 3.798376707510411,
+      "learning_rate": 1.986291029674497e-05,
+      "loss": 3.2303,
+      "step": 423
+    },
+    {
+      "epoch": 0.08157375787600404,
+      "grad_norm": 4.032188489653101,
+      "learning_rate": 1.986187999341128e-05,
+      "loss": 3.259,
+      "step": 424
+    },
+    {
+      "epoch": 0.08176614881439084,
+      "grad_norm": 2.7512321392444883,
+      "learning_rate": 1.9860845859837034e-05,
+      "loss": 2.9476,
+      "step": 425
+    },
+    {
+      "epoch": 0.08195853975277764,
+      "grad_norm": 4.239581592295298,
+      "learning_rate": 1.985980789642388e-05,
+      "loss": 3.2201,
+      "step": 426
+    },
+    {
+      "epoch": 0.08215093069116444,
+      "grad_norm": 4.370388170532908,
+      "learning_rate": 1.985876610357496e-05,
+      "loss": 3.6001,
+      "step": 427
+    },
+    {
+      "epoch": 0.08234332162955124,
+      "grad_norm": 4.227656599484725,
+      "learning_rate": 1.9857720481694887e-05,
+      "loss": 3.616,
+      "step": 428
+    },
+    {
+      "epoch": 0.08253571256793805,
+      "grad_norm": 3.6553375729141573,
+      "learning_rate": 1.9856671031189765e-05,
+      "loss": 3.5295,
+      "step": 429
+    },
+    {
+      "epoch": 0.08272810350632485,
+      "grad_norm": 3.763243449877697,
+      "learning_rate": 1.98556177524672e-05,
+      "loss": 3.4549,
+      "step": 430
+    },
+    {
+      "epoch": 0.08292049444471165,
+      "grad_norm": 4.299479966729342,
+      "learning_rate": 1.9854560645936262e-05,
+      "loss": 3.5363,
+      "step": 431
+    },
+    {
+      "epoch": 0.08311288538309845,
+      "grad_norm": 4.011600183487259,
+      "learning_rate": 1.9853499712007523e-05,
+      "loss": 3.5779,
+      "step": 432
+    },
+    {
+      "epoch": 0.08330527632148525,
+      "grad_norm": 3.8739906692172594,
+      "learning_rate": 1.9852434951093035e-05,
+      "loss": 3.5287,
+      "step": 433
+    },
+    {
+      "epoch": 0.08349766725987207,
+      "grad_norm": 3.926981071244615,
+      "learning_rate": 1.985136636360635e-05,
+      "loss": 3.3494,
+      "step": 434
+    },
+    {
+      "epoch": 0.08369005819825887,
+      "grad_norm": 3.8632970838720153,
+      "learning_rate": 1.985029394996248e-05,
+      "loss": 3.5918,
+      "step": 435
+    },
+    {
+      "epoch": 0.08388244913664566,
+      "grad_norm": 2.900798232366059,
+      "learning_rate": 1.9849217710577945e-05,
+      "loss": 3.0259,
+      "step": 436
+    },
+    {
+      "epoch": 0.08407484007503246,
+      "grad_norm": 4.1071575984109225,
+      "learning_rate": 1.9848137645870745e-05,
+      "loss": 3.2435,
+      "step": 437
+    },
+    {
+      "epoch": 0.08426723101341926,
+      "grad_norm": 3.944193931828056,
+      "learning_rate": 1.9847053756260363e-05,
+      "loss": 3.5384,
+      "step": 438
+    },
+    {
+      "epoch": 0.08445962195180606,
+      "grad_norm": 3.9498777589990364,
+      "learning_rate": 1.984596604216777e-05,
+      "loss": 3.3716,
+      "step": 439
+    },
+    {
+      "epoch": 0.08465201289019288,
+      "grad_norm": 3.8642269255635497,
+      "learning_rate": 1.984487450401542e-05,
+      "loss": 3.5756,
+      "step": 440
+    },
+    {
+      "epoch": 0.08484440382857968,
+      "grad_norm": 3.752402380966192,
+      "learning_rate": 1.9843779142227258e-05,
+      "loss": 3.395,
+      "step": 441
+    },
+    {
+      "epoch": 0.08503679476696648,
+      "grad_norm": 3.921588341489957,
+      "learning_rate": 1.9842679957228706e-05,
+      "loss": 3.6948,
+      "step": 442
+    },
+    {
+      "epoch": 0.08522918570535327,
+      "grad_norm": 3.7176571337442392,
+      "learning_rate": 1.9841576949446675e-05,
+      "loss": 3.4497,
+      "step": 443
+    },
+    {
+      "epoch": 0.08542157664374007,
+      "grad_norm": 3.9053694942483776,
+      "learning_rate": 1.984047011930956e-05,
+      "loss": 3.4399,
+      "step": 444
+    },
+    {
+      "epoch": 0.08561396758212689,
+      "grad_norm": 4.011986877200754,
+      "learning_rate": 1.9839359467247243e-05,
+      "loss": 3.2479,
+      "step": 445
+    },
+    {
+      "epoch": 0.08580635852051369,
+      "grad_norm": 3.956047164687606,
+      "learning_rate": 1.983824499369109e-05,
+      "loss": 3.7104,
+      "step": 446
+    },
+    {
+      "epoch": 0.08599874945890049,
+      "grad_norm": 4.0679196060824205,
+      "learning_rate": 1.9837126699073948e-05,
+      "loss": 3.5776,
+      "step": 447
+    },
+    {
+      "epoch": 0.08619114039728729,
+      "grad_norm": 3.736228925613589,
+      "learning_rate": 1.9836004583830146e-05,
+      "loss": 3.5278,
+      "step": 448
+    },
+    {
+      "epoch": 0.08638353133567409,
+      "grad_norm": 3.86021501721661,
+      "learning_rate": 1.9834878648395507e-05,
+      "loss": 3.6469,
+      "step": 449
+    },
+    {
+      "epoch": 0.0865759222740609,
+      "grad_norm": 3.7235063252414013,
+      "learning_rate": 1.9833748893207326e-05,
+      "loss": 3.4347,
+      "step": 450
+    },
+    {
+      "epoch": 0.0867683132124477,
+      "grad_norm": 3.9334052718440238,
+      "learning_rate": 1.9832615318704388e-05,
+      "loss": 3.2341,
+      "step": 451
+    },
+    {
+      "epoch": 0.0869607041508345,
+      "grad_norm": 3.6310604315616093,
+      "learning_rate": 1.9831477925326962e-05,
+      "loss": 3.4862,
+      "step": 452
+    },
+    {
+      "epoch": 0.0871530950892213,
+      "grad_norm": 4.098013174930056,
+      "learning_rate": 1.98303367135168e-05,
+      "loss": 3.5609,
+      "step": 453
+    },
+    {
+      "epoch": 0.0873454860276081,
+      "grad_norm": 2.5434820640944524,
+      "learning_rate": 1.9829191683717133e-05,
+      "loss": 2.8589,
+      "step": 454
+    },
+    {
+      "epoch": 0.0875378769659949,
+      "grad_norm": 2.5336873496739862,
+      "learning_rate": 1.9828042836372677e-05,
+      "loss": 3.0332,
+      "step": 455
+    },
+    {
+      "epoch": 0.08773026790438171,
+      "grad_norm": 4.102426889612231,
+      "learning_rate": 1.9826890171929634e-05,
+      "loss": 3.4248,
+      "step": 456
+    },
+    {
+      "epoch": 0.08792265884276851,
+      "grad_norm": 3.8750079625845184,
+      "learning_rate": 1.982573369083568e-05,
+      "loss": 3.1916,
+      "step": 457
+    },
+    {
+      "epoch": 0.08811504978115531,
+      "grad_norm": 4.256075701125683,
+      "learning_rate": 1.9824573393539984e-05,
+      "loss": 3.4743,
+      "step": 458
+    },
+    {
+      "epoch": 0.08830744071954211,
+      "grad_norm": 4.043118201071879,
+      "learning_rate": 1.982340928049319e-05,
+      "loss": 3.7704,
+      "step": 459
+    },
+    {
+      "epoch": 0.0884998316579289,
+      "grad_norm": 3.7951143970012278,
+      "learning_rate": 1.9822241352147426e-05,
+      "loss": 3.5927,
+      "step": 460
+    },
+    {
+      "epoch": 0.08869222259631572,
+      "grad_norm": 4.504902128406412,
+      "learning_rate": 1.9821069608956307e-05,
+      "loss": 3.4755,
+      "step": 461
+    },
+    {
+      "epoch": 0.08888461353470252,
+      "grad_norm": 4.075224057836491,
+      "learning_rate": 1.9819894051374917e-05,
+      "loss": 3.5348,
+      "step": 462
+    },
+    {
+      "epoch": 0.08907700447308932,
+      "grad_norm": 4.136715455418681,
+      "learning_rate": 1.981871467985983e-05,
+      "loss": 3.3743,
+      "step": 463
+    },
+    {
+      "epoch": 0.08926939541147612,
+      "grad_norm": 3.653061724310273,
+      "learning_rate": 1.9817531494869105e-05,
+      "loss": 3.3789,
+      "step": 464
+    },
+    {
+      "epoch": 0.08946178634986292,
+      "grad_norm": 3.8311496889033823,
+      "learning_rate": 1.9816344496862272e-05,
+      "loss": 3.4076,
+      "step": 465
+    },
+    {
+      "epoch": 0.08965417728824972,
+      "grad_norm": 2.479124146016505,
+      "learning_rate": 1.9815153686300352e-05,
+      "loss": 2.7875,
+      "step": 466
+    },
+    {
+      "epoch": 0.08984656822663653,
+      "grad_norm": 3.8957537561218203,
+      "learning_rate": 1.981395906364584e-05,
+      "loss": 3.5057,
+      "step": 467
+    },
+    {
+      "epoch": 0.09003895916502333,
+      "grad_norm": 4.033329007056783,
+      "learning_rate": 1.9812760629362714e-05,
+      "loss": 3.4297,
+      "step": 468
+    },
+    {
+      "epoch": 0.09023135010341013,
+      "grad_norm": 3.6057027929824925,
+      "learning_rate": 1.981155838391643e-05,
+      "loss": 3.2852,
+      "step": 469
+    },
+    {
+      "epoch": 0.09042374104179693,
+      "grad_norm": 3.814070646146285,
+      "learning_rate": 1.9810352327773935e-05,
+      "loss": 3.542,
+      "step": 470
+    },
+    {
+      "epoch": 0.09061613198018373,
+      "grad_norm": 4.160785849649063,
+      "learning_rate": 1.9809142461403635e-05,
+      "loss": 3.4473,
+      "step": 471
+    },
+    {
+      "epoch": 0.09080852291857054,
+      "grad_norm": 4.041053169009656,
+      "learning_rate": 1.9807928785275433e-05,
+      "loss": 3.5377,
+      "step": 472
+    },
+    {
+      "epoch": 0.09100091385695734,
+      "grad_norm": 3.892539400022939,
+      "learning_rate": 1.980671129986071e-05,
+      "loss": 3.3937,
+      "step": 473
+    },
+    {
+      "epoch": 0.09119330479534414,
+      "grad_norm": 3.8990823233233662,
+      "learning_rate": 1.9805490005632323e-05,
+      "loss": 3.642,
+      "step": 474
+    },
+    {
+      "epoch": 0.09138569573373094,
+      "grad_norm": 3.5298775070078463,
+      "learning_rate": 1.98042649030646e-05,
+      "loss": 3.1533,
+      "step": 475
+    },
+    {
+      "epoch": 0.09157808667211774,
+      "grad_norm": 3.739256683581738,
+      "learning_rate": 1.9803035992633366e-05,
+      "loss": 3.5223,
+      "step": 476
+    },
+    {
+      "epoch": 0.09177047761050454,
+      "grad_norm": 3.9248857596645235,
+      "learning_rate": 1.9801803274815915e-05,
+      "loss": 3.1959,
+      "step": 477
+    },
+    {
+      "epoch": 0.09196286854889135,
+      "grad_norm": 4.197695197922203,
+      "learning_rate": 1.9800566750091018e-05,
+      "loss": 3.5585,
+      "step": 478
+    },
+    {
+      "epoch": 0.09215525948727815,
+      "grad_norm": 3.9943795397691964,
+      "learning_rate": 1.9799326418938924e-05,
+      "loss": 3.3692,
+      "step": 479
+    },
+    {
+      "epoch": 0.09234765042566495,
+      "grad_norm": 4.4299981757935765,
+      "learning_rate": 1.979808228184137e-05,
+      "loss": 3.5715,
+      "step": 480
+    },
+    {
+      "epoch": 0.09254004136405175,
+      "grad_norm": 3.9875020894656865,
+      "learning_rate": 1.9796834339281557e-05,
+      "loss": 3.4909,
+      "step": 481
+    },
+    {
+      "epoch": 0.09273243230243855,
+      "grad_norm": 2.5498206738884193,
+      "learning_rate": 1.979558259174418e-05,
+      "loss": 2.8785,
+      "step": 482
+    },
+    {
+      "epoch": 0.09292482324082536,
+      "grad_norm": 4.233607674408517,
+      "learning_rate": 1.9794327039715395e-05,
+      "loss": 3.5097,
+      "step": 483
+    },
+    {
+      "epoch": 0.09311721417921216,
+      "grad_norm": 4.006494879505914,
+      "learning_rate": 1.979306768368285e-05,
+      "loss": 3.4646,
+      "step": 484
+    },
+    {
+      "epoch": 0.09330960511759896,
+      "grad_norm": 4.028357723221071,
+      "learning_rate": 1.9791804524135663e-05,
+      "loss": 3.2552,
+      "step": 485
+    },
+    {
+      "epoch": 0.09350199605598576,
+      "grad_norm": 2.496059264284846,
+      "learning_rate": 1.979053756156443e-05,
+      "loss": 2.835,
+      "step": 486
+    },
+    {
+      "epoch": 0.09369438699437256,
+      "grad_norm": 4.103231624690585,
+      "learning_rate": 1.9789266796461222e-05,
+      "loss": 3.5353,
+      "step": 487
+    },
+    {
+      "epoch": 0.09388677793275936,
+      "grad_norm": 4.030902584041719,
+      "learning_rate": 1.9787992229319594e-05,
+      "loss": 3.3925,
+      "step": 488
+    },
+    {
+      "epoch": 0.09407916887114617,
+      "grad_norm": 3.574008406748338,
+      "learning_rate": 1.978671386063457e-05,
+      "loss": 3.3375,
+      "step": 489
+    },
+    {
+      "epoch": 0.09427155980953297,
+      "grad_norm": 4.377663615178794,
+      "learning_rate": 1.978543169090265e-05,
+      "loss": 3.5363,
+      "step": 490
+    },
+    {
+      "epoch": 0.09446395074791977,
+      "grad_norm": 3.693629643505739,
+      "learning_rate": 1.9784145720621827e-05,
+      "loss": 3.6535,
+      "step": 491
+    },
+    {
+      "epoch": 0.09465634168630657,
+      "grad_norm": 3.87232266678281,
+      "learning_rate": 1.9782855950291542e-05,
+      "loss": 3.4199,
+      "step": 492
+    },
+    {
+      "epoch": 0.09484873262469337,
+      "grad_norm": 3.804180921945895,
+      "learning_rate": 1.978156238041274e-05,
+      "loss": 3.7146,
+      "step": 493
+    },
+    {
+      "epoch": 0.09504112356308018,
+      "grad_norm": 3.93639974807964,
+      "learning_rate": 1.9780265011487822e-05,
+      "loss": 3.4791,
+      "step": 494
+    },
+    {
+      "epoch": 0.09523351450146698,
+      "grad_norm": 3.534628653890325,
+      "learning_rate": 1.9778963844020668e-05,
+      "loss": 3.4336,
+      "step": 495
+    },
+    {
+      "epoch": 0.09542590543985378,
+      "grad_norm": 3.999598672304901,
+      "learning_rate": 1.977765887851664e-05,
+      "loss": 3.527,
+      "step": 496
+    },
+    {
+      "epoch": 0.09561829637824058,
+      "grad_norm": 3.9425373535267694,
+      "learning_rate": 1.977635011548257e-05,
+      "loss": 3.3544,
+      "step": 497
+    },
+    {
+      "epoch": 0.09581068731662738,
+      "grad_norm": 2.5446319639439845,
+      "learning_rate": 1.9775037555426772e-05,
+      "loss": 2.8217,
+      "step": 498
+    },
+    {
+      "epoch": 0.09600307825501418,
+      "grad_norm": 3.7362444860218837,
+      "learning_rate": 1.9773721198859024e-05,
+      "loss": 3.449,
+      "step": 499
+    },
+    {
+      "epoch": 0.096195469193401,
+      "grad_norm": 3.98523388484512,
+      "learning_rate": 1.9772401046290584e-05,
+      "loss": 3.4335,
+      "step": 500
+    },
+    {
+      "epoch": 0.0963878601317878,
+      "grad_norm": 3.776270075745232,
+      "learning_rate": 1.9771077098234187e-05,
+      "loss": 3.3569,
+      "step": 501
+    },
+    {
+      "epoch": 0.0965802510701746,
+      "grad_norm": 4.0316070765642715,
+      "learning_rate": 1.9769749355204034e-05,
+      "loss": 3.4556,
+      "step": 502
+    },
+    {
+      "epoch": 0.0967726420085614,
+      "grad_norm": 3.8080118510316776,
+      "learning_rate": 1.976841781771581e-05,
+      "loss": 3.4409,
+      "step": 503
+    },
+    {
+      "epoch": 0.0969650329469482,
+      "grad_norm": 3.884984013586109,
+      "learning_rate": 1.9767082486286667e-05,
+      "loss": 3.6602,
+      "step": 504
+    },
+    {
+      "epoch": 0.097157423885335,
+      "grad_norm": 3.8716423529281934,
+      "learning_rate": 1.9765743361435234e-05,
+      "loss": 3.3926,
+      "step": 505
+    },
+    {
+      "epoch": 0.0973498148237218,
+      "grad_norm": 4.276371321620769,
+      "learning_rate": 1.9764400443681607e-05,
+      "loss": 3.5185,
+      "step": 506
+    },
+    {
+      "epoch": 0.0975422057621086,
+      "grad_norm": 3.7579298228847504,
+      "learning_rate": 1.9763053733547367e-05,
+      "loss": 3.3856,
+      "step": 507
+    },
+    {
+      "epoch": 0.0977345967004954,
+      "grad_norm": 3.667219828038184,
+      "learning_rate": 1.976170323155555e-05,
+      "loss": 3.2837,
+      "step": 508
+    },
+    {
+      "epoch": 0.0979269876388822,
+      "grad_norm": 2.661240284888627,
+      "learning_rate": 1.976034893823069e-05,
+      "loss": 2.9806,
+      "step": 509
+    },
+    {
+      "epoch": 0.09811937857726902,
+      "grad_norm": 4.20104326352077,
+      "learning_rate": 1.975899085409876e-05,
+      "loss": 3.662,
+      "step": 510
+    },
+    {
+      "epoch": 0.09831176951565582,
+      "grad_norm": 4.059512982664822,
+      "learning_rate": 1.9757628979687247e-05,
+      "loss": 3.3928,
+      "step": 511
+    },
+    {
+      "epoch": 0.09850416045404262,
+      "grad_norm": 3.569596204739188,
+      "learning_rate": 1.975626331552507e-05,
+      "loss": 3.5144,
+      "step": 512
+    },
+    {
+      "epoch": 0.09869655139242942,
+      "grad_norm": 3.6999268476712994,
+      "learning_rate": 1.9754893862142643e-05,
+      "loss": 3.2305,
+      "step": 513
+    },
+    {
+      "epoch": 0.09888894233081622,
+      "grad_norm": 4.228674241003315,
+      "learning_rate": 1.9753520620071846e-05,
+      "loss": 3.1777,
+      "step": 514
+    },
+    {
+      "epoch": 0.09908133326920301,
+      "grad_norm": 4.096508120070463,
+      "learning_rate": 1.9752143589846027e-05,
+      "loss": 3.6346,
+      "step": 515
+    },
+    {
+      "epoch": 0.09927372420758983,
+      "grad_norm": 3.7247594310834042,
+      "learning_rate": 1.9750762772000014e-05,
+      "loss": 3.4207,
+      "step": 516
+    },
+    {
+      "epoch": 0.09946611514597663,
+      "grad_norm": 4.1739150115729196,
+      "learning_rate": 1.9749378167070097e-05,
+      "loss": 3.4517,
+      "step": 517
+    },
+    {
+      "epoch": 0.09965850608436343,
+      "grad_norm": 3.9298282572556067,
+      "learning_rate": 1.9747989775594044e-05,
+      "loss": 3.428,
+      "step": 518
+    },
+    {
+      "epoch": 0.09985089702275023,
+      "grad_norm": 4.235364579874633,
+      "learning_rate": 1.974659759811109e-05,
+      "loss": 3.7208,
+      "step": 519
+    },
+    {
+      "epoch": 0.10004328796113703,
+      "grad_norm": 4.257204726050071,
+      "learning_rate": 1.9745201635161938e-05,
+      "loss": 3.4978,
+      "step": 520
+    },
+    {
+      "epoch": 0.10023567889952384,
+      "grad_norm": 3.940546183242936,
+      "learning_rate": 1.9743801887288762e-05,
+      "loss": 3.4045,
+      "step": 521
+    },
+    {
+      "epoch": 0.10042806983791064,
+      "grad_norm": 3.843125455497617,
+      "learning_rate": 1.9742398355035212e-05,
+      "loss": 3.1228,
+      "step": 522
+    },
+    {
+      "epoch": 0.10062046077629744,
+      "grad_norm": 3.9533856646529193,
+      "learning_rate": 1.9740991038946404e-05,
+      "loss": 3.4495,
+      "step": 523
+    },
+    {
+      "epoch": 0.10081285171468424,
+      "grad_norm": 3.9797791575034824,
+      "learning_rate": 1.973957993956892e-05,
+      "loss": 3.3364,
+      "step": 524
+    },
+    {
+      "epoch": 0.10100524265307104,
+      "grad_norm": 3.809585124673704,
+      "learning_rate": 1.9738165057450817e-05,
+      "loss": 3.3171,
+      "step": 525
+    },
+    {
+      "epoch": 0.10119763359145784,
+      "grad_norm": 4.037211873795967,
+      "learning_rate": 1.9736746393141617e-05,
+      "loss": 3.7494,
+      "step": 526
+    },
+    {
+      "epoch": 0.10139002452984465,
+      "grad_norm": 4.217717613442144,
+      "learning_rate": 1.9735323947192317e-05,
+      "loss": 3.5143,
+      "step": 527
+    },
+    {
+      "epoch": 0.10158241546823145,
+      "grad_norm": 3.4983953855725933,
+      "learning_rate": 1.9733897720155377e-05,
+      "loss": 3.4486,
+      "step": 528
+    },
+    {
+      "epoch": 0.10177480640661825,
+      "grad_norm": 3.7001724122972477,
+      "learning_rate": 1.9732467712584723e-05,
+      "loss": 3.1859,
+      "step": 529
+    },
+    {
+      "epoch": 0.10196719734500505,
+      "grad_norm": 3.7511701036874587,
+      "learning_rate": 1.973103392503576e-05,
+      "loss": 3.5398,
+      "step": 530
+    },
+    {
+      "epoch": 0.10215958828339185,
+      "grad_norm": 3.8231294962562687,
+      "learning_rate": 1.9729596358065347e-05,
+      "loss": 3.4751,
+      "step": 531
+    },
+    {
+      "epoch": 0.10235197922177866,
+      "grad_norm": 3.664528745303133,
+      "learning_rate": 1.9728155012231825e-05,
+      "loss": 3.4917,
+      "step": 532
+    },
+    {
+      "epoch": 0.10254437016016546,
+      "grad_norm": 4.149810960560453,
+      "learning_rate": 1.9726709888094994e-05,
+      "loss": 3.2658,
+      "step": 533
+    },
+    {
+      "epoch": 0.10273676109855226,
+      "grad_norm": 4.212770079831097,
+      "learning_rate": 1.972526098621612e-05,
+      "loss": 3.3101,
+      "step": 534
+    },
+    {
+      "epoch": 0.10292915203693906,
+      "grad_norm": 4.113614127992525,
+      "learning_rate": 1.972380830715795e-05,
+      "loss": 3.3821,
+      "step": 535
+    },
+    {
+      "epoch": 0.10312154297532586,
+      "grad_norm": 3.8882160731219093,
+      "learning_rate": 1.9722351851484677e-05,
+      "loss": 3.3939,
+      "step": 536
+    },
+    {
+      "epoch": 0.10331393391371266,
+      "grad_norm": 3.7948237457076948,
+      "learning_rate": 1.9720891619761974e-05,
+      "loss": 3.6299,
+      "step": 537
+    },
+    {
+      "epoch": 0.10350632485209947,
+      "grad_norm": 3.784893607341369,
+      "learning_rate": 1.9719427612556982e-05,
+      "loss": 3.2872,
+      "step": 538
+    },
+    {
+      "epoch": 0.10369871579048627,
+      "grad_norm": 4.161503653227999,
+      "learning_rate": 1.9717959830438302e-05,
+      "loss": 3.3156,
+      "step": 539
+    },
+    {
+      "epoch": 0.10389110672887307,
+      "grad_norm": 3.688328947870827,
+      "learning_rate": 1.9716488273976006e-05,
+      "loss": 3.3155,
+      "step": 540
+    },
+    {
+      "epoch": 0.10408349766725987,
+      "grad_norm": 3.6643330725738865,
+      "learning_rate": 1.971501294374162e-05,
+      "loss": 3.2535,
+      "step": 541
+    },
+    {
+      "epoch": 0.10427588860564667,
+      "grad_norm": 3.925620489126276,
+      "learning_rate": 1.971353384030816e-05,
+      "loss": 3.5806,
+      "step": 542
+    },
+    {
+      "epoch": 0.10446827954403348,
+      "grad_norm": 3.6644710770355045,
+      "learning_rate": 1.9712050964250083e-05,
+      "loss": 3.7024,
+      "step": 543
+    },
+    {
+      "epoch": 0.10466067048242028,
+      "grad_norm": 4.017962963745318,
+      "learning_rate": 1.9710564316143323e-05,
+      "loss": 3.5019,
+      "step": 544
+    },
+    {
+      "epoch": 0.10485306142080708,
+      "grad_norm": 3.9748934281424475,
+      "learning_rate": 1.9709073896565276e-05,
+      "loss": 3.5279,
+      "step": 545
+    },
+    {
+      "epoch": 0.10504545235919388,
+      "grad_norm": 4.011630155344541,
+      "learning_rate": 1.9707579706094807e-05,
+      "loss": 3.4081,
+      "step": 546
+    },
+    {
+      "epoch": 0.10523784329758068,
+      "grad_norm": 3.969720584904166,
+      "learning_rate": 1.970608174531224e-05,
+      "loss": 3.4745,
+      "step": 547
+    },
+    {
+      "epoch": 0.10543023423596748,
+      "grad_norm": 3.606005427059065,
+      "learning_rate": 1.970458001479937e-05,
+      "loss": 3.0979,
+      "step": 548
+    },
+    {
+      "epoch": 0.10562262517435429,
+      "grad_norm": 4.045469218868416,
+      "learning_rate": 1.9703074515139445e-05,
+      "loss": 3.3946,
+      "step": 549
+    },
+    {
+      "epoch": 0.10581501611274109,
+      "grad_norm": 3.8069322438106163,
+      "learning_rate": 1.9701565246917184e-05,
+      "loss": 3.5141,
+      "step": 550
+    },
+    {
+      "epoch": 0.10600740705112789,
+      "grad_norm": 3.669928548454659,
+      "learning_rate": 1.9700052210718775e-05,
+      "loss": 3.4443,
+      "step": 551
+    },
+    {
+      "epoch": 0.10619979798951469,
+      "grad_norm": 3.9171510484118035,
+      "learning_rate": 1.969853540713186e-05,
+      "loss": 3.6701,
+      "step": 552
+    },
+    {
+      "epoch": 0.10639218892790149,
+      "grad_norm": 4.071017430700369,
+      "learning_rate": 1.9697014836745552e-05,
+      "loss": 3.5177,
+      "step": 553
+    },
+    {
+      "epoch": 0.1065845798662883,
+      "grad_norm": 4.266547173205893,
+      "learning_rate": 1.9695490500150418e-05,
+      "loss": 3.5378,
+      "step": 554
+    },
+    {
+      "epoch": 0.1067769708046751,
+      "grad_norm": 3.8285664471653544,
+      "learning_rate": 1.9693962397938495e-05,
+      "loss": 3.6287,
+      "step": 555
+    },
+    {
+      "epoch": 0.1069693617430619,
+      "grad_norm": 3.704085231375269,
+      "learning_rate": 1.9692430530703282e-05,
+      "loss": 3.4446,
+      "step": 556
+    },
+    {
+      "epoch": 0.1071617526814487,
+      "grad_norm": 4.192088052077803,
+      "learning_rate": 1.9690894899039735e-05,
+      "loss": 3.41,
+      "step": 557
+    },
+    {
+      "epoch": 0.1073541436198355,
+      "grad_norm": 4.145031493503153,
+      "learning_rate": 1.9689355503544277e-05,
+      "loss": 3.6728,
+      "step": 558
+    },
+    {
+      "epoch": 0.1075465345582223,
+      "grad_norm": 2.5023284857081576,
+      "learning_rate": 1.968781234481479e-05,
+      "loss": 2.6567,
+      "step": 559
+    },
+    {
+      "epoch": 0.10773892549660911,
+      "grad_norm": 3.8057278305132836,
+      "learning_rate": 1.9686265423450624e-05,
+      "loss": 3.7296,
+      "step": 560
+    },
+    {
+      "epoch": 0.10793131643499591,
+      "grad_norm": 4.085816286082312,
+      "learning_rate": 1.9684714740052584e-05,
+      "loss": 3.455,
+      "step": 561
+    },
+    {
+      "epoch": 0.10812370737338271,
+      "grad_norm": 4.197497602256956,
+      "learning_rate": 1.9683160295222934e-05,
+      "loss": 3.2806,
+      "step": 562
+    },
+    {
+      "epoch": 0.10831609831176951,
+      "grad_norm": 4.099550948663347,
+      "learning_rate": 1.9681602089565403e-05,
+      "loss": 3.4633,
+      "step": 563
+    },
+    {
+      "epoch": 0.10850848925015631,
+      "grad_norm": 3.871744373851531,
+      "learning_rate": 1.968004012368518e-05,
+      "loss": 3.4573,
+      "step": 564
+    },
+    {
+      "epoch": 0.10870088018854313,
+      "grad_norm": 3.7478430249928394,
+      "learning_rate": 1.967847439818892e-05,
+      "loss": 3.5944,
+      "step": 565
+    },
+    {
+      "epoch": 0.10889327112692992,
+      "grad_norm": 3.651559000654838,
+      "learning_rate": 1.9676904913684725e-05,
+      "loss": 3.3069,
+      "step": 566
+    },
+    {
+      "epoch": 0.10908566206531672,
+      "grad_norm": 3.797559968280414,
+      "learning_rate": 1.967533167078217e-05,
+      "loss": 3.449,
+      "step": 567
+    },
+    {
+      "epoch": 0.10927805300370352,
+      "grad_norm": 3.7987534129209206,
+      "learning_rate": 1.9673754670092283e-05,
+      "loss": 3.0108,
+      "step": 568
+    },
+    {
+      "epoch": 0.10947044394209032,
+      "grad_norm": 3.7672780028505,
+      "learning_rate": 1.9672173912227556e-05,
+      "loss": 3.6215,
+      "step": 569
+    },
+    {
+      "epoch": 0.10966283488047714,
+      "grad_norm": 2.6424366210420187,
+      "learning_rate": 1.967058939780193e-05,
+      "loss": 2.7987,
+      "step": 570
+    },
+    {
+      "epoch": 0.10985522581886394,
+      "grad_norm": 4.007056238216178,
+      "learning_rate": 1.966900112743082e-05,
+      "loss": 3.5677,
+      "step": 571
+    },
+    {
+      "epoch": 0.11004761675725074,
+      "grad_norm": 3.681765824259293,
+      "learning_rate": 1.966740910173108e-05,
+      "loss": 3.1244,
+      "step": 572
+    },
+    {
+      "epoch": 0.11024000769563753,
+      "grad_norm": 3.957626369418781,
+      "learning_rate": 1.9665813321321054e-05,
+      "loss": 3.3474,
+      "step": 573
+    },
+    {
+      "epoch": 0.11043239863402433,
+      "grad_norm": 3.9214513799422295,
+      "learning_rate": 1.9664213786820502e-05,
+      "loss": 3.4417,
+      "step": 574
+    },
+    {
+      "epoch": 0.11062478957241113,
+      "grad_norm": 3.548248619392042,
+      "learning_rate": 1.9662610498850684e-05,
+      "loss": 3.3698,
+      "step": 575
+    },
+    {
+      "epoch": 0.11081718051079795,
+      "grad_norm": 3.9946389877612707,
+      "learning_rate": 1.9661003458034288e-05,
+      "loss": 3.3542,
+      "step": 576
+    },
+    {
+      "epoch": 0.11100957144918475,
+      "grad_norm": 3.721626096259351,
+      "learning_rate": 1.965939266499547e-05,
+      "loss": 3.5496,
+      "step": 577
+    },
+    {
+      "epoch": 0.11120196238757155,
+      "grad_norm": 4.027025653633322,
+      "learning_rate": 1.9657778120359848e-05,
+      "loss": 3.4296,
+      "step": 578
+    },
+    {
+      "epoch": 0.11139435332595835,
+      "grad_norm": 3.958846798686176,
+      "learning_rate": 1.965615982475449e-05,
+      "loss": 3.2222,
+      "step": 579
+    },
+    {
+      "epoch": 0.11158674426434514,
+      "grad_norm": 3.576273325188134,
+      "learning_rate": 1.9654537778807924e-05,
+      "loss": 3.4524,
+      "step": 580
+    },
+    {
+      "epoch": 0.11177913520273196,
+      "grad_norm": 3.8044717356759508,
+      "learning_rate": 1.9652911983150135e-05,
+      "loss": 3.3657,
+      "step": 581
+    },
+    {
+      "epoch": 0.11197152614111876,
+      "grad_norm": 3.786040242393609,
+      "learning_rate": 1.965128243841256e-05,
+      "loss": 3.6601,
+      "step": 582
+    },
+    {
+      "epoch": 0.11216391707950556,
+      "grad_norm": 4.065643074510649,
+      "learning_rate": 1.96496491452281e-05,
+      "loss": 3.2806,
+      "step": 583
+    },
+    {
+      "epoch": 0.11235630801789236,
+      "grad_norm": 3.908528358571422,
+      "learning_rate": 1.9648012104231106e-05,
+      "loss": 3.6021,
+      "step": 584
+    },
+    {
+      "epoch": 0.11254869895627916,
+      "grad_norm": 3.846144977749209,
+      "learning_rate": 1.964637131605738e-05,
+      "loss": 3.5302,
+      "step": 585
+    },
+    {
+      "epoch": 0.11274108989466596,
+      "grad_norm": 3.8306544870281876,
+      "learning_rate": 1.9644726781344197e-05,
+      "loss": 3.2063,
+      "step": 586
+    },
+    {
+      "epoch": 0.11293348083305277,
+      "grad_norm": 3.63848284441587,
+      "learning_rate": 1.964307850073026e-05,
+      "loss": 3.3972,
+      "step": 587
+    },
+    {
+      "epoch": 0.11312587177143957,
+      "grad_norm": 3.6928537686650196,
+      "learning_rate": 1.964142647485576e-05,
+      "loss": 3.1231,
+      "step": 588
+    },
+    {
+      "epoch": 0.11331826270982637,
+      "grad_norm": 3.914662459762175,
+      "learning_rate": 1.9639770704362305e-05,
+      "loss": 3.1678,
+      "step": 589
+    },
+    {
+      "epoch": 0.11351065364821317,
+      "grad_norm": 3.763690773498089,
+      "learning_rate": 1.9638111189892994e-05,
+      "loss": 3.2837,
+      "step": 590
+    },
+    {
+      "epoch": 0.11370304458659997,
+      "grad_norm": 4.24216361115245,
+      "learning_rate": 1.9636447932092354e-05,
+      "loss": 3.7666,
+      "step": 591
+    },
+    {
+      "epoch": 0.11389543552498678,
+      "grad_norm": 3.847203895407255,
+      "learning_rate": 1.963478093160638e-05,
+      "loss": 3.266,
+      "step": 592
+    },
+    {
+      "epoch": 0.11408782646337358,
+      "grad_norm": 3.9589994395516737,
+      "learning_rate": 1.9633110189082515e-05,
+      "loss": 3.2244,
+      "step": 593
+    },
+    {
+      "epoch": 0.11428021740176038,
+      "grad_norm": 3.9150741894805425,
+      "learning_rate": 1.963143570516965e-05,
+      "loss": 3.3048,
+      "step": 594
+    },
+    {
+      "epoch": 0.11447260834014718,
+      "grad_norm": 2.542678087152977,
+      "learning_rate": 1.9629757480518144e-05,
+      "loss": 2.8083,
+      "step": 595
+    },
+    {
+      "epoch": 0.11466499927853398,
+      "grad_norm": 4.147748586669416,
+      "learning_rate": 1.9628075515779796e-05,
+      "loss": 3.3396,
+      "step": 596
+    },
+    {
+      "epoch": 0.11485739021692078,
+      "grad_norm": 3.9103841362829925,
+      "learning_rate": 1.962638981160786e-05,
+      "loss": 3.2086,
+      "step": 597
+    },
+    {
+      "epoch": 0.11504978115530759,
+      "grad_norm": 3.734059464582647,
+      "learning_rate": 1.9624700368657045e-05,
+      "loss": 3.3915,
+      "step": 598
+    },
+    {
+      "epoch": 0.11524217209369439,
+      "grad_norm": 3.869783588242207,
+      "learning_rate": 1.9623007187583518e-05,
+      "loss": 3.454,
+      "step": 599
+    },
+    {
+      "epoch": 0.11543456303208119,
+      "grad_norm": 3.678697421852201,
+      "learning_rate": 1.962131026904488e-05,
+      "loss": 3.4913,
+      "step": 600
+    },
+    {
+      "epoch": 0.11562695397046799,
+      "grad_norm": 4.066908023310014,
+      "learning_rate": 1.96196096137002e-05,
+      "loss": 3.1405,
+      "step": 601
+    },
+    {
+      "epoch": 0.11581934490885479,
+      "grad_norm": 3.740013726945565,
+      "learning_rate": 1.9617905222209998e-05,
+      "loss": 3.4924,
+      "step": 602
+    },
+    {
+      "epoch": 0.1160117358472416,
+      "grad_norm": 2.4754584776030617,
+      "learning_rate": 1.961619709523623e-05,
+      "loss": 3.0725,
+      "step": 603
+    },
+    {
+      "epoch": 0.1162041267856284,
+      "grad_norm": 3.8916338388725022,
+      "learning_rate": 1.9614485233442316e-05,
+      "loss": 3.6384,
+      "step": 604
+    },
+    {
+      "epoch": 0.1163965177240152,
+      "grad_norm": 4.123543815266122,
+      "learning_rate": 1.961276963749313e-05,
+      "loss": 3.2611,
+      "step": 605
+    },
+    {
+      "epoch": 0.116588908662402,
+      "grad_norm": 3.9415999892025604,
+      "learning_rate": 1.9611050308054982e-05,
+      "loss": 3.7221,
+      "step": 606
+    },
+    {
+      "epoch": 0.1167812996007888,
+      "grad_norm": 3.9714577199827987,
+      "learning_rate": 1.9609327245795642e-05,
+      "loss": 3.3412,
+      "step": 607
+    },
+    {
+      "epoch": 0.1169736905391756,
+      "grad_norm": 3.930302766766924,
+      "learning_rate": 1.9607600451384327e-05,
+      "loss": 3.5011,
+      "step": 608
+    },
+    {
+      "epoch": 0.11716608147756241,
+      "grad_norm": 3.610540756521898,
+      "learning_rate": 1.960586992549171e-05,
+      "loss": 3.4537,
+      "step": 609
+    },
+    {
+      "epoch": 0.11735847241594921,
+      "grad_norm": 3.958036123987214,
+      "learning_rate": 1.9604135668789897e-05,
+      "loss": 3.5074,
+      "step": 610
+    },
+    {
+      "epoch": 0.11755086335433601,
+      "grad_norm": 3.8590178174067056,
+      "learning_rate": 1.9602397681952462e-05,
+      "loss": 3.2923,
+      "step": 611
+    },
+    {
+      "epoch": 0.11774325429272281,
+      "grad_norm": 2.4396542436650375,
+      "learning_rate": 1.9600655965654413e-05,
+      "loss": 2.8096,
+      "step": 612
+    },
+    {
+      "epoch": 0.11793564523110961,
+      "grad_norm": 3.966893796974018,
+      "learning_rate": 1.959891052057222e-05,
+      "loss": 3.3364,
+      "step": 613
+    },
+    {
+      "epoch": 0.11812803616949642,
+      "grad_norm": 4.09830052089381,
+      "learning_rate": 1.9597161347383783e-05,
+      "loss": 3.3141,
+      "step": 614
+    },
+    {
+      "epoch": 0.11832042710788322,
+      "grad_norm": 4.005024211648436,
+      "learning_rate": 1.959540844676847e-05,
+      "loss": 3.2178,
+      "step": 615
+    },
+    {
+      "epoch": 0.11851281804627002,
+      "grad_norm": 3.7441662919286602,
+      "learning_rate": 1.9593651819407084e-05,
+      "loss": 3.4869,
+      "step": 616
+    },
+    {
+      "epoch": 0.11870520898465682,
+      "grad_norm": 3.796368806045701,
+      "learning_rate": 1.959189146598188e-05,
+      "loss": 3.4804,
+      "step": 617
+    },
+    {
+      "epoch": 0.11889759992304362,
+      "grad_norm": 3.730795820778722,
+      "learning_rate": 1.9590127387176556e-05,
+      "loss": 3.5284,
+      "step": 618
+    },
+    {
+      "epoch": 0.11908999086143042,
+      "grad_norm": 3.7845770889139017,
+      "learning_rate": 1.9588359583676263e-05,
+      "loss": 3.2563,
+      "step": 619
+    },
+    {
+      "epoch": 0.11928238179981723,
+      "grad_norm": 3.8676716175088433,
+      "learning_rate": 1.9586588056167595e-05,
+      "loss": 3.2073,
+      "step": 620
+    },
+    {
+      "epoch": 0.11947477273820403,
+      "grad_norm": 4.050224333768339,
+      "learning_rate": 1.958481280533859e-05,
+      "loss": 3.2945,
+      "step": 621
+    },
+    {
+      "epoch": 0.11966716367659083,
+      "grad_norm": 3.948282601696173,
+      "learning_rate": 1.958303383187874e-05,
+      "loss": 3.2916,
+      "step": 622
+    },
+    {
+      "epoch": 0.11985955461497763,
+      "grad_norm": 2.4296837616529103,
+      "learning_rate": 1.9581251136478974e-05,
+      "loss": 2.8839,
+      "step": 623
+    },
+    {
+      "epoch": 0.12005194555336443,
+      "grad_norm": 3.9228254262216913,
+      "learning_rate": 1.9579464719831668e-05,
+      "loss": 3.6049,
+      "step": 624
+    },
+    {
+      "epoch": 0.12024433649175124,
+      "grad_norm": 3.8547259341397595,
+      "learning_rate": 1.9577674582630653e-05,
+      "loss": 3.4498,
+      "step": 625
+    },
+    {
+      "epoch": 0.12043672743013804,
+      "grad_norm": 2.3711529113796095,
+      "learning_rate": 1.957588072557119e-05,
+      "loss": 2.7456,
+      "step": 626
+    },
+    {
+      "epoch": 0.12062911836852484,
+      "grad_norm": 4.225619562415427,
+      "learning_rate": 1.957408314935e-05,
+      "loss": 3.6149,
+      "step": 627
+    },
+    {
+      "epoch": 0.12082150930691164,
+      "grad_norm": 3.7492958304667097,
+      "learning_rate": 1.9572281854665233e-05,
+      "loss": 3.4971,
+      "step": 628
+    },
+    {
+      "epoch": 0.12101390024529844,
+      "grad_norm": 4.998632427191666,
+      "learning_rate": 1.95704768422165e-05,
+      "loss": 3.6341,
+      "step": 629
+    },
+    {
+      "epoch": 0.12120629118368526,
+      "grad_norm": 3.882218145235889,
+      "learning_rate": 1.956866811270484e-05,
+      "loss": 3.631,
+      "step": 630
+    },
+    {
+      "epoch": 0.12139868212207205,
+      "grad_norm": 2.2460929853326057,
+      "learning_rate": 1.9566855666832743e-05,
+      "loss": 2.8174,
+      "step": 631
+    },
+    {
+      "epoch": 0.12159107306045885,
+      "grad_norm": 4.077702614339351,
+      "learning_rate": 1.9565039505304145e-05,
+      "loss": 3.3573,
+      "step": 632
+    },
+    {
+      "epoch": 0.12178346399884565,
+      "grad_norm": 3.6746077534948176,
+      "learning_rate": 1.956321962882442e-05,
+      "loss": 3.3141,
+      "step": 633
+    },
+    {
+      "epoch": 0.12197585493723245,
+      "grad_norm": 2.4793280287846438,
+      "learning_rate": 1.956139603810039e-05,
+      "loss": 2.7447,
+      "step": 634
+    },
+    {
+      "epoch": 0.12216824587561925,
+      "grad_norm": 3.669735794492115,
+      "learning_rate": 1.9559568733840317e-05,
+      "loss": 3.4543,
+      "step": 635
+    },
+    {
+      "epoch": 0.12236063681400607,
+      "grad_norm": 4.308428659062334,
+      "learning_rate": 1.9557737716753898e-05,
+      "loss": 3.3498,
+      "step": 636
+    },
+    {
+      "epoch": 0.12255302775239287,
+      "grad_norm": 4.077717897963782,
+      "learning_rate": 1.9555902987552283e-05,
+      "loss": 3.726,
+      "step": 637
+    },
+    {
+      "epoch": 0.12274541869077966,
+      "grad_norm": 3.928131581519362,
+      "learning_rate": 1.9554064546948064e-05,
+      "loss": 3.4233,
+      "step": 638
+    },
+    {
+      "epoch": 0.12293780962916646,
+      "grad_norm": 4.033091815775968,
+      "learning_rate": 1.9552222395655262e-05,
+      "loss": 3.591,
+      "step": 639
+    },
+    {
+      "epoch": 0.12313020056755326,
+      "grad_norm": 3.934341009032088,
+      "learning_rate": 1.9550376534389355e-05,
+      "loss": 3.2933,
+      "step": 640
+    },
+    {
+      "epoch": 0.12332259150594008,
+      "grad_norm": 3.6966099699064303,
+      "learning_rate": 1.9548526963867253e-05,
+      "loss": 3.6508,
+      "step": 641
+    },
+    {
+      "epoch": 0.12351498244432688,
+      "grad_norm": 3.650793439990128,
+      "learning_rate": 1.9546673684807303e-05,
+      "loss": 3.1553,
+      "step": 642
+    },
+    {
+      "epoch": 0.12370737338271368,
+      "grad_norm": 3.890279813930762,
+      "learning_rate": 1.95448166979293e-05,
+      "loss": 3.2069,
+      "step": 643
+    },
+    {
+      "epoch": 0.12389976432110048,
+      "grad_norm": 3.780149905524223,
+      "learning_rate": 1.9542956003954477e-05,
+      "loss": 3.4439,
+      "step": 644
+    },
+    {
+      "epoch": 0.12409215525948727,
+      "grad_norm": 3.752470753954954,
+      "learning_rate": 1.9541091603605508e-05,
+      "loss": 3.3077,
+      "step": 645
+    },
+    {
+      "epoch": 0.12428454619787407,
+      "grad_norm": 3.6460522842740484,
+      "learning_rate": 1.95392234976065e-05,
+      "loss": 3.4217,
+      "step": 646
+    },
+    {
+      "epoch": 0.12447693713626089,
+      "grad_norm": 4.16028122115183,
+      "learning_rate": 1.9537351686683003e-05,
+      "loss": 3.1461,
+      "step": 647
+    },
+    {
+      "epoch": 0.12466932807464769,
+      "grad_norm": 3.868374581910461,
+      "learning_rate": 1.9535476171562014e-05,
+      "loss": 3.3216,
+      "step": 648
+    },
+    {
+      "epoch": 0.12486171901303449,
+      "grad_norm": 3.9915796844062936,
+      "learning_rate": 1.9533596952971955e-05,
+      "loss": 3.2443,
+      "step": 649
+    },
+    {
+      "epoch": 0.1250541099514213,
+      "grad_norm": 2.6706376971778782,
+      "learning_rate": 1.9531714031642698e-05,
+      "loss": 2.8459,
+      "step": 650
+    },
+    {
+      "epoch": 0.1252465008898081,
+      "grad_norm": 3.7654060610681106,
+      "learning_rate": 1.9529827408305542e-05,
+      "loss": 3.1009,
+      "step": 651
+    },
+    {
+      "epoch": 0.1254388918281949,
+      "grad_norm": 3.7750576590094598,
+      "learning_rate": 1.9527937083693233e-05,
+      "loss": 3.4383,
+      "step": 652
+    },
+    {
+      "epoch": 0.1256312827665817,
+      "grad_norm": 3.42677669440787,
+      "learning_rate": 1.952604305853995e-05,
+      "loss": 3.3553,
+      "step": 653
+    },
+    {
+      "epoch": 0.1258236737049685,
+      "grad_norm": 3.7835833156376224,
+      "learning_rate": 1.9524145333581315e-05,
+      "loss": 3.3097,
+      "step": 654
+    },
+    {
+      "epoch": 0.1260160646433553,
+      "grad_norm": 4.089706132244584,
+      "learning_rate": 1.9522243909554375e-05,
+      "loss": 3.2853,
+      "step": 655
+    },
+    {
+      "epoch": 0.1262084555817421,
+      "grad_norm": 3.949688316978291,
+      "learning_rate": 1.952033878719763e-05,
+      "loss": 3.3913,
+      "step": 656
+    },
+    {
+      "epoch": 0.1264008465201289,
+      "grad_norm": 3.8437080570536066,
+      "learning_rate": 1.9518429967251e-05,
+      "loss": 3.5656,
+      "step": 657
+    },
+    {
+      "epoch": 0.1265932374585157,
+      "grad_norm": 3.738228445256269,
+      "learning_rate": 1.9516517450455853e-05,
+      "loss": 3.3455,
+      "step": 658
+    },
+    {
+      "epoch": 0.1267856283969025,
+      "grad_norm": 3.5275638613357785,
+      "learning_rate": 1.951460123755499e-05,
+      "loss": 3.1694,
+      "step": 659
+    },
+    {
+      "epoch": 0.12697801933528932,
+      "grad_norm": 3.492521663026377,
+      "learning_rate": 1.9512681329292635e-05,
+      "loss": 3.417,
+      "step": 660
+    },
+    {
+      "epoch": 0.12717041027367612,
+      "grad_norm": 3.7574626098038575,
+      "learning_rate": 1.951075772641447e-05,
+      "loss": 3.3815,
+      "step": 661
+    },
+    {
+      "epoch": 0.12736280121206292,
+      "grad_norm": 4.255995538005145,
+      "learning_rate": 1.95088304296676e-05,
+      "loss": 3.2917,
+      "step": 662
+    },
+    {
+      "epoch": 0.12755519215044972,
+      "grad_norm": 3.7007107239886765,
+      "learning_rate": 1.950689943980056e-05,
+      "loss": 3.4311,
+      "step": 663
+    },
+    {
+      "epoch": 0.12774758308883652,
+      "grad_norm": 2.498581554478525,
+      "learning_rate": 1.9504964757563322e-05,
+      "loss": 2.6608,
+      "step": 664
+    },
+    {
+      "epoch": 0.12793997402722332,
+      "grad_norm": 3.9948503306297964,
+      "learning_rate": 1.95030263837073e-05,
+      "loss": 3.373,
+      "step": 665
+    },
+    {
+      "epoch": 0.12813236496561012,
+      "grad_norm": 4.137307729349788,
+      "learning_rate": 1.9501084318985335e-05,
+      "loss": 3.6572,
+      "step": 666
+    },
+    {
+      "epoch": 0.12832475590399692,
+      "grad_norm": 2.5627581654685305,
+      "learning_rate": 1.94991385641517e-05,
+      "loss": 2.819,
+      "step": 667
+    },
+    {
+      "epoch": 0.12851714684238372,
+      "grad_norm": 4.416542842567331,
+      "learning_rate": 1.9497189119962105e-05,
+      "loss": 3.6537,
+      "step": 668
+    },
+    {
+      "epoch": 0.12870953778077052,
+      "grad_norm": 3.6356552807572022,
+      "learning_rate": 1.9495235987173693e-05,
+      "loss": 3.1776,
+      "step": 669
+    },
+    {
+      "epoch": 0.12890192871915732,
+      "grad_norm": 3.8183209393738746,
+      "learning_rate": 1.949327916654504e-05,
+      "loss": 3.5277,
+      "step": 670
+    },
+    {
+      "epoch": 0.12909431965754414,
+      "grad_norm": 4.010072048242557,
+      "learning_rate": 1.949131865883614e-05,
+      "loss": 3.4049,
+      "step": 671
+    },
+    {
+      "epoch": 0.12928671059593094,
+      "grad_norm": 4.125187258206206,
+      "learning_rate": 1.948935446480845e-05,
+      "loss": 3.48,
+      "step": 672
+    },
+    {
+      "epoch": 0.12947910153431774,
+      "grad_norm": 3.6134305859619014,
+      "learning_rate": 1.948738658522483e-05,
+      "loss": 3.1533,
+      "step": 673
+    },
+    {
+      "epoch": 0.12967149247270454,
+      "grad_norm": 3.9003512110340006,
+      "learning_rate": 1.9485415020849583e-05,
+      "loss": 3.3644,
+      "step": 674
+    },
+    {
+      "epoch": 0.12986388341109134,
+      "grad_norm": 4.182717490133172,
+      "learning_rate": 1.9483439772448444e-05,
+      "loss": 3.5639,
+      "step": 675
+    },
+    {
+      "epoch": 0.13005627434947814,
+      "grad_norm": 3.916550310845153,
+      "learning_rate": 1.9481460840788573e-05,
+      "loss": 3.2973,
+      "step": 676
+    },
+    {
+      "epoch": 0.13024866528786494,
+      "grad_norm": 3.8116833556327543,
+      "learning_rate": 1.9479478226638565e-05,
+      "loss": 3.5043,
+      "step": 677
+    },
+    {
+      "epoch": 0.13044105622625174,
+      "grad_norm": 3.864092596525119,
+      "learning_rate": 1.947749193076845e-05,
+      "loss": 3.4497,
+      "step": 678
+    },
+    {
+      "epoch": 0.13063344716463854,
+      "grad_norm": 3.5762321330199254,
+      "learning_rate": 1.9475501953949674e-05,
+      "loss": 3.3806,
+      "step": 679
+    },
+    {
+      "epoch": 0.13082583810302534,
+      "grad_norm": 3.6986221344087125,
+      "learning_rate": 1.9473508296955126e-05,
+      "loss": 3.6081,
+      "step": 680
+    },
+    {
+      "epoch": 0.13101822904141214,
+      "grad_norm": 4.271448227010434,
+      "learning_rate": 1.9471510960559122e-05,
+      "loss": 3.5149,
+      "step": 681
+    },
+    {
+      "epoch": 0.13121061997979896,
+      "grad_norm": 3.877276774234732,
+      "learning_rate": 1.9469509945537395e-05,
+      "loss": 3.5408,
+      "step": 682
+    },
+    {
+      "epoch": 0.13140301091818576,
+      "grad_norm": 3.578673060846642,
+      "learning_rate": 1.9467505252667126e-05,
+      "loss": 3.0981,
+      "step": 683
+    },
+    {
+      "epoch": 0.13159540185657256,
+      "grad_norm": 2.553492531390485,
+      "learning_rate": 1.9465496882726913e-05,
+      "loss": 2.7914,
+      "step": 684
+    },
+    {
+      "epoch": 0.13178779279495936,
+      "grad_norm": 3.56489190081433,
+      "learning_rate": 1.946348483649678e-05,
+      "loss": 3.1511,
+      "step": 685
+    },
+    {
+      "epoch": 0.13198018373334616,
+      "grad_norm": 4.044968834534828,
+      "learning_rate": 1.9461469114758184e-05,
+      "loss": 3.5398,
+      "step": 686
+    },
+    {
+      "epoch": 0.13217257467173296,
+      "grad_norm": 3.402806100528008,
+      "learning_rate": 1.9459449718294008e-05,
+      "loss": 3.3305,
+      "step": 687
+    },
+    {
+      "epoch": 0.13236496561011976,
+      "grad_norm": 3.794451690110137,
+      "learning_rate": 1.945742664788856e-05,
+      "loss": 3.1684,
+      "step": 688
+    },
+    {
+      "epoch": 0.13255735654850656,
+      "grad_norm": 3.635460009224197,
+      "learning_rate": 1.9455399904327585e-05,
+      "loss": 3.4611,
+      "step": 689
+    },
+    {
+      "epoch": 0.13274974748689336,
+      "grad_norm": 4.383861662078943,
+      "learning_rate": 1.945336948839824e-05,
+      "loss": 3.8425,
+      "step": 690
+    },
+    {
+      "epoch": 0.13294213842528016,
+      "grad_norm": 3.7310946825455833,
+      "learning_rate": 1.9451335400889114e-05,
+      "loss": 3.3041,
+      "step": 691
+    },
+    {
+      "epoch": 0.13313452936366696,
+      "grad_norm": 3.8918376570838316,
+      "learning_rate": 1.944929764259023e-05,
+      "loss": 3.2306,
+      "step": 692
+    },
+    {
+      "epoch": 0.1333269203020538,
+      "grad_norm": 3.4184232592136254,
+      "learning_rate": 1.9447256214293026e-05,
+      "loss": 3.1729,
+      "step": 693
+    },
+    {
+      "epoch": 0.13351931124044059,
+      "grad_norm": 3.8360039829149746,
+      "learning_rate": 1.9445211116790365e-05,
+      "loss": 3.5258,
+      "step": 694
+    },
+    {
+      "epoch": 0.13371170217882739,
+      "grad_norm": 3.6329861607960923,
+      "learning_rate": 1.9443162350876544e-05,
+      "loss": 3.3225,
+      "step": 695
+    },
+    {
+      "epoch": 0.13390409311721418,
+      "grad_norm": 4.067419440858311,
+      "learning_rate": 1.944110991734728e-05,
+      "loss": 3.5562,
+      "step": 696
+    },
+    {
+      "epoch": 0.13409648405560098,
+      "grad_norm": 3.6798213394352524,
+      "learning_rate": 1.9439053816999717e-05,
+      "loss": 3.0848,
+      "step": 697
+    },
+    {
+      "epoch": 0.13428887499398778,
+      "grad_norm": 3.6443339968432893,
+      "learning_rate": 1.9436994050632415e-05,
+      "loss": 3.3822,
+      "step": 698
+    },
+    {
+      "epoch": 0.13448126593237458,
+      "grad_norm": 3.7306783750528476,
+      "learning_rate": 1.9434930619045367e-05,
+      "loss": 3.1656,
+      "step": 699
+    },
+    {
+      "epoch": 0.13467365687076138,
+      "grad_norm": 3.557228183092432,
+      "learning_rate": 1.9432863523039986e-05,
+      "loss": 3.0755,
+      "step": 700
+    },
+    {
+      "epoch": 0.13486604780914818,
+      "grad_norm": 3.747839542517907,
+      "learning_rate": 1.9430792763419105e-05,
+      "loss": 3.2924,
+      "step": 701
+    },
+    {
+      "epoch": 0.13505843874753498,
+      "grad_norm": 3.9665349693244867,
+      "learning_rate": 1.942871834098699e-05,
+      "loss": 3.4184,
+      "step": 702
+    },
+    {
+      "epoch": 0.13525082968592178,
+      "grad_norm": 3.849550700227978,
+      "learning_rate": 1.9426640256549313e-05,
+      "loss": 3.3251,
+      "step": 703
+    },
+    {
+      "epoch": 0.1354432206243086,
+      "grad_norm": 3.5603689192754153,
+      "learning_rate": 1.9424558510913186e-05,
+      "loss": 3.223,
+      "step": 704
+    },
+    {
+      "epoch": 0.1356356115626954,
+      "grad_norm": 3.7003136310242706,
+      "learning_rate": 1.9422473104887133e-05,
+      "loss": 2.9962,
+      "step": 705
+    },
+    {
+      "epoch": 0.1358280025010822,
+      "grad_norm": 3.8312447936543315,
+      "learning_rate": 1.9420384039281103e-05,
+      "loss": 3.4346,
+      "step": 706
+    },
+    {
+      "epoch": 0.136020393439469,
+      "grad_norm": 4.053797540711221,
+      "learning_rate": 1.941829131490646e-05,
+      "loss": 3.4218,
+      "step": 707
+    },
+    {
+      "epoch": 0.1362127843778558,
+      "grad_norm": 3.4353495433536447,
+      "learning_rate": 1.9416194932576e-05,
+      "loss": 3.2556,
+      "step": 708
+    },
+    {
+      "epoch": 0.1364051753162426,
+      "grad_norm": 4.339016017059798,
+      "learning_rate": 1.941409489310393e-05,
+      "loss": 3.3998,
+      "step": 709
+    },
+    {
+      "epoch": 0.1365975662546294,
+      "grad_norm": 3.536292890109313,
+      "learning_rate": 1.9411991197305878e-05,
+      "loss": 3.2612,
+      "step": 710
+    },
+    {
+      "epoch": 0.1367899571930162,
+      "grad_norm": 3.7496824145582077,
+      "learning_rate": 1.9409883845998905e-05,
+      "loss": 3.4701,
+      "step": 711
+    },
+    {
+      "epoch": 0.136982348131403,
+      "grad_norm": 3.662504266130211,
+      "learning_rate": 1.9407772840001473e-05,
+      "loss": 3.2443,
+      "step": 712
+    },
+    {
+      "epoch": 0.1371747390697898,
+      "grad_norm": 4.053091142955789,
+      "learning_rate": 1.9405658180133477e-05,
+      "loss": 3.2616,
+      "step": 713
+    },
+    {
+      "epoch": 0.1373671300081766,
+      "grad_norm": 4.172510479918806,
+      "learning_rate": 1.9403539867216226e-05,
+      "loss": 3.483,
+      "step": 714
+    },
+    {
+      "epoch": 0.13755952094656343,
+      "grad_norm": 4.079493366714864,
+      "learning_rate": 1.9401417902072447e-05,
+      "loss": 3.5146,
+      "step": 715
+    },
+    {
+      "epoch": 0.13775191188495023,
+      "grad_norm": 3.758447147841402,
+      "learning_rate": 1.9399292285526286e-05,
+      "loss": 3.2306,
+      "step": 716
+    },
+    {
+      "epoch": 0.13794430282333703,
+      "grad_norm": 3.4457034328816123,
+      "learning_rate": 1.939716301840331e-05,
+      "loss": 3.4389,
+      "step": 717
+    },
+    {
+      "epoch": 0.13813669376172383,
+      "grad_norm": 3.888483362135414,
+      "learning_rate": 1.9395030101530504e-05,
+      "loss": 3.4809,
+      "step": 718
+    },
+    {
+      "epoch": 0.13832908470011063,
+      "grad_norm": 3.7402393687386666,
+      "learning_rate": 1.939289353573626e-05,
+      "loss": 3.4792,
+      "step": 719
+    },
+    {
+      "epoch": 0.13852147563849743,
+      "grad_norm": 3.6208508687947045,
+      "learning_rate": 1.9390753321850404e-05,
+      "loss": 3.2297,
+      "step": 720
+    },
+    {
+      "epoch": 0.13871386657688423,
+      "grad_norm": 3.8793052538889254,
+      "learning_rate": 1.938860946070417e-05,
+      "loss": 3.402,
+      "step": 721
+    },
+    {
+      "epoch": 0.13890625751527103,
+      "grad_norm": 3.7412116322028406,
+      "learning_rate": 1.93864619531302e-05,
+      "loss": 3.5437,
+      "step": 722
+    },
+    {
+      "epoch": 0.13909864845365782,
+      "grad_norm": 3.7711033112138748,
+      "learning_rate": 1.9384310799962575e-05,
+      "loss": 3.5282,
+      "step": 723
+    },
+    {
+      "epoch": 0.13929103939204462,
+      "grad_norm": 3.620497707201849,
+      "learning_rate": 1.9382156002036764e-05,
+      "loss": 3.2366,
+      "step": 724
+    },
+    {
+      "epoch": 0.13948343033043142,
+      "grad_norm": 3.896487181595723,
+      "learning_rate": 1.9379997560189677e-05,
+      "loss": 3.5194,
+      "step": 725
+    },
+    {
+      "epoch": 0.13967582126881825,
+      "grad_norm": 2.6335246012979137,
+      "learning_rate": 1.937783547525962e-05,
+      "loss": 2.7729,
+      "step": 726
+    },
+    {
+      "epoch": 0.13986821220720505,
+      "grad_norm": 4.24678616346325,
+      "learning_rate": 1.9375669748086326e-05,
+      "loss": 3.3255,
+      "step": 727
+    },
+    {
+      "epoch": 0.14006060314559185,
+      "grad_norm": 3.559838040468837,
+      "learning_rate": 1.937350037951094e-05,
+      "loss": 3.2986,
+      "step": 728
+    },
+    {
+      "epoch": 0.14025299408397865,
+      "grad_norm": 3.8991692787033263,
+      "learning_rate": 1.9371327370376018e-05,
+      "loss": 3.2577,
+      "step": 729
+    },
+    {
+      "epoch": 0.14044538502236545,
+      "grad_norm": 3.7960381817952173,
+      "learning_rate": 1.936915072152553e-05,
+      "loss": 3.647,
+      "step": 730
+    },
+    {
+      "epoch": 0.14063777596075225,
+      "grad_norm": 3.6560689947531384,
+      "learning_rate": 1.936697043380486e-05,
+      "loss": 3.2781,
+      "step": 731
+    },
+    {
+      "epoch": 0.14083016689913905,
+      "grad_norm": 3.888334612613286,
+      "learning_rate": 1.936478650806081e-05,
+      "loss": 3.3404,
+      "step": 732
+    },
+    {
+      "epoch": 0.14102255783752585,
+      "grad_norm": 3.803497422106739,
+      "learning_rate": 1.936259894514159e-05,
+      "loss": 3.5095,
+      "step": 733
+    },
+    {
+      "epoch": 0.14121494877591265,
+      "grad_norm": 3.6169278962979297,
+      "learning_rate": 1.9360407745896828e-05,
+      "loss": 3.0966,
+      "step": 734
+    },
+    {
+      "epoch": 0.14140733971429945,
+      "grad_norm": 3.5311152960470764,
+      "learning_rate": 1.9358212911177556e-05,
+      "loss": 3.0481,
+      "step": 735
+    },
+    {
+      "epoch": 0.14159973065268625,
+      "grad_norm": 3.578694934300673,
+      "learning_rate": 1.935601444183622e-05,
+      "loss": 3.7095,
+      "step": 736
+    },
+    {
+      "epoch": 0.14179212159107307,
+      "grad_norm": 4.174436819000354,
+      "learning_rate": 1.935381233872669e-05,
+      "loss": 3.4741,
+      "step": 737
+    },
+    {
+      "epoch": 0.14198451252945987,
+      "grad_norm": 3.677060181620812,
+      "learning_rate": 1.935160660270423e-05,
+      "loss": 3.2872,
+      "step": 738
+    },
+    {
+      "epoch": 0.14217690346784667,
+      "grad_norm": 3.6499246704739514,
+      "learning_rate": 1.934939723462552e-05,
+      "loss": 3.3332,
+      "step": 739
+    },
+    {
+      "epoch": 0.14236929440623347,
+      "grad_norm": 3.8323562015057435,
+      "learning_rate": 1.9347184235348663e-05,
+      "loss": 3.2478,
+      "step": 740
+    },
+    {
+      "epoch": 0.14256168534462027,
+      "grad_norm": 4.284226916025821,
+      "learning_rate": 1.9344967605733154e-05,
+      "loss": 3.1334,
+      "step": 741
+    },
+    {
+      "epoch": 0.14275407628300707,
+      "grad_norm": 3.9858530822420377,
+      "learning_rate": 1.934274734663991e-05,
+      "loss": 3.5429,
+      "step": 742
+    },
+    {
+      "epoch": 0.14294646722139387,
+      "grad_norm": 3.575775617074263,
+      "learning_rate": 1.934052345893125e-05,
+      "loss": 3.3682,
+      "step": 743
+    },
+    {
+      "epoch": 0.14313885815978067,
+      "grad_norm": 3.6995888155121297,
+      "learning_rate": 1.9338295943470915e-05,
+      "loss": 3.1854,
+      "step": 744
+    },
+    {
+      "epoch": 0.14333124909816747,
+      "grad_norm": 3.887903273673699,
+      "learning_rate": 1.9336064801124034e-05,
+      "loss": 3.3596,
+      "step": 745
+    },
+    {
+      "epoch": 0.14352364003655427,
+      "grad_norm": 4.241768895784738,
+      "learning_rate": 1.933383003275717e-05,
+      "loss": 3.4625,
+      "step": 746
+    },
+    {
+      "epoch": 0.14371603097494107,
+      "grad_norm": 3.8627809253487757,
+      "learning_rate": 1.933159163923827e-05,
+      "loss": 3.5094,
+      "step": 747
+    },
+    {
+      "epoch": 0.1439084219133279,
+      "grad_norm": 4.081253501154426,
+      "learning_rate": 1.9329349621436708e-05,
+      "loss": 3.3011,
+      "step": 748
+    },
+    {
+      "epoch": 0.1441008128517147,
+      "grad_norm": 3.878585025803719,
+      "learning_rate": 1.9327103980223255e-05,
+      "loss": 3.3285,
+      "step": 749
+    },
+    {
+      "epoch": 0.1442932037901015,
+      "grad_norm": 2.264241580227417,
+      "learning_rate": 1.932485471647009e-05,
+      "loss": 2.7237,
+      "step": 750
+    },
+    {
+      "epoch": 0.1444855947284883,
+      "grad_norm": 3.7569892451981874,
+      "learning_rate": 1.9322601831050804e-05,
+      "loss": 3.4087,
+      "step": 751
+    },
+    {
+      "epoch": 0.1446779856668751,
+      "grad_norm": 3.7421702643595043,
+      "learning_rate": 1.9320345324840396e-05,
+      "loss": 3.1942,
+      "step": 752
+    },
+    {
+      "epoch": 0.1448703766052619,
+      "grad_norm": 3.6678353304277027,
+      "learning_rate": 1.9318085198715257e-05,
+      "loss": 3.3187,
+      "step": 753
+    },
+    {
+      "epoch": 0.1450627675436487,
+      "grad_norm": 3.7663201248679274,
+      "learning_rate": 1.93158214535532e-05,
+      "loss": 3.0819,
+      "step": 754
+    },
+    {
+      "epoch": 0.1452551584820355,
+      "grad_norm": 3.6575782975706077,
+      "learning_rate": 1.9313554090233436e-05,
+      "loss": 3.5002,
+      "step": 755
+    },
+    {
+      "epoch": 0.1454475494204223,
+      "grad_norm": 3.776682275780765,
+      "learning_rate": 1.9311283109636586e-05,
+      "loss": 3.3964,
+      "step": 756
+    },
+    {
+      "epoch": 0.1456399403588091,
+      "grad_norm": 3.82053468889084,
+      "learning_rate": 1.9309008512644668e-05,
+      "loss": 3.3724,
+      "step": 757
+    },
+    {
+      "epoch": 0.1458323312971959,
+      "grad_norm": 3.7223518058783966,
+      "learning_rate": 1.930673030014111e-05,
+      "loss": 3.0924,
+      "step": 758
+    },
+    {
+      "epoch": 0.14602472223558272,
+      "grad_norm": 2.3879939689369634,
+      "learning_rate": 1.930444847301075e-05,
+      "loss": 2.5766,
+      "step": 759
+    },
+    {
+      "epoch": 0.14621711317396952,
+      "grad_norm": 3.7195509758283545,
+      "learning_rate": 1.9302163032139813e-05,
+      "loss": 3.2034,
+      "step": 760
+    },
+    {
+      "epoch": 0.14640950411235631,
+      "grad_norm": 3.772175225155343,
+      "learning_rate": 1.9299873978415947e-05,
+      "loss": 3.3691,
+      "step": 761
+    },
+    {
+      "epoch": 0.14660189505074311,
+      "grad_norm": 3.824221223156336,
+      "learning_rate": 1.9297581312728187e-05,
+      "loss": 3.325,
+      "step": 762
+    },
+    {
+      "epoch": 0.1467942859891299,
+      "grad_norm": 4.085970393246644,
+      "learning_rate": 1.929528503596698e-05,
+      "loss": 3.3356,
+      "step": 763
+    },
+    {
+      "epoch": 0.1469866769275167,
+      "grad_norm": 3.830859645596577,
+      "learning_rate": 1.929298514902418e-05,
+      "loss": 3.4829,
+      "step": 764
+    },
+    {
+      "epoch": 0.1471790678659035,
+      "grad_norm": 3.858220459668656,
+      "learning_rate": 1.929068165279303e-05,
+      "loss": 3.5516,
+      "step": 765
+    },
+    {
+      "epoch": 0.1473714588042903,
+      "grad_norm": 3.8698160028161075,
+      "learning_rate": 1.928837454816818e-05,
+      "loss": 3.1732,
+      "step": 766
+    },
+    {
+      "epoch": 0.1475638497426771,
+      "grad_norm": 4.065574889213117,
+      "learning_rate": 1.9286063836045687e-05,
+      "loss": 3.3727,
+      "step": 767
+    },
+    {
+      "epoch": 0.1477562406810639,
+      "grad_norm": 3.80430456655814,
+      "learning_rate": 1.9283749517323e-05,
+      "loss": 3.2601,
+      "step": 768
+    },
+    {
+      "epoch": 0.1479486316194507,
+      "grad_norm": 3.6921301442376,
+      "learning_rate": 1.928143159289898e-05,
+      "loss": 3.3608,
+      "step": 769
+    },
+    {
+      "epoch": 0.14814102255783754,
+      "grad_norm": 3.7240514729021545,
+      "learning_rate": 1.927911006367388e-05,
+      "loss": 3.1868,
+      "step": 770
+    },
+    {
+      "epoch": 0.14833341349622434,
+      "grad_norm": 3.7791800181227133,
+      "learning_rate": 1.927678493054935e-05,
+      "loss": 3.4391,
+      "step": 771
+    },
+    {
+      "epoch": 0.14852580443461114,
+      "grad_norm": 3.6797186105085355,
+      "learning_rate": 1.9274456194428454e-05,
+      "loss": 3.2541,
+      "step": 772
+    },
+    {
+      "epoch": 0.14871819537299794,
+      "grad_norm": 3.5423902192397385,
+      "learning_rate": 1.9272123856215643e-05,
+      "loss": 3.4177,
+      "step": 773
+    },
+    {
+      "epoch": 0.14891058631138474,
+      "grad_norm": 3.768728547497981,
+      "learning_rate": 1.9269787916816764e-05,
+      "loss": 3.484,
+      "step": 774
+    },
+    {
+      "epoch": 0.14910297724977153,
+      "grad_norm": 3.731033784460436,
+      "learning_rate": 1.9267448377139074e-05,
+      "loss": 3.2785,
+      "step": 775
+    },
+    {
+      "epoch": 0.14929536818815833,
+      "grad_norm": 3.7705406511477304,
+      "learning_rate": 1.9265105238091227e-05,
+      "loss": 3.1696,
+      "step": 776
+    },
+    {
+      "epoch": 0.14948775912654513,
+      "grad_norm": 3.8852857377162726,
+      "learning_rate": 1.9262758500583265e-05,
+      "loss": 3.4354,
+      "step": 777
+    },
+    {
+      "epoch": 0.14968015006493193,
+      "grad_norm": 4.0500020663294825,
+      "learning_rate": 1.9260408165526638e-05,
+      "loss": 2.9902,
+      "step": 778
+    },
+    {
+      "epoch": 0.14987254100331873,
+      "grad_norm": 3.6378494464710163,
+      "learning_rate": 1.9258054233834184e-05,
+      "loss": 3.3949,
+      "step": 779
+    },
+    {
+      "epoch": 0.15006493194170556,
+      "grad_norm": 3.9142096255271226,
+      "learning_rate": 1.9255696706420147e-05,
+      "loss": 3.4239,
+      "step": 780
+    },
+    {
+      "epoch": 0.15025732288009236,
+      "grad_norm": 3.8983533801081323,
+      "learning_rate": 1.9253335584200165e-05,
+      "loss": 3.4702,
+      "step": 781
+    },
+    {
+      "epoch": 0.15044971381847916,
+      "grad_norm": 3.893885619025196,
+      "learning_rate": 1.9250970868091268e-05,
+      "loss": 3.3291,
+      "step": 782
+    },
+    {
+      "epoch": 0.15064210475686596,
+      "grad_norm": 3.6428061125371998,
+      "learning_rate": 1.924860255901188e-05,
+      "loss": 3.3607,
+      "step": 783
+    },
+    {
+      "epoch": 0.15083449569525276,
+      "grad_norm": 3.8192126028973266,
+      "learning_rate": 1.9246230657881834e-05,
+      "loss": 3.4876,
+      "step": 784
+    },
+    {
+      "epoch": 0.15102688663363956,
+      "grad_norm": 3.9479296480743056,
+      "learning_rate": 1.9243855165622345e-05,
+      "loss": 3.35,
+      "step": 785
+    },
+    {
+      "epoch": 0.15121927757202636,
+      "grad_norm": 4.026783318469912,
+      "learning_rate": 1.9241476083156026e-05,
+      "loss": 3.2086,
+      "step": 786
+    },
+    {
+      "epoch": 0.15141166851041316,
+      "grad_norm": 4.020018656643367,
+      "learning_rate": 1.9239093411406885e-05,
+      "loss": 3.5327,
+      "step": 787
+    },
+    {
+      "epoch": 0.15160405944879995,
+      "grad_norm": 3.8216559980768015,
+      "learning_rate": 1.9236707151300326e-05,
+      "loss": 3.3721,
+      "step": 788
+    },
+    {
+      "epoch": 0.15179645038718675,
+      "grad_norm": 3.646008784736545,
+      "learning_rate": 1.9234317303763145e-05,
+      "loss": 3.4488,
+      "step": 789
+    },
+    {
+      "epoch": 0.15198884132557355,
+      "grad_norm": 3.3067626262122576,
+      "learning_rate": 1.9231923869723528e-05,
+      "loss": 3.2627,
+      "step": 790
+    },
+    {
+      "epoch": 0.15218123226396038,
+      "grad_norm": 3.669772561921337,
+      "learning_rate": 1.922952685011106e-05,
+      "loss": 3.4919,
+      "step": 791
+    },
+    {
+      "epoch": 0.15237362320234718,
+      "grad_norm": 3.9197363477669973,
+      "learning_rate": 1.9227126245856716e-05,
+      "loss": 3.3121,
+      "step": 792
+    },
+    {
+      "epoch": 0.15256601414073398,
+      "grad_norm": 3.908767151103692,
+      "learning_rate": 1.922472205789286e-05,
+      "loss": 3.2103,
+      "step": 793
+    },
+    {
+      "epoch": 0.15275840507912078,
+      "grad_norm": 3.8164501500455916,
+      "learning_rate": 1.9222314287153255e-05,
+      "loss": 3.3073,
+      "step": 794
+    },
+    {
+      "epoch": 0.15295079601750758,
+      "grad_norm": 3.936497888718697,
+      "learning_rate": 1.9219902934573048e-05,
+      "loss": 3.4139,
+      "step": 795
+    },
+    {
+      "epoch": 0.15314318695589438,
+      "grad_norm": 3.549401784693579,
+      "learning_rate": 1.9217488001088784e-05,
+      "loss": 3.3944,
+      "step": 796
+    },
+    {
+      "epoch": 0.15333557789428118,
+      "grad_norm": 3.5863361824452897,
+      "learning_rate": 1.9215069487638396e-05,
+      "loss": 3.4804,
+      "step": 797
+    },
+    {
+      "epoch": 0.15352796883266798,
+      "grad_norm": 3.86456456453494,
+      "learning_rate": 1.92126473951612e-05,
+      "loss": 3.35,
+      "step": 798
+    },
+    {
+      "epoch": 0.15372035977105478,
+      "grad_norm": 3.5853842020186346,
+      "learning_rate": 1.921022172459791e-05,
+      "loss": 3.4303,
+      "step": 799
+    },
+    {
+      "epoch": 0.15391275070944158,
+      "grad_norm": 3.759202988154235,
+      "learning_rate": 1.920779247689064e-05,
+      "loss": 3.2807,
+      "step": 800
+    },
+    {
+      "epoch": 0.15410514164782838,
+      "grad_norm": 3.8634866448774097,
+      "learning_rate": 1.9205359652982868e-05,
+      "loss": 3.5123,
+      "step": 801
+    },
+    {
+      "epoch": 0.1542975325862152,
+      "grad_norm": 3.709831910743789,
+      "learning_rate": 1.9202923253819482e-05,
+      "loss": 3.3008,
+      "step": 802
+    },
+    {
+      "epoch": 0.154489923524602,
+      "grad_norm": 3.868097731409169,
+      "learning_rate": 1.920048328034675e-05,
+      "loss": 3.4325,
+      "step": 803
+    },
+    {
+      "epoch": 0.1546823144629888,
+      "grad_norm": 3.5057194470472313,
+      "learning_rate": 1.9198039733512326e-05,
+      "loss": 3.0671,
+      "step": 804
+    },
+    {
+      "epoch": 0.1548747054013756,
+      "grad_norm": 4.065070841854985,
+      "learning_rate": 1.9195592614265262e-05,
+      "loss": 3.3361,
+      "step": 805
+    },
+    {
+      "epoch": 0.1550670963397624,
+      "grad_norm": 3.8601776567209574,
+      "learning_rate": 1.9193141923555984e-05,
+      "loss": 3.4139,
+      "step": 806
+    },
+    {
+      "epoch": 0.1552594872781492,
+      "grad_norm": 3.796396504413359,
+      "learning_rate": 1.919068766233632e-05,
+      "loss": 3.2524,
+      "step": 807
+    },
+    {
+      "epoch": 0.155451878216536,
+      "grad_norm": 3.723651764577327,
+      "learning_rate": 1.9188229831559468e-05,
+      "loss": 3.2077,
+      "step": 808
+    },
+    {
+      "epoch": 0.1556442691549228,
+      "grad_norm": 3.7905005655878683,
+      "learning_rate": 1.9185768432180026e-05,
+      "loss": 3.2734,
+      "step": 809
+    },
+    {
+      "epoch": 0.1558366600933096,
+      "grad_norm": 4.007665728511415,
+      "learning_rate": 1.9183303465153972e-05,
+      "loss": 3.4597,
+      "step": 810
+    },
+    {
+      "epoch": 0.1560290510316964,
+      "grad_norm": 3.57268308815684,
+      "learning_rate": 1.9180834931438673e-05,
+      "loss": 3.3218,
+      "step": 811
+    },
+    {
+      "epoch": 0.1562214419700832,
+      "grad_norm": 3.5731709832978136,
+      "learning_rate": 1.917836283199288e-05,
+      "loss": 3.299,
+      "step": 812
+    },
+    {
+      "epoch": 0.15641383290847002,
+      "grad_norm": 4.046193739032613,
+      "learning_rate": 1.917588716777672e-05,
+      "loss": 3.1967,
+      "step": 813
+    },
+    {
+      "epoch": 0.15660622384685682,
+      "grad_norm": 3.60142220773021,
+      "learning_rate": 1.917340793975172e-05,
+      "loss": 3.2939,
+      "step": 814
+    },
+    {
+      "epoch": 0.15679861478524362,
+      "grad_norm": 4.009842050885506,
+      "learning_rate": 1.917092514888078e-05,
+      "loss": 3.3291,
+      "step": 815
+    },
+    {
+      "epoch": 0.15699100572363042,
+      "grad_norm": 3.5437507673421753,
+      "learning_rate": 1.9168438796128193e-05,
+      "loss": 3.536,
+      "step": 816
+    },
+    {
+      "epoch": 0.15718339666201722,
+      "grad_norm": 3.835258677960973,
+      "learning_rate": 1.9165948882459623e-05,
+      "loss": 3.0638,
+      "step": 817
+    },
+    {
+      "epoch": 0.15737578760040402,
+      "grad_norm": 3.8349276880441874,
+      "learning_rate": 1.9163455408842123e-05,
+      "loss": 3.4811,
+      "step": 818
+    },
+    {
+      "epoch": 0.15756817853879082,
+      "grad_norm": 3.90580251044361,
+      "learning_rate": 1.9160958376244138e-05,
+      "loss": 3.0676,
+      "step": 819
+    },
+    {
+      "epoch": 0.15776056947717762,
+      "grad_norm": 4.250496588205761,
+      "learning_rate": 1.9158457785635478e-05,
+      "loss": 3.514,
+      "step": 820
+    },
+    {
+      "epoch": 0.15795296041556442,
+      "grad_norm": 3.988632703709733,
+      "learning_rate": 1.915595363798735e-05,
+      "loss": 3.4423,
+      "step": 821
+    },
+    {
+      "epoch": 0.15814535135395122,
+      "grad_norm": 4.317367889090759,
+      "learning_rate": 1.915344593427233e-05,
+      "loss": 3.6166,
+      "step": 822
+    },
+    {
+      "epoch": 0.15833774229233802,
+      "grad_norm": 3.757929723726165,
+      "learning_rate": 1.9150934675464384e-05,
+      "loss": 3.2099,
+      "step": 823
+    },
+    {
+      "epoch": 0.15853013323072485,
+      "grad_norm": 3.6882079831784362,
+      "learning_rate": 1.9148419862538858e-05,
+      "loss": 2.961,
+      "step": 824
+    },
+    {
+      "epoch": 0.15872252416911165,
+      "grad_norm": 2.4232185708541794,
+      "learning_rate": 1.9145901496472474e-05,
+      "loss": 2.8242,
+      "step": 825
+    },
+    {
+      "epoch": 0.15891491510749844,
+      "grad_norm": 4.066261517913169,
+      "learning_rate": 1.9143379578243335e-05,
+      "loss": 3.4002,
+      "step": 826
+    },
+    {
+      "epoch": 0.15910730604588524,
+      "grad_norm": 4.077593163631176,
+      "learning_rate": 1.914085410883093e-05,
+      "loss": 2.9748,
+      "step": 827
+    },
+    {
+      "epoch": 0.15929969698427204,
+      "grad_norm": 4.010684029970076,
+      "learning_rate": 1.9138325089216118e-05,
+      "loss": 3.2321,
+      "step": 828
+    },
+    {
+      "epoch": 0.15949208792265884,
+      "grad_norm": 4.013616019715655,
+      "learning_rate": 1.913579252038114e-05,
+      "loss": 3.2489,
+      "step": 829
+    },
+    {
+      "epoch": 0.15968447886104564,
+      "grad_norm": 3.538666643320513,
+      "learning_rate": 1.9133256403309627e-05,
+      "loss": 3.3619,
+      "step": 830
+    },
+    {
+      "epoch": 0.15987686979943244,
+      "grad_norm": 3.4941762587956244,
+      "learning_rate": 1.913071673898656e-05,
+      "loss": 3.2055,
+      "step": 831
+    },
+    {
+      "epoch": 0.16006926073781924,
+      "grad_norm": 3.856220116921987,
+      "learning_rate": 1.912817352839833e-05,
+      "loss": 3.3974,
+      "step": 832
+    },
+    {
+      "epoch": 0.16026165167620604,
+      "grad_norm": 4.112013707465101,
+      "learning_rate": 1.9125626772532683e-05,
+      "loss": 3.1503,
+      "step": 833
+    },
+    {
+      "epoch": 0.16045404261459284,
+      "grad_norm": 3.9033182440169014,
+      "learning_rate": 1.9123076472378753e-05,
+      "loss": 3.6394,
+      "step": 834
+    },
+    {
+      "epoch": 0.16064643355297967,
+      "grad_norm": 4.095892545073012,
+      "learning_rate": 1.9120522628927047e-05,
+      "loss": 3.5815,
+      "step": 835
+    },
+    {
+      "epoch": 0.16083882449136647,
+      "grad_norm": 3.8384699781496536,
+      "learning_rate": 1.9117965243169444e-05,
+      "loss": 3.1393,
+      "step": 836
+    },
+    {
+      "epoch": 0.16103121542975327,
+      "grad_norm": 3.6723719188561756,
+      "learning_rate": 1.9115404316099212e-05,
+      "loss": 3.1213,
+      "step": 837
+    },
+    {
+      "epoch": 0.16122360636814007,
+      "grad_norm": 4.168805740975727,
+      "learning_rate": 1.9112839848710978e-05,
+      "loss": 3.2848,
+      "step": 838
+    },
+    {
+      "epoch": 0.16141599730652686,
+      "grad_norm": 3.906660605712675,
+      "learning_rate": 1.9110271842000755e-05,
+      "loss": 3.3915,
+      "step": 839
+    },
+    {
+      "epoch": 0.16160838824491366,
+      "grad_norm": 3.7182502065286904,
+      "learning_rate": 1.9107700296965926e-05,
+      "loss": 3.435,
+      "step": 840
+    },
+    {
+      "epoch": 0.16180077918330046,
+      "grad_norm": 3.7288905374890082,
+      "learning_rate": 1.910512521460525e-05,
+      "loss": 3.4563,
+      "step": 841
+    },
+    {
+      "epoch": 0.16199317012168726,
+      "grad_norm": 3.99718689537971,
+      "learning_rate": 1.9102546595918857e-05,
+      "loss": 3.2347,
+      "step": 842
+    },
+    {
+      "epoch": 0.16218556106007406,
+      "grad_norm": 3.666838712484358,
+      "learning_rate": 1.9099964441908258e-05,
+      "loss": 3.4011,
+      "step": 843
+    },
+    {
+      "epoch": 0.16237795199846086,
+      "grad_norm": 3.8810789986068763,
+      "learning_rate": 1.9097378753576327e-05,
+      "loss": 3.3636,
+      "step": 844
+    },
+    {
+      "epoch": 0.16257034293684766,
+      "grad_norm": 3.910302845484145,
+      "learning_rate": 1.9094789531927315e-05,
+      "loss": 3.5101,
+      "step": 845
+    },
+    {
+      "epoch": 0.1627627338752345,
+      "grad_norm": 3.9129142429204045,
+      "learning_rate": 1.909219677796685e-05,
+      "loss": 3.3531,
+      "step": 846
+    },
+    {
+      "epoch": 0.1629551248136213,
+      "grad_norm": 3.64976006766142,
+      "learning_rate": 1.9089600492701926e-05,
+      "loss": 3.4376,
+      "step": 847
+    },
+    {
+      "epoch": 0.1631475157520081,
+      "grad_norm": 3.6261261284818285,
+      "learning_rate": 1.908700067714091e-05,
+      "loss": 3.4446,
+      "step": 848
+    },
+    {
+      "epoch": 0.1633399066903949,
+      "grad_norm": 3.673976257193007,
+      "learning_rate": 1.9084397332293537e-05,
+      "loss": 3.2565,
+      "step": 849
+    },
+    {
+      "epoch": 0.1635322976287817,
+      "grad_norm": 3.774178714100663,
+      "learning_rate": 1.9081790459170926e-05,
+      "loss": 3.5573,
+      "step": 850
+    },
+    {
+      "epoch": 0.16372468856716849,
+      "grad_norm": 3.506068665925256,
+      "learning_rate": 1.9079180058785547e-05,
+      "loss": 3.1248,
+      "step": 851
+    },
+    {
+      "epoch": 0.16391707950555529,
+      "grad_norm": 3.841924084348077,
+      "learning_rate": 1.9076566132151255e-05,
+      "loss": 3.2609,
+      "step": 852
+    },
+    {
+      "epoch": 0.16410947044394208,
+      "grad_norm": 3.623388473949473,
+      "learning_rate": 1.907394868028326e-05,
+      "loss": 3.1476,
+      "step": 853
+    },
+    {
+      "epoch": 0.16430186138232888,
+      "grad_norm": 3.5690975525184045,
+      "learning_rate": 1.9071327704198163e-05,
+      "loss": 3.6811,
+      "step": 854
+    },
+    {
+      "epoch": 0.16449425232071568,
+      "grad_norm": 4.289252175642952,
+      "learning_rate": 1.906870320491391e-05,
+      "loss": 3.3999,
+      "step": 855
+    },
+    {
+      "epoch": 0.16468664325910248,
+      "grad_norm": 3.8070972910670013,
+      "learning_rate": 1.9066075183449835e-05,
+      "loss": 3.2767,
+      "step": 856
+    },
+    {
+      "epoch": 0.1648790341974893,
+      "grad_norm": 3.799222792553937,
+      "learning_rate": 1.9063443640826624e-05,
+      "loss": 3.1813,
+      "step": 857
+    },
+    {
+      "epoch": 0.1650714251358761,
+      "grad_norm": 2.3502576200118943,
+      "learning_rate": 1.906080857806634e-05,
+      "loss": 2.773,
+      "step": 858
+    },
+    {
+      "epoch": 0.1652638160742629,
+      "grad_norm": 4.140891370901141,
+      "learning_rate": 1.905816999619242e-05,
+      "loss": 3.4434,
+      "step": 859
+    },
+    {
+      "epoch": 0.1654562070126497,
+      "grad_norm": 3.798358860542677,
+      "learning_rate": 1.9055527896229642e-05,
+      "loss": 3.2424,
+      "step": 860
+    },
+    {
+      "epoch": 0.1656485979510365,
+      "grad_norm": 3.5613489072099815,
+      "learning_rate": 1.905288227920418e-05,
+      "loss": 3.3535,
+      "step": 861
+    },
+    {
+      "epoch": 0.1658409888894233,
+      "grad_norm": 3.838230826325324,
+      "learning_rate": 1.9050233146143554e-05,
+      "loss": 3.4646,
+      "step": 862
+    },
+    {
+      "epoch": 0.1660333798278101,
+      "grad_norm": 4.068850871685197,
+      "learning_rate": 1.9047580498076663e-05,
+      "loss": 3.5008,
+      "step": 863
+    },
+    {
+      "epoch": 0.1662257707661969,
+      "grad_norm": 3.84645525221868,
+      "learning_rate": 1.904492433603376e-05,
+      "loss": 3.4962,
+      "step": 864
+    },
+    {
+      "epoch": 0.1664181617045837,
+      "grad_norm": 3.783604562047447,
+      "learning_rate": 1.904226466104647e-05,
+      "loss": 3.6662,
+      "step": 865
+    },
+    {
+      "epoch": 0.1666105526429705,
+      "grad_norm": 3.9756112160869996,
+      "learning_rate": 1.903960147414778e-05,
+      "loss": 3.3766,
+      "step": 866
+    },
+    {
+      "epoch": 0.1668029435813573,
+      "grad_norm": 3.599341899272511,
+      "learning_rate": 1.903693477637204e-05,
+      "loss": 3.5607,
+      "step": 867
+    },
+    {
+      "epoch": 0.16699533451974413,
+      "grad_norm": 3.3431213556068213,
+      "learning_rate": 1.9034264568754967e-05,
+      "loss": 3.1606,
+      "step": 868
+    },
+    {
+      "epoch": 0.16718772545813093,
+      "grad_norm": 3.765566667947436,
+      "learning_rate": 1.9031590852333637e-05,
+      "loss": 3.2887,
+      "step": 869
+    },
+    {
+      "epoch": 0.16738011639651773,
+      "grad_norm": 3.536379749429491,
+      "learning_rate": 1.9028913628146487e-05,
+      "loss": 3.0671,
+      "step": 870
+    },
+    {
+      "epoch": 0.16757250733490453,
+      "grad_norm": 4.201927975881617,
+      "learning_rate": 1.902623289723333e-05,
+      "loss": 3.3642,
+      "step": 871
+    },
+    {
+      "epoch": 0.16776489827329133,
+      "grad_norm": 3.880654909264599,
+      "learning_rate": 1.902354866063532e-05,
+      "loss": 3.5291,
+      "step": 872
+    },
+    {
+      "epoch": 0.16795728921167813,
+      "grad_norm": 3.9008023689225624,
+      "learning_rate": 1.9020860919394992e-05,
+      "loss": 3.1752,
+      "step": 873
+    },
+    {
+      "epoch": 0.16814968015006493,
+      "grad_norm": 3.719555062703872,
+      "learning_rate": 1.9018169674556228e-05,
+      "loss": 3.1591,
+      "step": 874
+    },
+    {
+      "epoch": 0.16834207108845173,
+      "grad_norm": 3.5819531730340715,
+      "learning_rate": 1.901547492716428e-05,
+      "loss": 3.2805,
+      "step": 875
+    },
+    {
+      "epoch": 0.16853446202683853,
+      "grad_norm": 3.824004538516887,
+      "learning_rate": 1.9012776678265756e-05,
+      "loss": 3.588,
+      "step": 876
+    },
+    {
+      "epoch": 0.16872685296522533,
+      "grad_norm": 3.9228173626308704,
+      "learning_rate": 1.9010074928908624e-05,
+      "loss": 3.1218,
+      "step": 877
+    },
+    {
+      "epoch": 0.16891924390361213,
+      "grad_norm": 3.7154959917229045,
+      "learning_rate": 1.900736968014221e-05,
+      "loss": 3.3514,
+      "step": 878
+    },
+    {
+      "epoch": 0.16911163484199895,
+      "grad_norm": 3.7498436374734934,
+      "learning_rate": 1.9004660933017208e-05,
+      "loss": 3.4181,
+      "step": 879
+    },
+    {
+      "epoch": 0.16930402578038575,
+      "grad_norm": 2.5958196162435043,
+      "learning_rate": 1.900194868858566e-05,
+      "loss": 2.7701,
+      "step": 880
+    },
+    {
+      "epoch": 0.16949641671877255,
+      "grad_norm": 3.453580693616908,
+      "learning_rate": 1.8999232947900968e-05,
+      "loss": 3.1022,
+      "step": 881
+    },
+    {
+      "epoch": 0.16968880765715935,
+      "grad_norm": 3.578487840097027,
+      "learning_rate": 1.89965137120179e-05,
+      "loss": 3.542,
+      "step": 882
+    },
+    {
+      "epoch": 0.16988119859554615,
+      "grad_norm": 3.9787187702113957,
+      "learning_rate": 1.899379098199257e-05,
+      "loss": 3.6026,
+      "step": 883
+    },
+    {
+      "epoch": 0.17007358953393295,
+      "grad_norm": 3.7702223106805812,
+      "learning_rate": 1.899106475888246e-05,
+      "loss": 3.223,
+      "step": 884
+    },
+    {
+      "epoch": 0.17026598047231975,
+      "grad_norm": 3.847430284205849,
+      "learning_rate": 1.89883350437464e-05,
+      "loss": 3.339,
+      "step": 885
+    },
+    {
+      "epoch": 0.17045837141070655,
+      "grad_norm": 4.153817019845932,
+      "learning_rate": 1.8985601837644586e-05,
+      "loss": 3.2387,
+      "step": 886
+    },
+    {
+      "epoch": 0.17065076234909335,
+      "grad_norm": 3.9444501249972825,
+      "learning_rate": 1.8982865141638557e-05,
+      "loss": 3.2684,
+      "step": 887
+    },
+    {
+      "epoch": 0.17084315328748015,
+      "grad_norm": 3.6560281697029926,
+      "learning_rate": 1.8980124956791216e-05,
+      "loss": 3.5155,
+      "step": 888
+    },
+    {
+      "epoch": 0.17103554422586695,
+      "grad_norm": 3.3342208284569184,
+      "learning_rate": 1.8977381284166818e-05,
+      "loss": 3.3734,
+      "step": 889
+    },
+    {
+      "epoch": 0.17122793516425378,
+      "grad_norm": 3.9212332712051317,
+      "learning_rate": 1.897463412483098e-05,
+      "loss": 3.3063,
+      "step": 890
+    },
+    {
+      "epoch": 0.17142032610264057,
+      "grad_norm": 2.3452605734044725,
+      "learning_rate": 1.897188347985066e-05,
+      "loss": 2.7017,
+      "step": 891
+    },
+    {
+      "epoch": 0.17161271704102737,
+      "grad_norm": 3.732475712406152,
+      "learning_rate": 1.896912935029418e-05,
+      "loss": 3.3527,
+      "step": 892
+    },
+    {
+      "epoch": 0.17180510797941417,
+      "grad_norm": 2.4669611300871046,
+      "learning_rate": 1.896637173723121e-05,
+      "loss": 2.9544,
+      "step": 893
+    },
+    {
+      "epoch": 0.17199749891780097,
+      "grad_norm": 3.7842712876259186,
+      "learning_rate": 1.8963610641732777e-05,
+      "loss": 3.4949,
+      "step": 894
+    },
+    {
+      "epoch": 0.17218988985618777,
+      "grad_norm": 2.3946358404297317,
+      "learning_rate": 1.8960846064871257e-05,
+      "loss": 2.7898,
+      "step": 895
+    },
+    {
+      "epoch": 0.17238228079457457,
+      "grad_norm": 3.532902291284836,
+      "learning_rate": 1.8958078007720387e-05,
+      "loss": 3.3312,
+      "step": 896
+    },
+    {
+      "epoch": 0.17257467173296137,
+      "grad_norm": 3.9389751845951224,
+      "learning_rate": 1.895530647135524e-05,
+      "loss": 3.5202,
+      "step": 897
+    },
+    {
+      "epoch": 0.17276706267134817,
+      "grad_norm": 3.496674410919265,
+      "learning_rate": 1.8952531456852248e-05,
+      "loss": 3.3487,
+      "step": 898
+    },
+    {
+      "epoch": 0.17295945360973497,
+      "grad_norm": 3.6861467418605987,
+      "learning_rate": 1.8949752965289197e-05,
+      "loss": 3.1999,
+      "step": 899
+    },
+    {
+      "epoch": 0.1731518445481218,
+      "grad_norm": 3.7815316448489966,
+      "learning_rate": 1.894697099774523e-05,
+      "loss": 3.3929,
+      "step": 900
+    },
+    {
+      "epoch": 0.1733442354865086,
+      "grad_norm": 3.571979804420861,
+      "learning_rate": 1.894418555530082e-05,
+      "loss": 3.28,
+      "step": 901
+    },
+    {
+      "epoch": 0.1735366264248954,
+      "grad_norm": 3.7907113457870865,
+      "learning_rate": 1.89413966390378e-05,
+      "loss": 3.0804,
+      "step": 902
+    },
+    {
+      "epoch": 0.1737290173632822,
+      "grad_norm": 2.3768966526237305,
+      "learning_rate": 1.8938604250039362e-05,
+      "loss": 2.6133,
+      "step": 903
+    },
+    {
+      "epoch": 0.173921408301669,
+      "grad_norm": 3.6872068296055045,
+      "learning_rate": 1.8935808389390032e-05,
+      "loss": 3.1062,
+      "step": 904
+    },
+    {
+      "epoch": 0.1741137992400558,
+      "grad_norm": 3.9753424843304392,
+      "learning_rate": 1.893300905817569e-05,
+      "loss": 3.4384,
+      "step": 905
+    },
+    {
+      "epoch": 0.1743061901784426,
+      "grad_norm": 3.773710834628998,
+      "learning_rate": 1.8930206257483566e-05,
+      "loss": 3.4324,
+      "step": 906
+    },
+    {
+      "epoch": 0.1744985811168294,
+      "grad_norm": 3.7474020750083987,
+      "learning_rate": 1.8927399988402233e-05,
+      "loss": 3.242,
+      "step": 907
+    },
+    {
+      "epoch": 0.1746909720552162,
+      "grad_norm": 3.9809633047334314,
+      "learning_rate": 1.8924590252021614e-05,
+      "loss": 3.255,
+      "step": 908
+    },
+    {
+      "epoch": 0.174883362993603,
+      "grad_norm": 3.657348044339167,
+      "learning_rate": 1.8921777049432985e-05,
+      "loss": 3.2938,
+      "step": 909
+    },
+    {
+      "epoch": 0.1750757539319898,
+      "grad_norm": 3.6905693292174493,
+      "learning_rate": 1.8918960381728947e-05,
+      "loss": 3.3751,
+      "step": 910
+    },
+    {
+      "epoch": 0.17526814487037662,
+      "grad_norm": 3.5028864847714356,
+      "learning_rate": 1.8916140250003475e-05,
+      "loss": 3.1017,
+      "step": 911
+    },
+    {
+      "epoch": 0.17546053580876342,
+      "grad_norm": 4.0445860911862015,
+      "learning_rate": 1.891331665535187e-05,
+      "loss": 2.8897,
+      "step": 912
+    },
+    {
+      "epoch": 0.17565292674715022,
+      "grad_norm": 3.6971987328235913,
+      "learning_rate": 1.8910489598870784e-05,
+      "loss": 3.2658,
+      "step": 913
+    },
+    {
+      "epoch": 0.17584531768553702,
+      "grad_norm": 3.8771824085481437,
+      "learning_rate": 1.8907659081658214e-05,
+      "loss": 3.1795,
+      "step": 914
+    },
+    {
+      "epoch": 0.17603770862392382,
+      "grad_norm": 2.642611627691317,
+      "learning_rate": 1.8904825104813497e-05,
+      "loss": 2.9471,
+      "step": 915
+    },
+    {
+      "epoch": 0.17623009956231062,
+      "grad_norm": 3.912849813816243,
+      "learning_rate": 1.8901987669437322e-05,
+      "loss": 3.537,
+      "step": 916
+    },
+    {
+      "epoch": 0.17642249050069742,
+      "grad_norm": 4.062132648021204,
+      "learning_rate": 1.889914677663171e-05,
+      "loss": 3.6837,
+      "step": 917
+    },
+    {
+      "epoch": 0.17661488143908421,
+      "grad_norm": 3.9442158882551266,
+      "learning_rate": 1.8896302427500042e-05,
+      "loss": 3.3822,
+      "step": 918
+    },
+    {
+      "epoch": 0.17680727237747101,
+      "grad_norm": 3.62869226663621,
+      "learning_rate": 1.8893454623147017e-05,
+      "loss": 3.2202,
+      "step": 919
+    },
+    {
+      "epoch": 0.1769996633158578,
+      "grad_norm": 3.7052020597615876,
+      "learning_rate": 1.88906033646787e-05,
+      "loss": 3.335,
+      "step": 920
+    },
+    {
+      "epoch": 0.1771920542542446,
+      "grad_norm": 3.9359648289731815,
+      "learning_rate": 1.8887748653202478e-05,
+      "loss": 3.2959,
+      "step": 921
+    },
+    {
+      "epoch": 0.17738444519263144,
+      "grad_norm": 3.548739657007135,
+      "learning_rate": 1.8884890489827097e-05,
+      "loss": 3.3676,
+      "step": 922
+    },
+    {
+      "epoch": 0.17757683613101824,
+      "grad_norm": 3.4631778233405437,
+      "learning_rate": 1.8882028875662627e-05,
+      "loss": 3.2249,
+      "step": 923
+    },
+    {
+      "epoch": 0.17776922706940504,
+      "grad_norm": 3.865883713021327,
+      "learning_rate": 1.8879163811820493e-05,
+      "loss": 3.2998,
+      "step": 924
+    },
+    {
+      "epoch": 0.17796161800779184,
+      "grad_norm": 3.847450748531889,
+      "learning_rate": 1.8876295299413445e-05,
+      "loss": 3.3638,
+      "step": 925
+    },
+    {
+      "epoch": 0.17815400894617864,
+      "grad_norm": 3.6505819107478144,
+      "learning_rate": 1.8873423339555584e-05,
+      "loss": 3.2984,
+      "step": 926
+    },
+    {
+      "epoch": 0.17834639988456544,
+      "grad_norm": 3.8396517264549166,
+      "learning_rate": 1.8870547933362352e-05,
+      "loss": 3.1302,
+      "step": 927
+    },
+    {
+      "epoch": 0.17853879082295224,
+      "grad_norm": 3.349902649366339,
+      "learning_rate": 1.886766908195051e-05,
+      "loss": 3.2886,
+      "step": 928
+    },
+    {
+      "epoch": 0.17873118176133904,
+      "grad_norm": 3.6134100325950085,
+      "learning_rate": 1.8864786786438187e-05,
+      "loss": 3.1922,
+      "step": 929
+    },
+    {
+      "epoch": 0.17892357269972584,
+      "grad_norm": 2.844814320510446,
+      "learning_rate": 1.886190104794482e-05,
+      "loss": 2.8476,
+      "step": 930
+    },
+    {
+      "epoch": 0.17911596363811264,
+      "grad_norm": 3.5228907223438855,
+      "learning_rate": 1.8859011867591203e-05,
+      "loss": 3.4008,
+      "step": 931
+    },
+    {
+      "epoch": 0.17930835457649943,
+      "grad_norm": 3.529435816775825,
+      "learning_rate": 1.885611924649946e-05,
+      "loss": 3.4716,
+      "step": 932
+    },
+    {
+      "epoch": 0.17950074551488626,
+      "grad_norm": 3.558045097412259,
+      "learning_rate": 1.885322318579305e-05,
+      "loss": 3.1968,
+      "step": 933
+    },
+    {
+      "epoch": 0.17969313645327306,
+      "grad_norm": 3.513730028963813,
+      "learning_rate": 1.8850323686596766e-05,
+      "loss": 3.2703,
+      "step": 934
+    },
+    {
+      "epoch": 0.17988552739165986,
+      "grad_norm": 3.799904218805346,
+      "learning_rate": 1.8847420750036748e-05,
+      "loss": 3.518,
+      "step": 935
+    },
+    {
+      "epoch": 0.18007791833004666,
+      "grad_norm": 3.767348666894908,
+      "learning_rate": 1.884451437724046e-05,
+      "loss": 3.3184,
+      "step": 936
+    },
+    {
+      "epoch": 0.18027030926843346,
+      "grad_norm": 3.60908412753015,
+      "learning_rate": 1.8841604569336702e-05,
+      "loss": 3.1177,
+      "step": 937
+    },
+    {
+      "epoch": 0.18046270020682026,
+      "grad_norm": 3.8173528323686194,
+      "learning_rate": 1.883869132745561e-05,
+      "loss": 3.4607,
+      "step": 938
+    },
+    {
+      "epoch": 0.18065509114520706,
+      "grad_norm": 3.7642346435877254,
+      "learning_rate": 1.883577465272866e-05,
+      "loss": 3.2594,
+      "step": 939
+    },
+    {
+      "epoch": 0.18084748208359386,
+      "grad_norm": 2.6465860365231393,
+      "learning_rate": 1.8832854546288642e-05,
+      "loss": 2.7692,
+      "step": 940
+    },
+    {
+      "epoch": 0.18103987302198066,
+      "grad_norm": 3.5121580982214184,
+      "learning_rate": 1.8829931009269707e-05,
+      "loss": 3.1475,
+      "step": 941
+    },
+    {
+      "epoch": 0.18123226396036746,
+      "grad_norm": 3.778872326126429,
+      "learning_rate": 1.882700404280731e-05,
+      "loss": 2.9584,
+      "step": 942
+    },
+    {
+      "epoch": 0.18142465489875426,
+      "grad_norm": 3.8169274407192852,
+      "learning_rate": 1.8824073648038258e-05,
+      "loss": 3.3047,
+      "step": 943
+    },
+    {
+      "epoch": 0.18161704583714108,
+      "grad_norm": 3.505117172638852,
+      "learning_rate": 1.882113982610068e-05,
+      "loss": 3.4158,
+      "step": 944
+    },
+    {
+      "epoch": 0.18180943677552788,
+      "grad_norm": 3.6194246203842115,
+      "learning_rate": 1.881820257813404e-05,
+      "loss": 3.2396,
+      "step": 945
+    },
+    {
+      "epoch": 0.18200182771391468,
+      "grad_norm": 3.6898058594166723,
+      "learning_rate": 1.8815261905279133e-05,
+      "loss": 3.3277,
+      "step": 946
+    },
+    {
+      "epoch": 0.18219421865230148,
+      "grad_norm": 3.638331870677263,
+      "learning_rate": 1.8812317808678075e-05,
+      "loss": 3.1913,
+      "step": 947
+    },
+    {
+      "epoch": 0.18238660959068828,
+      "grad_norm": 3.863960344537927,
+      "learning_rate": 1.8809370289474327e-05,
+      "loss": 3.5927,
+      "step": 948
+    },
+    {
+      "epoch": 0.18257900052907508,
+      "grad_norm": 4.2221297268380376,
+      "learning_rate": 1.8806419348812673e-05,
+      "loss": 3.3191,
+      "step": 949
+    },
+    {
+      "epoch": 0.18277139146746188,
+      "grad_norm": 3.3326723113136985,
+      "learning_rate": 1.8803464987839217e-05,
+      "loss": 3.1879,
+      "step": 950
+    },
+    {
+      "epoch": 0.18296378240584868,
+      "grad_norm": 3.5299573707038094,
+      "learning_rate": 1.88005072077014e-05,
+      "loss": 3.3861,
+      "step": 951
+    },
+    {
+      "epoch": 0.18315617334423548,
+      "grad_norm": 3.505957101332907,
+      "learning_rate": 1.8797546009547996e-05,
+      "loss": 3.4116,
+      "step": 952
+    },
+    {
+      "epoch": 0.18334856428262228,
+      "grad_norm": 3.8449954236454538,
+      "learning_rate": 1.879458139452909e-05,
+      "loss": 3.6255,
+      "step": 953
+    },
+    {
+      "epoch": 0.18354095522100908,
+      "grad_norm": 3.6563714006716994,
+      "learning_rate": 1.8791613363796118e-05,
+      "loss": 3.2292,
+      "step": 954
+    },
+    {
+      "epoch": 0.1837333461593959,
+      "grad_norm": 3.620276291068154,
+      "learning_rate": 1.8788641918501817e-05,
+      "loss": 3.0894,
+      "step": 955
+    },
+    {
+      "epoch": 0.1839257370977827,
+      "grad_norm": 3.6321740560674436,
+      "learning_rate": 1.8785667059800264e-05,
+      "loss": 3.0285,
+      "step": 956
+    },
+    {
+      "epoch": 0.1841181280361695,
+      "grad_norm": 2.4848185625586785,
+      "learning_rate": 1.8782688788846865e-05,
+      "loss": 2.8891,
+      "step": 957
+    },
+    {
+      "epoch": 0.1843105189745563,
+      "grad_norm": 3.743411235340086,
+      "learning_rate": 1.877970710679834e-05,
+      "loss": 3.3287,
+      "step": 958
+    },
+    {
+      "epoch": 0.1845029099129431,
+      "grad_norm": 3.628159858940091,
+      "learning_rate": 1.877672201481275e-05,
+      "loss": 3.2454,
+      "step": 959
+    },
+    {
+      "epoch": 0.1846953008513299,
+      "grad_norm": 3.8685446779182175,
+      "learning_rate": 1.877373351404946e-05,
+      "loss": 3.3356,
+      "step": 960
+    },
+    {
+      "epoch": 0.1848876917897167,
+      "grad_norm": 4.031047177943223,
+      "learning_rate": 1.8770741605669173e-05,
+      "loss": 3.2247,
+      "step": 961
+    },
+    {
+      "epoch": 0.1850800827281035,
+      "grad_norm": 3.7362243259987684,
+      "learning_rate": 1.876774629083391e-05,
+      "loss": 3.5658,
+      "step": 962
+    },
+    {
+      "epoch": 0.1852724736664903,
+      "grad_norm": 3.978190963874401,
+      "learning_rate": 1.8764747570707017e-05,
+      "loss": 3.5073,
+      "step": 963
+    },
+    {
+      "epoch": 0.1854648646048771,
+      "grad_norm": 3.7283654675237514,
+      "learning_rate": 1.8761745446453167e-05,
+      "loss": 3.4356,
+      "step": 964
+    },
+    {
+      "epoch": 0.1856572555432639,
+      "grad_norm": 3.4989704239515516,
+      "learning_rate": 1.875873991923835e-05,
+      "loss": 3.0339,
+      "step": 965
+    },
+    {
+      "epoch": 0.18584964648165073,
+      "grad_norm": 3.635561891608815,
+      "learning_rate": 1.875573099022987e-05,
+      "loss": 3.4635,
+      "step": 966
+    },
+    {
+      "epoch": 0.18604203742003753,
+      "grad_norm": 3.609765265745457,
+      "learning_rate": 1.8752718660596367e-05,
+      "loss": 3.314,
+      "step": 967
+    },
+    {
+      "epoch": 0.18623442835842433,
+      "grad_norm": 3.6879975493305093,
+      "learning_rate": 1.8749702931507797e-05,
+      "loss": 3.4329,
+      "step": 968
+    },
+    {
+      "epoch": 0.18642681929681112,
+      "grad_norm": 3.761944143268723,
+      "learning_rate": 1.874668380413543e-05,
+      "loss": 3.4629,
+      "step": 969
+    },
+    {
+      "epoch": 0.18661921023519792,
+      "grad_norm": 3.759177859226735,
+      "learning_rate": 1.8743661279651856e-05,
+      "loss": 3.4633,
+      "step": 970
+    },
+    {
+      "epoch": 0.18681160117358472,
+      "grad_norm": 3.80812324779773,
+      "learning_rate": 1.8740635359231e-05,
+      "loss": 3.5043,
+      "step": 971
+    },
+    {
+      "epoch": 0.18700399211197152,
+      "grad_norm": 3.8758164654383322,
+      "learning_rate": 1.8737606044048086e-05,
+      "loss": 3.2462,
+      "step": 972
+    },
+    {
+      "epoch": 0.18719638305035832,
+      "grad_norm": 3.7653475291871787,
+      "learning_rate": 1.873457333527967e-05,
+      "loss": 3.3824,
+      "step": 973
+    },
+    {
+      "epoch": 0.18738877398874512,
+      "grad_norm": 3.3251307297796964,
+      "learning_rate": 1.873153723410362e-05,
+      "loss": 3.2417,
+      "step": 974
+    },
+    {
+      "epoch": 0.18758116492713192,
+      "grad_norm": 3.7412810855102414,
+      "learning_rate": 1.8728497741699115e-05,
+      "loss": 3.4674,
+      "step": 975
+    },
+    {
+      "epoch": 0.18777355586551872,
+      "grad_norm": 3.764171256202672,
+      "learning_rate": 1.872545485924667e-05,
+      "loss": 3.3952,
+      "step": 976
+    },
+    {
+      "epoch": 0.18796594680390555,
+      "grad_norm": 3.9949460120113294,
+      "learning_rate": 1.8722408587928104e-05,
+      "loss": 3.0292,
+      "step": 977
+    },
+    {
+      "epoch": 0.18815833774229235,
+      "grad_norm": 3.842188919878382,
+      "learning_rate": 1.8719358928926546e-05,
+      "loss": 3.2077,
+      "step": 978
+    },
+    {
+      "epoch": 0.18835072868067915,
+      "grad_norm": 3.882080038070223,
+      "learning_rate": 1.8716305883426456e-05,
+      "loss": 3.4547,
+      "step": 979
+    },
+    {
+      "epoch": 0.18854311961906595,
+      "grad_norm": 3.7241452281899723,
+      "learning_rate": 1.87132494526136e-05,
+      "loss": 3.3897,
+      "step": 980
+    },
+    {
+      "epoch": 0.18873551055745275,
+      "grad_norm": 3.5641742069369653,
+      "learning_rate": 1.8710189637675055e-05,
+      "loss": 3.3291,
+      "step": 981
+    },
+    {
+      "epoch": 0.18892790149583955,
+      "grad_norm": 3.519450820725552,
+      "learning_rate": 1.8707126439799225e-05,
+      "loss": 3.1263,
+      "step": 982
+    },
+    {
+      "epoch": 0.18912029243422634,
+      "grad_norm": 3.4998682331873723,
+      "learning_rate": 1.870405986017582e-05,
+      "loss": 3.2317,
+      "step": 983
+    },
+    {
+      "epoch": 0.18931268337261314,
+      "grad_norm": 2.7772123932364132,
+      "learning_rate": 1.8700989899995857e-05,
+      "loss": 2.8833,
+      "step": 984
+    },
+    {
+      "epoch": 0.18950507431099994,
+      "grad_norm": 3.9459943155189174,
+      "learning_rate": 1.8697916560451682e-05,
+      "loss": 3.1636,
+      "step": 985
+    },
+    {
+      "epoch": 0.18969746524938674,
+      "grad_norm": 3.401652826226265,
+      "learning_rate": 1.869483984273694e-05,
+      "loss": 3.2266,
+      "step": 986
+    },
+    {
+      "epoch": 0.18988985618777354,
+      "grad_norm": 3.7440360061728915,
+      "learning_rate": 1.8691759748046594e-05,
+      "loss": 3.2045,
+      "step": 987
+    },
+    {
+      "epoch": 0.19008224712616037,
+      "grad_norm": 3.380494330842233,
+      "learning_rate": 1.8688676277576916e-05,
+      "loss": 2.9217,
+      "step": 988
+    },
+    {
+      "epoch": 0.19027463806454717,
+      "grad_norm": 3.599271798950916,
+      "learning_rate": 1.868558943252549e-05,
+      "loss": 3.2403,
+      "step": 989
+    },
+    {
+      "epoch": 0.19046702900293397,
+      "grad_norm": 2.175679996183849,
+      "learning_rate": 1.868249921409122e-05,
+      "loss": 2.7465,
+      "step": 990
+    },
+    {
+      "epoch": 0.19065941994132077,
+      "grad_norm": 3.560969033835681,
+      "learning_rate": 1.8679405623474294e-05,
+      "loss": 3.1901,
+      "step": 991
+    },
+    {
+      "epoch": 0.19085181087970757,
+      "grad_norm": 3.852305331842743,
+      "learning_rate": 1.8676308661876242e-05,
+      "loss": 3.4346,
+      "step": 992
+    },
+    {
+      "epoch": 0.19104420181809437,
+      "grad_norm": 3.6331790361721263,
+      "learning_rate": 1.8673208330499884e-05,
+      "loss": 3.402,
+      "step": 993
+    },
+    {
+      "epoch": 0.19123659275648117,
+      "grad_norm": 3.778265629095022,
+      "learning_rate": 1.8670104630549348e-05,
+      "loss": 3.3602,
+      "step": 994
+    },
+    {
+      "epoch": 0.19142898369486797,
+      "grad_norm": 3.634835361700186,
+      "learning_rate": 1.866699756323008e-05,
+      "loss": 3.6459,
+      "step": 995
+    },
+    {
+      "epoch": 0.19162137463325477,
+      "grad_norm": 3.710686674370289,
+      "learning_rate": 1.866388712974883e-05,
+      "loss": 3.1749,
+      "step": 996
+    },
+    {
+      "epoch": 0.19181376557164156,
+      "grad_norm": 3.4897242209275507,
+      "learning_rate": 1.866077333131365e-05,
+      "loss": 3.2747,
+      "step": 997
+    },
+    {
+      "epoch": 0.19200615651002836,
+      "grad_norm": 3.7315824045922668,
+      "learning_rate": 1.8657656169133908e-05,
+      "loss": 3.262,
+      "step": 998
+    },
+    {
+      "epoch": 0.1921985474484152,
+      "grad_norm": 3.833137501260972,
+      "learning_rate": 1.8654535644420276e-05,
+      "loss": 3.3925,
+      "step": 999
+    },
+    {
+      "epoch": 0.192390938386802,
+      "grad_norm": 3.7813347978439666,
+      "learning_rate": 1.8651411758384718e-05,
+      "loss": 3.3222,
+      "step": 1000
+    },
+    {
+      "epoch": 0.1925833293251888,
+      "grad_norm": 3.6530438419347924,
+      "learning_rate": 1.8648284512240527e-05,
+      "loss": 3.2956,
+      "step": 1001
+    },
+    {
+      "epoch": 0.1927757202635756,
+      "grad_norm": 3.95333384576967,
+      "learning_rate": 1.8645153907202285e-05,
+      "loss": 3.3077,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1929681112019624,
+      "grad_norm": 3.2925303848996803,
+      "learning_rate": 1.8642019944485884e-05,
+      "loss": 3.2576,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1931605021403492,
+      "grad_norm": 3.529726171320256,
+      "learning_rate": 1.863888262530852e-05,
+      "loss": 3.3958,
+      "step": 1004
+    },
+    {
+      "epoch": 0.193352893078736,
+      "grad_norm": 3.479236936197514,
+      "learning_rate": 1.863574195088869e-05,
+      "loss": 3.2905,
+      "step": 1005
+    },
+    {
+      "epoch": 0.1935452840171228,
+      "grad_norm": 3.6639571361989858,
+      "learning_rate": 1.8632597922446195e-05,
+      "loss": 3.1664,
+      "step": 1006
+    },
+    {
+      "epoch": 0.1937376749555096,
+      "grad_norm": 3.438593661411922,
+      "learning_rate": 1.8629450541202142e-05,
+      "loss": 3.1502,
+      "step": 1007
+    },
+    {
+      "epoch": 0.1939300658938964,
+      "grad_norm": 3.8561897467710646,
+      "learning_rate": 1.8626299808378933e-05,
+      "loss": 3.2916,
+      "step": 1008
+    },
+    {
+      "epoch": 0.19412245683228319,
+      "grad_norm": 3.766856814131168,
+      "learning_rate": 1.862314572520028e-05,
+      "loss": 3.3581,
+      "step": 1009
+    },
+    {
+      "epoch": 0.19431484777067,
+      "grad_norm": 3.3984161282294076,
+      "learning_rate": 1.861998829289119e-05,
+      "loss": 3.3115,
+      "step": 1010
+    },
+    {
+      "epoch": 0.1945072387090568,
+      "grad_norm": 3.913643231132839,
+      "learning_rate": 1.861682751267798e-05,
+      "loss": 3.2178,
+      "step": 1011
+    },
+    {
+      "epoch": 0.1946996296474436,
+      "grad_norm": 3.5180413536338655,
+      "learning_rate": 1.861366338578825e-05,
+      "loss": 3.1398,
+      "step": 1012
+    },
+    {
+      "epoch": 0.1948920205858304,
+      "grad_norm": 3.4963392500031762,
+      "learning_rate": 1.8610495913450922e-05,
+      "loss": 3.3579,
+      "step": 1013
+    },
+    {
+      "epoch": 0.1950844115242172,
+      "grad_norm": 3.632881649811181,
+      "learning_rate": 1.8607325096896197e-05,
+      "loss": 3.1529,
+      "step": 1014
+    },
+    {
+      "epoch": 0.195276802462604,
+      "grad_norm": 3.4973368613547047,
+      "learning_rate": 1.8604150937355588e-05,
+      "loss": 3.3052,
+      "step": 1015
+    },
+    {
+      "epoch": 0.1954691934009908,
+      "grad_norm": 3.854191564374744,
+      "learning_rate": 1.86009734360619e-05,
+      "loss": 3.4566,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1956615843393776,
+      "grad_norm": 3.783141219427456,
+      "learning_rate": 1.8597792594249237e-05,
+      "loss": 3.2546,
+      "step": 1017
+    },
+    {
+      "epoch": 0.1958539752777644,
+      "grad_norm": 2.491988600627279,
+      "learning_rate": 1.8594608413153e-05,
+      "loss": 2.6835,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1960463662161512,
+      "grad_norm": 2.4806332201031402,
+      "learning_rate": 1.8591420894009897e-05,
+      "loss": 2.7807,
+      "step": 1019
+    },
+    {
+      "epoch": 0.19623875715453803,
+      "grad_norm": 3.5565971709501514,
+      "learning_rate": 1.8588230038057913e-05,
+      "loss": 3.3128,
+      "step": 1020
+    },
+    {
+      "epoch": 0.19643114809292483,
+      "grad_norm": 3.7377507252466042,
+      "learning_rate": 1.8585035846536347e-05,
+      "loss": 3.3658,
+      "step": 1021
+    },
+    {
+      "epoch": 0.19662353903131163,
+      "grad_norm": 3.9043061280516658,
+      "learning_rate": 1.8581838320685782e-05,
+      "loss": 3.3929,
+      "step": 1022
+    },
+    {
+      "epoch": 0.19681592996969843,
+      "grad_norm": 3.926980740174967,
+      "learning_rate": 1.8578637461748105e-05,
+      "loss": 3.6829,
+      "step": 1023
+    },
+    {
+      "epoch": 0.19700832090808523,
+      "grad_norm": 3.8382220065083006,
+      "learning_rate": 1.857543327096649e-05,
+      "loss": 3.4138,
+      "step": 1024
+    },
+    {
+      "epoch": 0.19720071184647203,
+      "grad_norm": 3.621244876315297,
+      "learning_rate": 1.85722257495854e-05,
+      "loss": 3.3486,
+      "step": 1025
+    },
+    {
+      "epoch": 0.19739310278485883,
+      "grad_norm": 3.771629288082905,
+      "learning_rate": 1.856901489885061e-05,
+      "loss": 3.6227,
+      "step": 1026
+    },
+    {
+      "epoch": 0.19758549372324563,
+      "grad_norm": 3.888987021732522,
+      "learning_rate": 1.856580072000918e-05,
+      "loss": 3.3009,
+      "step": 1027
+    },
+    {
+      "epoch": 0.19777788466163243,
+      "grad_norm": 3.6997464416678065,
+      "learning_rate": 1.8562583214309447e-05,
+      "loss": 3.3893,
+      "step": 1028
+    },
+    {
+      "epoch": 0.19797027560001923,
+      "grad_norm": 3.9957460957305706,
+      "learning_rate": 1.855936238300106e-05,
+      "loss": 3.4376,
+      "step": 1029
+    },
+    {
+      "epoch": 0.19816266653840603,
+      "grad_norm": 3.75521674552145,
+      "learning_rate": 1.8556138227334957e-05,
+      "loss": 3.4112,
+      "step": 1030
+    },
+    {
+      "epoch": 0.19835505747679286,
+      "grad_norm": 3.468798943544283,
+      "learning_rate": 1.855291074856336e-05,
+      "loss": 3.2921,
+      "step": 1031
+    },
+    {
+      "epoch": 0.19854744841517966,
+      "grad_norm": 3.592208582317363,
+      "learning_rate": 1.8549679947939778e-05,
+      "loss": 3.2812,
+      "step": 1032
+    },
+    {
+      "epoch": 0.19873983935356646,
+      "grad_norm": 3.720284927903332,
+      "learning_rate": 1.8546445826719023e-05,
+      "loss": 3.2748,
+      "step": 1033
+    },
+    {
+      "epoch": 0.19893223029195325,
+      "grad_norm": 4.035069420300631,
+      "learning_rate": 1.8543208386157195e-05,
+      "loss": 3.4026,
+      "step": 1034
+    },
+    {
+      "epoch": 0.19912462123034005,
+      "grad_norm": 3.448873798443782,
+      "learning_rate": 1.853996762751167e-05,
+      "loss": 3.28,
+      "step": 1035
+    },
+    {
+      "epoch": 0.19931701216872685,
+      "grad_norm": 3.421084621421924,
+      "learning_rate": 1.8536723552041124e-05,
+      "loss": 3.048,
+      "step": 1036
+    },
+    {
+      "epoch": 0.19950940310711365,
+      "grad_norm": 2.653199046043925,
+      "learning_rate": 1.853347616100552e-05,
+      "loss": 2.7984,
+      "step": 1037
+    },
+    {
+      "epoch": 0.19970179404550045,
+      "grad_norm": 4.0537033554070705,
+      "learning_rate": 1.8530225455666103e-05,
+      "loss": 3.6408,
+      "step": 1038
+    },
+    {
+      "epoch": 0.19989418498388725,
+      "grad_norm": 3.8647521505146334,
+      "learning_rate": 1.8526971437285416e-05,
+      "loss": 3.4454,
+      "step": 1039
+    },
+    {
+      "epoch": 0.20008657592227405,
+      "grad_norm": 3.812144348486101,
+      "learning_rate": 1.8523714107127278e-05,
+      "loss": 3.2739,
+      "step": 1040
+    },
+    {
+      "epoch": 0.20027896686066085,
+      "grad_norm": 3.66800637899219,
+      "learning_rate": 1.8520453466456797e-05,
+      "loss": 3.1953,
+      "step": 1041
+    },
+    {
+      "epoch": 0.20047135779904768,
+      "grad_norm": 2.5084857171016424,
+      "learning_rate": 1.8517189516540376e-05,
+      "loss": 2.8469,
+      "step": 1042
+    },
+    {
+      "epoch": 0.20066374873743448,
+      "grad_norm": 3.723926752287868,
+      "learning_rate": 1.8513922258645687e-05,
+      "loss": 3.1398,
+      "step": 1043
+    },
+    {
+      "epoch": 0.20085613967582128,
+      "grad_norm": 3.7346223814980934,
+      "learning_rate": 1.8510651694041702e-05,
+      "loss": 3.1229,
+      "step": 1044
+    },
+    {
+      "epoch": 0.20104853061420808,
+      "grad_norm": 3.7923722464703205,
+      "learning_rate": 1.8507377823998664e-05,
+      "loss": 3.2298,
+      "step": 1045
+    },
+    {
+      "epoch": 0.20124092155259488,
+      "grad_norm": 3.6024841213651215,
+      "learning_rate": 1.850410064978811e-05,
+      "loss": 3.1737,
+      "step": 1046
+    },
+    {
+      "epoch": 0.20143331249098168,
+      "grad_norm": 3.514986827646717,
+      "learning_rate": 1.8500820172682858e-05,
+      "loss": 3.4881,
+      "step": 1047
+    },
+    {
+      "epoch": 0.20162570342936847,
+      "grad_norm": 3.763168004471287,
+      "learning_rate": 1.8497536393957005e-05,
+      "loss": 3.2618,
+      "step": 1048
+    },
+    {
+      "epoch": 0.20181809436775527,
+      "grad_norm": 3.3565451427104285,
+      "learning_rate": 1.8494249314885932e-05,
+      "loss": 3.2789,
+      "step": 1049
+    },
+    {
+      "epoch": 0.20201048530614207,
+      "grad_norm": 3.722822424702454,
+      "learning_rate": 1.8490958936746304e-05,
+      "loss": 3.3118,
+      "step": 1050
+    },
+    {
+      "epoch": 0.20220287624452887,
+      "grad_norm": 3.4703313648094722,
+      "learning_rate": 1.848766526081607e-05,
+      "loss": 3.2949,
+      "step": 1051
+    },
+    {
+      "epoch": 0.20239526718291567,
+      "grad_norm": 3.3860476483086006,
+      "learning_rate": 1.8484368288374452e-05,
+      "loss": 3.0235,
+      "step": 1052
+    },
+    {
+      "epoch": 0.2025876581213025,
+      "grad_norm": 3.9072589314723825,
+      "learning_rate": 1.8481068020701954e-05,
+      "loss": 3.1916,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2027800490596893,
+      "grad_norm": 3.7808641845204702,
+      "learning_rate": 1.8477764459080364e-05,
+      "loss": 3.5465,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2029724399980761,
+      "grad_norm": 3.6929810932189104,
+      "learning_rate": 1.8474457604792746e-05,
+      "loss": 3.2289,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2031648309364629,
+      "grad_norm": 3.60334996279392,
+      "learning_rate": 1.8471147459123447e-05,
+      "loss": 3.1884,
+      "step": 1056
+    },
+    {
+      "epoch": 0.2033572218748497,
+      "grad_norm": 3.636637031792294,
+      "learning_rate": 1.8467834023358088e-05,
+      "loss": 3.5079,
+      "step": 1057
+    },
+    {
+      "epoch": 0.2035496128132365,
+      "grad_norm": 3.7795064913837186,
+      "learning_rate": 1.846451729878357e-05,
+      "loss": 3.3518,
+      "step": 1058
+    },
+    {
+      "epoch": 0.2037420037516233,
+      "grad_norm": 3.6547750028372294,
+      "learning_rate": 1.846119728668807e-05,
+      "loss": 3.2087,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2039343946900101,
+      "grad_norm": 2.65744198427758,
+      "learning_rate": 1.845787398836104e-05,
+      "loss": 2.9168,
+      "step": 1060
+    },
+    {
+      "epoch": 0.2041267856283969,
+      "grad_norm": 3.4712309425550743,
+      "learning_rate": 1.8454547405093212e-05,
+      "loss": 3.2617,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2043191765667837,
+      "grad_norm": 3.6463273006402686,
+      "learning_rate": 1.8451217538176597e-05,
+      "loss": 3.2772,
+      "step": 1062
+    },
+    {
+      "epoch": 0.2045115675051705,
+      "grad_norm": 3.387304632790974,
+      "learning_rate": 1.844788438890447e-05,
+      "loss": 3.2592,
+      "step": 1063
+    },
+    {
+      "epoch": 0.20470395844355732,
+      "grad_norm": 3.558043481085913,
+      "learning_rate": 1.8444547958571396e-05,
+      "loss": 3.307,
+      "step": 1064
+    },
+    {
+      "epoch": 0.20489634938194412,
+      "grad_norm": 3.3498202546067857,
+      "learning_rate": 1.84412082484732e-05,
+      "loss": 3.4513,
+      "step": 1065
+    },
+    {
+      "epoch": 0.20508874032033092,
+      "grad_norm": 3.502921823770039,
+      "learning_rate": 1.8437865259906987e-05,
+      "loss": 3.1403,
+      "step": 1066
+    },
+    {
+      "epoch": 0.20528113125871772,
+      "grad_norm": 3.827196841158253,
+      "learning_rate": 1.8434518994171136e-05,
+      "loss": 3.0626,
+      "step": 1067
+    },
+    {
+      "epoch": 0.20547352219710452,
+      "grad_norm": 3.999064858765071,
+      "learning_rate": 1.84311694525653e-05,
+      "loss": 3.3179,
+      "step": 1068
+    },
+    {
+      "epoch": 0.20566591313549132,
+      "grad_norm": 3.8685414333320947,
+      "learning_rate": 1.84278166363904e-05,
+      "loss": 3.6509,
+      "step": 1069
+    },
+    {
+      "epoch": 0.20585830407387812,
+      "grad_norm": 3.8089090591401336,
+      "learning_rate": 1.8424460546948632e-05,
+      "loss": 3.1937,
+      "step": 1070
+    },
+    {
+      "epoch": 0.20605069501226492,
+      "grad_norm": 3.816902742874493,
+      "learning_rate": 1.8421101185543463e-05,
+      "loss": 3.2598,
+      "step": 1071
+    },
+    {
+      "epoch": 0.20624308595065172,
+      "grad_norm": 3.6490394562356205,
+      "learning_rate": 1.841773855347963e-05,
+      "loss": 3.447,
+      "step": 1072
+    },
+    {
+      "epoch": 0.20643547688903852,
+      "grad_norm": 3.4784780113908185,
+      "learning_rate": 1.841437265206314e-05,
+      "loss": 3.2568,
+      "step": 1073
+    },
+    {
+      "epoch": 0.20662786782742532,
+      "grad_norm": 3.638943718426881,
+      "learning_rate": 1.841100348260127e-05,
+      "loss": 3.1666,
+      "step": 1074
+    },
+    {
+      "epoch": 0.20682025876581214,
+      "grad_norm": 3.6329968025529995,
+      "learning_rate": 1.840763104640257e-05,
+      "loss": 3.1322,
+      "step": 1075
+    },
+    {
+      "epoch": 0.20701264970419894,
+      "grad_norm": 3.754802240192697,
+      "learning_rate": 1.8404255344776853e-05,
+      "loss": 3.4042,
+      "step": 1076
+    },
+    {
+      "epoch": 0.20720504064258574,
+      "grad_norm": 3.3007335395577284,
+      "learning_rate": 1.84008763790352e-05,
+      "loss": 3.1617,
+      "step": 1077
+    },
+    {
+      "epoch": 0.20739743158097254,
+      "grad_norm": 3.5686486880493984,
+      "learning_rate": 1.8397494150489963e-05,
+      "loss": 3.2491,
+      "step": 1078
+    },
+    {
+      "epoch": 0.20758982251935934,
+      "grad_norm": 3.438791056294158,
+      "learning_rate": 1.8394108660454766e-05,
+      "loss": 3.3265,
+      "step": 1079
+    },
+    {
+      "epoch": 0.20778221345774614,
+      "grad_norm": 3.3886625396142853,
+      "learning_rate": 1.8390719910244487e-05,
+      "loss": 3.0809,
+      "step": 1080
+    },
+    {
+      "epoch": 0.20797460439613294,
+      "grad_norm": 3.6767638338628124,
+      "learning_rate": 1.8387327901175286e-05,
+      "loss": 3.3294,
+      "step": 1081
+    },
+    {
+      "epoch": 0.20816699533451974,
+      "grad_norm": 3.947587297956204,
+      "learning_rate": 1.838393263456457e-05,
+      "loss": 3.2335,
+      "step": 1082
+    },
+    {
+      "epoch": 0.20835938627290654,
+      "grad_norm": 3.7170489513554177,
+      "learning_rate": 1.838053411173103e-05,
+      "loss": 3.324,
+      "step": 1083
+    },
+    {
+      "epoch": 0.20855177721129334,
+      "grad_norm": 3.697648813169382,
+      "learning_rate": 1.8377132333994606e-05,
+      "loss": 3.4174,
+      "step": 1084
+    },
+    {
+      "epoch": 0.20874416814968014,
+      "grad_norm": 2.454078723338447,
+      "learning_rate": 1.837372730267652e-05,
+      "loss": 2.7743,
+      "step": 1085
+    },
+    {
+      "epoch": 0.20893655908806696,
+      "grad_norm": 3.983524845627747,
+      "learning_rate": 1.8370319019099236e-05,
+      "loss": 3.2964,
+      "step": 1086
+    },
+    {
+      "epoch": 0.20912895002645376,
+      "grad_norm": 4.017651741005177,
+      "learning_rate": 1.8366907484586497e-05,
+      "loss": 3.4723,
+      "step": 1087
+    },
+    {
+      "epoch": 0.20932134096484056,
+      "grad_norm": 3.5416052346880287,
+      "learning_rate": 1.83634927004633e-05,
+      "loss": 3.265,
+      "step": 1088
+    },
+    {
+      "epoch": 0.20951373190322736,
+      "grad_norm": 4.59798760453932,
+      "learning_rate": 1.8360074668055915e-05,
+      "loss": 2.9131,
+      "step": 1089
+    },
+    {
+      "epoch": 0.20970612284161416,
+      "grad_norm": 4.191257159742864,
+      "learning_rate": 1.8356653388691857e-05,
+      "loss": 3.4458,
+      "step": 1090
+    },
+    {
+      "epoch": 0.20989851378000096,
+      "grad_norm": 3.4167232974459103,
+      "learning_rate": 1.8353228863699922e-05,
+      "loss": 3.156,
+      "step": 1091
+    },
+    {
+      "epoch": 0.21009090471838776,
+      "grad_norm": 3.750714319609836,
+      "learning_rate": 1.8349801094410148e-05,
+      "loss": 3.2541,
+      "step": 1092
+    },
+    {
+      "epoch": 0.21028329565677456,
+      "grad_norm": 3.7697417791462815,
+      "learning_rate": 1.8346370082153843e-05,
+      "loss": 2.9436,
+      "step": 1093
+    },
+    {
+      "epoch": 0.21047568659516136,
+      "grad_norm": 3.787164124551854,
+      "learning_rate": 1.8342935828263574e-05,
+      "loss": 3.0144,
+      "step": 1094
+    },
+    {
+      "epoch": 0.21066807753354816,
+      "grad_norm": 3.5394330092441204,
+      "learning_rate": 1.8339498334073166e-05,
+      "loss": 3.2821,
+      "step": 1095
+    },
+    {
+      "epoch": 0.21086046847193496,
+      "grad_norm": 3.379815184057389,
+      "learning_rate": 1.83360576009177e-05,
+      "loss": 3.1312,
+      "step": 1096
+    },
+    {
+      "epoch": 0.21105285941032179,
+      "grad_norm": 3.7457918602775497,
+      "learning_rate": 1.833261363013352e-05,
+      "loss": 3.3936,
+      "step": 1097
+    },
+    {
+      "epoch": 0.21124525034870859,
+      "grad_norm": 3.789524347215857,
+      "learning_rate": 1.832916642305822e-05,
+      "loss": 3.3639,
+      "step": 1098
+    },
+    {
+      "epoch": 0.21143764128709538,
+      "grad_norm": 3.7630621778739575,
+      "learning_rate": 1.832571598103066e-05,
+      "loss": 3.3222,
+      "step": 1099
+    },
+    {
+      "epoch": 0.21163003222548218,
+      "grad_norm": 3.923851111841389,
+      "learning_rate": 1.8322262305390948e-05,
+      "loss": 3.3961,
+      "step": 1100
+    },
+    {
+      "epoch": 0.21182242316386898,
+      "grad_norm": 3.7266743969421325,
+      "learning_rate": 1.8318805397480455e-05,
+      "loss": 3.4283,
+      "step": 1101
+    },
+    {
+      "epoch": 0.21201481410225578,
+      "grad_norm": 3.5424213624171403,
+      "learning_rate": 1.8315345258641802e-05,
+      "loss": 3.3251,
+      "step": 1102
+    },
+    {
+      "epoch": 0.21220720504064258,
+      "grad_norm": 3.610828990037535,
+      "learning_rate": 1.8311881890218873e-05,
+      "loss": 3.4798,
+      "step": 1103
+    },
+    {
+      "epoch": 0.21239959597902938,
+      "grad_norm": 3.5430347629535395,
+      "learning_rate": 1.830841529355679e-05,
+      "loss": 3.3374,
+      "step": 1104
+    },
+    {
+      "epoch": 0.21259198691741618,
+      "grad_norm": 3.647889781041136,
+      "learning_rate": 1.8304945470001948e-05,
+      "loss": 3.4843,
+      "step": 1105
+    },
+    {
+      "epoch": 0.21278437785580298,
+      "grad_norm": 3.7798713419581955,
+      "learning_rate": 1.8301472420901985e-05,
+      "loss": 3.2531,
+      "step": 1106
+    },
+    {
+      "epoch": 0.21297676879418978,
+      "grad_norm": 2.5025915952213036,
+      "learning_rate": 1.8297996147605787e-05,
+      "loss": 2.7806,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2131691597325766,
+      "grad_norm": 3.6459502159482415,
+      "learning_rate": 1.829451665146351e-05,
+      "loss": 3.234,
+      "step": 1108
+    },
+    {
+      "epoch": 0.2133615506709634,
+      "grad_norm": 3.720132676168959,
+      "learning_rate": 1.8291033933826535e-05,
+      "loss": 3.2532,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2135539416093502,
+      "grad_norm": 3.603698453763478,
+      "learning_rate": 1.8287547996047523e-05,
+      "loss": 3.2349,
+      "step": 1110
+    },
+    {
+      "epoch": 0.213746332547737,
+      "grad_norm": 3.7417484422720957,
+      "learning_rate": 1.8284058839480363e-05,
+      "loss": 3.2723,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2139387234861238,
+      "grad_norm": 3.6078706557178024,
+      "learning_rate": 1.8280566465480206e-05,
+      "loss": 3.115,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2141311144245106,
+      "grad_norm": 3.6504412945493185,
+      "learning_rate": 1.8277070875403455e-05,
+      "loss": 3.0297,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2143235053628974,
+      "grad_norm": 3.755649943794121,
+      "learning_rate": 1.8273572070607756e-05,
+      "loss": 3.3713,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2145158963012842,
+      "grad_norm": 4.176709392465833,
+      "learning_rate": 1.8270070052451995e-05,
+      "loss": 3.2771,
+      "step": 1115
+    },
+    {
+      "epoch": 0.214708287239671,
+      "grad_norm": 3.932113841863053,
+      "learning_rate": 1.8266564822296323e-05,
+      "loss": 3.3402,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2149006781780578,
+      "grad_norm": 3.459662515093711,
+      "learning_rate": 1.826305638150213e-05,
+      "loss": 3.1933,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2150930691164446,
+      "grad_norm": 3.7430054638061283,
+      "learning_rate": 1.825954473143205e-05,
+      "loss": 3.2808,
+      "step": 1118
+    },
+    {
+      "epoch": 0.21528546005483143,
+      "grad_norm": 2.4855206320728778,
+      "learning_rate": 1.8256029873449976e-05,
+      "loss": 2.8493,
+      "step": 1119
+    },
+    {
+      "epoch": 0.21547785099321823,
+      "grad_norm": 3.6465865151746897,
+      "learning_rate": 1.825251180892103e-05,
+      "loss": 3.3145,
+      "step": 1120
+    },
+    {
+      "epoch": 0.21567024193160503,
+      "grad_norm": 3.829666929887576,
+      "learning_rate": 1.8248990539211596e-05,
+      "loss": 3.3964,
+      "step": 1121
+    },
+    {
+      "epoch": 0.21586263286999183,
+      "grad_norm": 2.902687756720181,
+      "learning_rate": 1.8245466065689282e-05,
+      "loss": 2.7776,
+      "step": 1122
+    },
+    {
+      "epoch": 0.21605502380837863,
+      "grad_norm": 3.5795215895976717,
+      "learning_rate": 1.824193838972297e-05,
+      "loss": 3.3175,
+      "step": 1123
+    },
+    {
+      "epoch": 0.21624741474676543,
+      "grad_norm": 3.5642092473642055,
+      "learning_rate": 1.823840751268275e-05,
+      "loss": 3.1275,
+      "step": 1124
+    },
+    {
+      "epoch": 0.21643980568515223,
+      "grad_norm": 3.8185084581325124,
+      "learning_rate": 1.8234873435939987e-05,
+      "loss": 3.2753,
+      "step": 1125
+    },
+    {
+      "epoch": 0.21663219662353903,
+      "grad_norm": 2.311324671698934,
+      "learning_rate": 1.8231336160867275e-05,
+      "loss": 2.8352,
+      "step": 1126
+    },
+    {
+      "epoch": 0.21682458756192582,
+      "grad_norm": 3.732946006110883,
+      "learning_rate": 1.8227795688838446e-05,
+      "loss": 3.3323,
+      "step": 1127
+    },
+    {
+      "epoch": 0.21701697850031262,
+      "grad_norm": 3.8180370078692016,
+      "learning_rate": 1.822425202122858e-05,
+      "loss": 3.4266,
+      "step": 1128
+    },
+    {
+      "epoch": 0.21720936943869945,
+      "grad_norm": 3.584271525398372,
+      "learning_rate": 1.8220705159413996e-05,
+      "loss": 2.9952,
+      "step": 1129
+    },
+    {
+      "epoch": 0.21740176037708625,
+      "grad_norm": 3.3622536989155583,
+      "learning_rate": 1.8217155104772256e-05,
+      "loss": 3.4404,
+      "step": 1130
+    },
+    {
+      "epoch": 0.21759415131547305,
+      "grad_norm": 3.823307120996351,
+      "learning_rate": 1.8213601858682158e-05,
+      "loss": 3.0944,
+      "step": 1131
+    },
+    {
+      "epoch": 0.21778654225385985,
+      "grad_norm": 3.164253266869665,
+      "learning_rate": 1.8210045422523744e-05,
+      "loss": 3.1197,
+      "step": 1132
+    },
+    {
+      "epoch": 0.21797893319224665,
+      "grad_norm": 3.4423707363371703,
+      "learning_rate": 1.8206485797678294e-05,
+      "loss": 3.2787,
+      "step": 1133
+    },
+    {
+      "epoch": 0.21817132413063345,
+      "grad_norm": 3.9685916514916397,
+      "learning_rate": 1.820292298552832e-05,
+      "loss": 3.4923,
+      "step": 1134
+    },
+    {
+      "epoch": 0.21836371506902025,
+      "grad_norm": 4.03965380559882,
+      "learning_rate": 1.819935698745759e-05,
+      "loss": 3.0117,
+      "step": 1135
+    },
+    {
+      "epoch": 0.21855610600740705,
+      "grad_norm": 3.6058786114430843,
+      "learning_rate": 1.8195787804851076e-05,
+      "loss": 3.3693,
+      "step": 1136
+    },
+    {
+      "epoch": 0.21874849694579385,
+      "grad_norm": 3.8235461452782027,
+      "learning_rate": 1.8192215439095025e-05,
+      "loss": 3.3758,
+      "step": 1137
+    },
+    {
+      "epoch": 0.21894088788418065,
+      "grad_norm": 3.1928562308130912,
+      "learning_rate": 1.8188639891576893e-05,
+      "loss": 3.117,
+      "step": 1138
+    },
+    {
+      "epoch": 0.21913327882256745,
+      "grad_norm": 3.6835333907984915,
+      "learning_rate": 1.8185061163685386e-05,
+      "loss": 3.4716,
+      "step": 1139
+    },
+    {
+      "epoch": 0.21932566976095427,
+      "grad_norm": 2.3597801460789833,
+      "learning_rate": 1.818147925681044e-05,
+      "loss": 2.6651,
+      "step": 1140
+    },
+    {
+      "epoch": 0.21951806069934107,
+      "grad_norm": 4.03259883670465,
+      "learning_rate": 1.8177894172343227e-05,
+      "loss": 3.1774,
+      "step": 1141
+    },
+    {
+      "epoch": 0.21971045163772787,
+      "grad_norm": 3.699034405103865,
+      "learning_rate": 1.817430591167615e-05,
+      "loss": 2.9907,
+      "step": 1142
+    },
+    {
+      "epoch": 0.21990284257611467,
+      "grad_norm": 3.9512152511924277,
+      "learning_rate": 1.8170714476202848e-05,
+      "loss": 3.2757,
+      "step": 1143
+    },
+    {
+      "epoch": 0.22009523351450147,
+      "grad_norm": 3.700846689540025,
+      "learning_rate": 1.8167119867318197e-05,
+      "loss": 3.6028,
+      "step": 1144
+    },
+    {
+      "epoch": 0.22028762445288827,
+      "grad_norm": 4.024495938298642,
+      "learning_rate": 1.81635220864183e-05,
+      "loss": 3.3616,
+      "step": 1145
+    },
+    {
+      "epoch": 0.22048001539127507,
+      "grad_norm": 3.7794710880809195,
+      "learning_rate": 1.8159921134900486e-05,
+      "loss": 3.486,
+      "step": 1146
+    },
+    {
+      "epoch": 0.22067240632966187,
+      "grad_norm": 3.3394860036325236,
+      "learning_rate": 1.8156317014163337e-05,
+      "loss": 3.3156,
+      "step": 1147
+    },
+    {
+      "epoch": 0.22086479726804867,
+      "grad_norm": 3.773118374117237,
+      "learning_rate": 1.8152709725606642e-05,
+      "loss": 3.3154,
+      "step": 1148
+    },
+    {
+      "epoch": 0.22105718820643547,
+      "grad_norm": 3.728574792638062,
+      "learning_rate": 1.8149099270631434e-05,
+      "loss": 3.2453,
+      "step": 1149
+    },
+    {
+      "epoch": 0.22124957914482227,
+      "grad_norm": 3.7318897571145753,
+      "learning_rate": 1.8145485650639973e-05,
+      "loss": 3.3738,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2214419700832091,
+      "grad_norm": 3.4201501649227897,
+      "learning_rate": 1.8141868867035745e-05,
+      "loss": 3.4207,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2216343610215959,
+      "grad_norm": 3.458344870862233,
+      "learning_rate": 1.8138248921223465e-05,
+      "loss": 3.1383,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2218267519599827,
+      "grad_norm": 3.4723888847746016,
+      "learning_rate": 1.8134625814609084e-05,
+      "loss": 3.3182,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2220191428983695,
+      "grad_norm": 3.6065189734609175,
+      "learning_rate": 1.8130999548599767e-05,
+      "loss": 3.1431,
+      "step": 1154
+    },
+    {
+      "epoch": 0.2222115338367563,
+      "grad_norm": 3.7998977195408523,
+      "learning_rate": 1.8127370124603927e-05,
+      "loss": 3.3298,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2224039247751431,
+      "grad_norm": 3.748405180128046,
+      "learning_rate": 1.8123737544031178e-05,
+      "loss": 3.0319,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2225963157135299,
+      "grad_norm": 3.8754924886288147,
+      "learning_rate": 1.8120101808292373e-05,
+      "loss": 3.2765,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2227887066519167,
+      "grad_norm": 2.385838907123001,
+      "learning_rate": 1.81164629187996e-05,
+      "loss": 2.8617,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2229810975903035,
+      "grad_norm": 3.652061445560373,
+      "learning_rate": 1.811282087696615e-05,
+      "loss": 3.4552,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2231734885286903,
+      "grad_norm": 3.8215980496478252,
+      "learning_rate": 1.8109175684206558e-05,
+      "loss": 3.2307,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2233658794670771,
+      "grad_norm": 3.776811925815661,
+      "learning_rate": 1.8105527341936574e-05,
+      "loss": 3.1895,
+      "step": 1161
+    },
+    {
+      "epoch": 0.22355827040546392,
+      "grad_norm": 3.6725113406951237,
+      "learning_rate": 1.810187585157317e-05,
+      "loss": 3.3874,
+      "step": 1162
+    },
+    {
+      "epoch": 0.22375066134385072,
+      "grad_norm": 3.5202527971107576,
+      "learning_rate": 1.8098221214534543e-05,
+      "loss": 3.2391,
+      "step": 1163
+    },
+    {
+      "epoch": 0.22394305228223751,
+      "grad_norm": 3.7278530751349743,
+      "learning_rate": 1.8094563432240107e-05,
+      "loss": 3.2792,
+      "step": 1164
+    },
+    {
+      "epoch": 0.22413544322062431,
+      "grad_norm": 3.63164154889994,
+      "learning_rate": 1.8090902506110513e-05,
+      "loss": 3.1332,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2243278341590111,
+      "grad_norm": 3.447620692051021,
+      "learning_rate": 1.8087238437567614e-05,
+      "loss": 3.1867,
+      "step": 1166
+    },
+    {
+      "epoch": 0.2245202250973979,
+      "grad_norm": 3.839244656219176,
+      "learning_rate": 1.8083571228034498e-05,
+      "loss": 3.2413,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2247126160357847,
+      "grad_norm": 3.5545013920109825,
+      "learning_rate": 1.807990087893546e-05,
+      "loss": 3.3006,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2249050069741715,
+      "grad_norm": 3.3661919201466457,
+      "learning_rate": 1.807622739169603e-05,
+      "loss": 3.3642,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2250973979125583,
+      "grad_norm": 3.7732111001981696,
+      "learning_rate": 1.807255076774294e-05,
+      "loss": 3.2222,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2252897888509451,
+      "grad_norm": 3.540739540268087,
+      "learning_rate": 1.8068871008504153e-05,
+      "loss": 3.2107,
+      "step": 1171
+    },
+    {
+      "epoch": 0.2254821797893319,
+      "grad_norm": 2.340770270632457,
+      "learning_rate": 1.8065188115408844e-05,
+      "loss": 2.5624,
+      "step": 1172
+    },
+    {
+      "epoch": 0.22567457072771874,
+      "grad_norm": 3.735750511537603,
+      "learning_rate": 1.8061502089887406e-05,
+      "loss": 3.2648,
+      "step": 1173
+    },
+    {
+      "epoch": 0.22586696166610554,
+      "grad_norm": 3.718528424444349,
+      "learning_rate": 1.805781293337145e-05,
+      "loss": 3.1233,
+      "step": 1174
+    },
+    {
+      "epoch": 0.22605935260449234,
+      "grad_norm": 3.655205895460516,
+      "learning_rate": 1.8054120647293798e-05,
+      "loss": 3.3154,
+      "step": 1175
+    },
+    {
+      "epoch": 0.22625174354287914,
+      "grad_norm": 3.8657338571050395,
+      "learning_rate": 1.8050425233088496e-05,
+      "loss": 3.5108,
+      "step": 1176
+    },
+    {
+      "epoch": 0.22644413448126594,
+      "grad_norm": 3.280540829734142,
+      "learning_rate": 1.80467266921908e-05,
+      "loss": 3.1047,
+      "step": 1177
+    },
+    {
+      "epoch": 0.22663652541965273,
+      "grad_norm": 3.5083932357363423,
+      "learning_rate": 1.8043025026037178e-05,
+      "loss": 3.3671,
+      "step": 1178
+    },
+    {
+      "epoch": 0.22682891635803953,
+      "grad_norm": 3.813268308879681,
+      "learning_rate": 1.8039320236065314e-05,
+      "loss": 2.909,
+      "step": 1179
+    },
+    {
+      "epoch": 0.22702130729642633,
+      "grad_norm": 3.6227512242277977,
+      "learning_rate": 1.803561232371411e-05,
+      "loss": 3.0294,
+      "step": 1180
+    },
+    {
+      "epoch": 0.22721369823481313,
+      "grad_norm": 4.102201348087501,
+      "learning_rate": 1.803190129042367e-05,
+      "loss": 3.3029,
+      "step": 1181
+    },
+    {
+      "epoch": 0.22740608917319993,
+      "grad_norm": 3.905819413381657,
+      "learning_rate": 1.8028187137635325e-05,
+      "loss": 3.1419,
+      "step": 1182
+    },
+    {
+      "epoch": 0.22759848011158673,
+      "grad_norm": 3.7273824526522708,
+      "learning_rate": 1.8024469866791602e-05,
+      "loss": 3.2607,
+      "step": 1183
+    },
+    {
+      "epoch": 0.22779087104997356,
+      "grad_norm": 3.4302452758137125,
+      "learning_rate": 1.802074947933625e-05,
+      "loss": 2.9422,
+      "step": 1184
+    },
+    {
+      "epoch": 0.22798326198836036,
+      "grad_norm": 3.896770658068382,
+      "learning_rate": 1.801702597671422e-05,
+      "loss": 3.3165,
+      "step": 1185
+    },
+    {
+      "epoch": 0.22817565292674716,
+      "grad_norm": 3.887079412910981,
+      "learning_rate": 1.8013299360371685e-05,
+      "loss": 3.2146,
+      "step": 1186
+    },
+    {
+      "epoch": 0.22836804386513396,
+      "grad_norm": 3.5750409397770193,
+      "learning_rate": 1.8009569631756013e-05,
+      "loss": 3.4012,
+      "step": 1187
+    },
+    {
+      "epoch": 0.22856043480352076,
+      "grad_norm": 3.6538608669722787,
+      "learning_rate": 1.8005836792315793e-05,
+      "loss": 3.1355,
+      "step": 1188
+    },
+    {
+      "epoch": 0.22875282574190756,
+      "grad_norm": 3.810512506739271,
+      "learning_rate": 1.800210084350081e-05,
+      "loss": 3.3147,
+      "step": 1189
+    },
+    {
+      "epoch": 0.22894521668029436,
+      "grad_norm": 3.5659671361892102,
+      "learning_rate": 1.799836178676207e-05,
+      "loss": 3.3401,
+      "step": 1190
+    },
+    {
+      "epoch": 0.22913760761868116,
+      "grad_norm": 3.7428470759162544,
+      "learning_rate": 1.799461962355178e-05,
+      "loss": 3.3341,
+      "step": 1191
+    },
+    {
+      "epoch": 0.22932999855706795,
+      "grad_norm": 3.5918040353157665,
+      "learning_rate": 1.7990874355323345e-05,
+      "loss": 3.2578,
+      "step": 1192
+    },
+    {
+      "epoch": 0.22952238949545475,
+      "grad_norm": 2.4683272286665328,
+      "learning_rate": 1.7987125983531393e-05,
+      "loss": 2.7909,
+      "step": 1193
+    },
+    {
+      "epoch": 0.22971478043384155,
+      "grad_norm": 3.6546223167902165,
+      "learning_rate": 1.7983374509631742e-05,
+      "loss": 2.9416,
+      "step": 1194
+    },
+    {
+      "epoch": 0.22990717137222838,
+      "grad_norm": 3.544122524690546,
+      "learning_rate": 1.7979619935081424e-05,
+      "loss": 3.4315,
+      "step": 1195
+    },
+    {
+      "epoch": 0.23009956231061518,
+      "grad_norm": 4.140652728705349,
+      "learning_rate": 1.797586226133867e-05,
+      "loss": 3.3225,
+      "step": 1196
+    },
+    {
+      "epoch": 0.23029195324900198,
+      "grad_norm": 3.458061407167826,
+      "learning_rate": 1.7972101489862924e-05,
+      "loss": 3.2251,
+      "step": 1197
+    },
+    {
+      "epoch": 0.23048434418738878,
+      "grad_norm": 3.722180508255122,
+      "learning_rate": 1.7968337622114824e-05,
+      "loss": 3.4521,
+      "step": 1198
+    },
+    {
+      "epoch": 0.23067673512577558,
+      "grad_norm": 3.3402231434957286,
+      "learning_rate": 1.7964570659556206e-05,
+      "loss": 2.9212,
+      "step": 1199
+    },
+    {
+      "epoch": 0.23086912606416238,
+      "grad_norm": 3.544229919017743,
+      "learning_rate": 1.796080060365012e-05,
+      "loss": 3.4461,
+      "step": 1200
+    },
+    {
+      "epoch": 0.23106151700254918,
+      "grad_norm": 3.570429597673848,
+      "learning_rate": 1.7957027455860815e-05,
+      "loss": 3.4735,
+      "step": 1201
+    },
+    {
+      "epoch": 0.23125390794093598,
+      "grad_norm": 3.3779758572825775,
+      "learning_rate": 1.795325121765373e-05,
+      "loss": 3.0721,
+      "step": 1202
+    },
+    {
+      "epoch": 0.23144629887932278,
+      "grad_norm": 3.7136543187333584,
+      "learning_rate": 1.794947189049552e-05,
+      "loss": 3.246,
+      "step": 1203
+    },
+    {
+      "epoch": 0.23163868981770958,
+      "grad_norm": 3.85206937550551,
+      "learning_rate": 1.7945689475854033e-05,
+      "loss": 3.1097,
+      "step": 1204
+    },
+    {
+      "epoch": 0.23183108075609637,
+      "grad_norm": 3.661468154849637,
+      "learning_rate": 1.7941903975198305e-05,
+      "loss": 3.2463,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2320234716944832,
+      "grad_norm": 3.853595922039581,
+      "learning_rate": 1.7938115389998595e-05,
+      "loss": 2.8962,
+      "step": 1206
+    },
+    {
+      "epoch": 0.23221586263287,
+      "grad_norm": 3.8214569201197066,
+      "learning_rate": 1.7934323721726334e-05,
+      "loss": 3.1934,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2324082535712568,
+      "grad_norm": 3.7666414043460423,
+      "learning_rate": 1.7930528971854166e-05,
+      "loss": 3.5081,
+      "step": 1208
+    },
+    {
+      "epoch": 0.2326006445096436,
+      "grad_norm": 3.6321183678891824,
+      "learning_rate": 1.792673114185593e-05,
+      "loss": 3.1487,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2327930354480304,
+      "grad_norm": 3.63015249348842,
+      "learning_rate": 1.7922930233206656e-05,
+      "loss": 3.2473,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2329854263864172,
+      "grad_norm": 3.8904699500867364,
+      "learning_rate": 1.7919126247382576e-05,
+      "loss": 3.1662,
+      "step": 1211
+    },
+    {
+      "epoch": 0.233177817324804,
+      "grad_norm": 3.5744716949642013,
+      "learning_rate": 1.791531918586112e-05,
+      "loss": 3.3149,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2333702082631908,
+      "grad_norm": 3.6562102435131476,
+      "learning_rate": 1.7911509050120892e-05,
+      "loss": 3.2977,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2335625992015776,
+      "grad_norm": 3.5406650438404017,
+      "learning_rate": 1.7907695841641716e-05,
+      "loss": 3.5762,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2337549901399644,
+      "grad_norm": 3.421370113969954,
+      "learning_rate": 1.7903879561904597e-05,
+      "loss": 3.1075,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2339473810783512,
+      "grad_norm": 3.5256648683581746,
+      "learning_rate": 1.790006021239173e-05,
+      "loss": 3.1988,
+      "step": 1216
+    },
+    {
+      "epoch": 0.23413977201673802,
+      "grad_norm": 3.6592919537765307,
+      "learning_rate": 1.789623779458651e-05,
+      "loss": 3.2118,
+      "step": 1217
+    },
+    {
+      "epoch": 0.23433216295512482,
+      "grad_norm": 3.426736419740435,
+      "learning_rate": 1.789241230997352e-05,
+      "loss": 3.2201,
+      "step": 1218
+    },
+    {
+      "epoch": 0.23452455389351162,
+      "grad_norm": 3.468466994467266,
+      "learning_rate": 1.7888583760038534e-05,
+      "loss": 3.2687,
+      "step": 1219
+    },
+    {
+      "epoch": 0.23471694483189842,
+      "grad_norm": 3.8088616117307326,
+      "learning_rate": 1.7884752146268513e-05,
+      "loss": 3.3056,
+      "step": 1220
+    },
+    {
+      "epoch": 0.23490933577028522,
+      "grad_norm": 3.9345853972836053,
+      "learning_rate": 1.7880917470151614e-05,
+      "loss": 3.1702,
+      "step": 1221
+    },
+    {
+      "epoch": 0.23510172670867202,
+      "grad_norm": 3.538227542328572,
+      "learning_rate": 1.7877079733177185e-05,
+      "loss": 3.3349,
+      "step": 1222
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 3.7738564738089906,
+      "learning_rate": 1.7873238936835754e-05,
+      "loss": 3.275,
+      "step": 1223
+    },
+    {
+      "epoch": 0.23548650858544562,
+      "grad_norm": 3.642851093887507,
+      "learning_rate": 1.786939508261904e-05,
+      "loss": 3.1769,
+      "step": 1224
+    },
+    {
+      "epoch": 0.23567889952383242,
+      "grad_norm": 3.7407100132949913,
+      "learning_rate": 1.786554817201996e-05,
+      "loss": 3.2466,
+      "step": 1225
+    },
+    {
+      "epoch": 0.23587129046221922,
+      "grad_norm": 3.684021005128528,
+      "learning_rate": 1.78616982065326e-05,
+      "loss": 3.4386,
+      "step": 1226
+    },
+    {
+      "epoch": 0.23606368140060602,
+      "grad_norm": 3.2780544452914104,
+      "learning_rate": 1.785784518765225e-05,
+      "loss": 3.4094,
+      "step": 1227
+    },
+    {
+      "epoch": 0.23625607233899285,
+      "grad_norm": 3.3437309855531288,
+      "learning_rate": 1.7853989116875373e-05,
+      "loss": 3.1323,
+      "step": 1228
+    },
+    {
+      "epoch": 0.23644846327737964,
+      "grad_norm": 3.5032254504256612,
+      "learning_rate": 1.7850129995699626e-05,
+      "loss": 3.1531,
+      "step": 1229
+    },
+    {
+      "epoch": 0.23664085421576644,
+      "grad_norm": 3.6982344145468464,
+      "learning_rate": 1.7846267825623843e-05,
+      "loss": 3.4765,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23683324515415324,
+      "grad_norm": 3.148461522703182,
+      "learning_rate": 1.7842402608148053e-05,
+      "loss": 3.0939,
+      "step": 1231
+    },
+    {
+      "epoch": 0.23702563609254004,
+      "grad_norm": 3.416225265690513,
+      "learning_rate": 1.7838534344773453e-05,
+      "loss": 3.4387,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23721802703092684,
+      "grad_norm": 3.6280505063728605,
+      "learning_rate": 1.7834663037002444e-05,
+      "loss": 3.1976,
+      "step": 1233
+    },
+    {
+      "epoch": 0.23741041796931364,
+      "grad_norm": 3.49536631208839,
+      "learning_rate": 1.7830788686338586e-05,
+      "loss": 3.5853,
+      "step": 1234
+    },
+    {
+      "epoch": 0.23760280890770044,
+      "grad_norm": 3.4545620158116326,
+      "learning_rate": 1.7826911294286636e-05,
+      "loss": 3.42,
+      "step": 1235
+    },
+    {
+      "epoch": 0.23779519984608724,
+      "grad_norm": 3.6396403680101095,
+      "learning_rate": 1.782303086235253e-05,
+      "loss": 3.5826,
+      "step": 1236
+    },
+    {
+      "epoch": 0.23798759078447404,
+      "grad_norm": 3.5566471617619997,
+      "learning_rate": 1.781914739204338e-05,
+      "loss": 3.0901,
+      "step": 1237
+    },
+    {
+      "epoch": 0.23817998172286084,
+      "grad_norm": 3.5228643048468467,
+      "learning_rate": 1.7815260884867486e-05,
+      "loss": 3.3107,
+      "step": 1238
+    },
+    {
+      "epoch": 0.23837237266124767,
+      "grad_norm": 3.753512661791852,
+      "learning_rate": 1.781137134233432e-05,
+      "loss": 3.2985,
+      "step": 1239
+    },
+    {
+      "epoch": 0.23856476359963447,
+      "grad_norm": 3.5983461950443214,
+      "learning_rate": 1.7807478765954532e-05,
+      "loss": 3.025,
+      "step": 1240
+    },
+    {
+      "epoch": 0.23875715453802127,
+      "grad_norm": 3.610312213509601,
+      "learning_rate": 1.7803583157239958e-05,
+      "loss": 3.532,
+      "step": 1241
+    },
+    {
+      "epoch": 0.23894954547640807,
+      "grad_norm": 3.779550600043719,
+      "learning_rate": 1.7799684517703605e-05,
+      "loss": 3.2155,
+      "step": 1242
+    },
+    {
+      "epoch": 0.23914193641479486,
+      "grad_norm": 3.469224629931652,
+      "learning_rate": 1.779578284885966e-05,
+      "loss": 3.3263,
+      "step": 1243
+    },
+    {
+      "epoch": 0.23933432735318166,
+      "grad_norm": 3.3883621289112456,
+      "learning_rate": 1.779187815222349e-05,
+      "loss": 3.2731,
+      "step": 1244
+    },
+    {
+      "epoch": 0.23952671829156846,
+      "grad_norm": 3.381521849246229,
+      "learning_rate": 1.778797042931163e-05,
+      "loss": 3.174,
+      "step": 1245
+    },
+    {
+      "epoch": 0.23971910922995526,
+      "grad_norm": 3.6594035031509766,
+      "learning_rate": 1.7784059681641798e-05,
+      "loss": 3.2058,
+      "step": 1246
+    },
+    {
+      "epoch": 0.23991150016834206,
+      "grad_norm": 3.578263588229496,
+      "learning_rate": 1.778014591073288e-05,
+      "loss": 3.2923,
+      "step": 1247
+    },
+    {
+      "epoch": 0.24010389110672886,
+      "grad_norm": 3.5626691322864636,
+      "learning_rate": 1.777622911810494e-05,
+      "loss": 3.0991,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2402962820451157,
+      "grad_norm": 3.4134678708294883,
+      "learning_rate": 1.777230930527922e-05,
+      "loss": 3.251,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2404886729835025,
+      "grad_norm": 3.370575115401204,
+      "learning_rate": 1.7768386473778124e-05,
+      "loss": 3.2923,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2406810639218893,
+      "grad_norm": 3.8376062274494376,
+      "learning_rate": 1.7764460625125236e-05,
+      "loss": 3.3162,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2408734548602761,
+      "grad_norm": 3.676568898011241,
+      "learning_rate": 1.776053176084531e-05,
+      "loss": 3.3818,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2410658457986629,
+      "grad_norm": 2.761493477923443,
+      "learning_rate": 1.7756599882464274e-05,
+      "loss": 2.8916,
+      "step": 1253
+    },
+    {
+      "epoch": 0.24125823673704969,
+      "grad_norm": 3.485692095365088,
+      "learning_rate": 1.7752664991509224e-05,
+      "loss": 2.9124,
+      "step": 1254
+    },
+    {
+      "epoch": 0.24145062767543649,
+      "grad_norm": 4.010119160356957,
+      "learning_rate": 1.7748727089508423e-05,
+      "loss": 3.2991,
+      "step": 1255
+    },
+    {
+      "epoch": 0.24164301861382328,
+      "grad_norm": 3.910981407663038,
+      "learning_rate": 1.7744786177991307e-05,
+      "loss": 3.339,
+      "step": 1256
+    },
+    {
+      "epoch": 0.24183540955221008,
+      "grad_norm": 3.6149846140535176,
+      "learning_rate": 1.774084225848849e-05,
+      "loss": 3.3851,
+      "step": 1257
+    },
+    {
+      "epoch": 0.24202780049059688,
+      "grad_norm": 3.6788313117238043,
+      "learning_rate": 1.773689533253173e-05,
+      "loss": 3.4519,
+      "step": 1258
+    },
+    {
+      "epoch": 0.24222019142898368,
+      "grad_norm": 3.5556316581252823,
+      "learning_rate": 1.7732945401653978e-05,
+      "loss": 3.1903,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2424125823673705,
+      "grad_norm": 3.702759804995198,
+      "learning_rate": 1.7728992467389342e-05,
+      "loss": 3.1276,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2426049733057573,
+      "grad_norm": 3.330205476405611,
+      "learning_rate": 1.7725036531273087e-05,
+      "loss": 3.2022,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2427973642441441,
+      "grad_norm": 3.7230762289275248,
+      "learning_rate": 1.7721077594841663e-05,
+      "loss": 3.0985,
+      "step": 1262
+    },
+    {
+      "epoch": 0.2429897551825309,
+      "grad_norm": 2.4250614374645263,
+      "learning_rate": 1.771711565963267e-05,
+      "loss": 2.745,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2431821461209177,
+      "grad_norm": 3.4758712041328628,
+      "learning_rate": 1.7713150727184878e-05,
+      "loss": 3.143,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2433745370593045,
+      "grad_norm": 3.750634762910219,
+      "learning_rate": 1.770918279903822e-05,
+      "loss": 3.2823,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2435669279976913,
+      "grad_norm": 3.4032819272773436,
+      "learning_rate": 1.77052118767338e-05,
+      "loss": 3.3783,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2437593189360781,
+      "grad_norm": 3.2630617120207646,
+      "learning_rate": 1.7701237961813874e-05,
+      "loss": 3.0467,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2439517098744649,
+      "grad_norm": 3.354443378165428,
+      "learning_rate": 1.7697261055821864e-05,
+      "loss": 3.0717,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2441441008128517,
+      "grad_norm": 3.7484785857963017,
+      "learning_rate": 1.7693281160302354e-05,
+      "loss": 3.3377,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2443364917512385,
+      "grad_norm": 3.7382556529264273,
+      "learning_rate": 1.7689298276801095e-05,
+      "loss": 3.451,
+      "step": 1270
+    },
+    {
+      "epoch": 0.24452888268962533,
+      "grad_norm": 3.301692321456445,
+      "learning_rate": 1.7685312406864986e-05,
+      "loss": 3.3316,
+      "step": 1271
+    },
+    {
+      "epoch": 0.24472127362801213,
+      "grad_norm": 3.508711032570739,
+      "learning_rate": 1.7681323552042094e-05,
+      "loss": 3.2802,
+      "step": 1272
+    },
+    {
+      "epoch": 0.24491366456639893,
+      "grad_norm": 3.6779783221217124,
+      "learning_rate": 1.767733171388165e-05,
+      "loss": 3.3672,
+      "step": 1273
+    },
+    {
+      "epoch": 0.24510605550478573,
+      "grad_norm": 3.5151006732092087,
+      "learning_rate": 1.7673336893934033e-05,
+      "loss": 3.4685,
+      "step": 1274
+    },
+    {
+      "epoch": 0.24529844644317253,
+      "grad_norm": 3.524760590647865,
+      "learning_rate": 1.7669339093750786e-05,
+      "loss": 3.5569,
+      "step": 1275
+    },
+    {
+      "epoch": 0.24549083738155933,
+      "grad_norm": 4.650612636351101,
+      "learning_rate": 1.766533831488461e-05,
+      "loss": 2.9633,
+      "step": 1276
+    },
+    {
+      "epoch": 0.24568322831994613,
+      "grad_norm": 3.659392539858296,
+      "learning_rate": 1.7661334558889357e-05,
+      "loss": 3.4155,
+      "step": 1277
+    },
+    {
+      "epoch": 0.24587561925833293,
+      "grad_norm": 3.350610367771228,
+      "learning_rate": 1.7657327827320046e-05,
+      "loss": 3.2799,
+      "step": 1278
+    },
+    {
+      "epoch": 0.24606801019671973,
+      "grad_norm": 3.468499691543615,
+      "learning_rate": 1.765331812173284e-05,
+      "loss": 3.4902,
+      "step": 1279
+    },
+    {
+      "epoch": 0.24626040113510653,
+      "grad_norm": 2.383848031285396,
+      "learning_rate": 1.7649305443685068e-05,
+      "loss": 2.7025,
+      "step": 1280
+    },
+    {
+      "epoch": 0.24645279207349333,
+      "grad_norm": 2.5271874655238893,
+      "learning_rate": 1.76452897947352e-05,
+      "loss": 2.7292,
+      "step": 1281
+    },
+    {
+      "epoch": 0.24664518301188015,
+      "grad_norm": 3.913155540634312,
+      "learning_rate": 1.7641271176442876e-05,
+      "loss": 3.4697,
+      "step": 1282
+    },
+    {
+      "epoch": 0.24683757395026695,
+      "grad_norm": 3.586731799127687,
+      "learning_rate": 1.7637249590368878e-05,
+      "loss": 3.1996,
+      "step": 1283
+    },
+    {
+      "epoch": 0.24702996488865375,
+      "grad_norm": 3.7750314991236134,
+      "learning_rate": 1.763322503807514e-05,
+      "loss": 3.493,
+      "step": 1284
+    },
+    {
+      "epoch": 0.24722235582704055,
+      "grad_norm": 3.9401532811822744,
+      "learning_rate": 1.7629197521124758e-05,
+      "loss": 3.3227,
+      "step": 1285
+    },
+    {
+      "epoch": 0.24741474676542735,
+      "grad_norm": 3.738709047749914,
+      "learning_rate": 1.7625167041081967e-05,
+      "loss": 3.298,
+      "step": 1286
+    },
+    {
+      "epoch": 0.24760713770381415,
+      "grad_norm": 3.7730266725960746,
+      "learning_rate": 1.7621133599512163e-05,
+      "loss": 2.9929,
+      "step": 1287
+    },
+    {
+      "epoch": 0.24779952864220095,
+      "grad_norm": 3.6988506447558693,
+      "learning_rate": 1.761709719798189e-05,
+      "loss": 3.2784,
+      "step": 1288
+    },
+    {
+      "epoch": 0.24799191958058775,
+      "grad_norm": 3.439699719725171,
+      "learning_rate": 1.761305783805883e-05,
+      "loss": 3.5196,
+      "step": 1289
+    },
+    {
+      "epoch": 0.24818431051897455,
+      "grad_norm": 3.7647472695974957,
+      "learning_rate": 1.7609015521311836e-05,
+      "loss": 2.9478,
+      "step": 1290
+    },
+    {
+      "epoch": 0.24837670145736135,
+      "grad_norm": 3.457933991100311,
+      "learning_rate": 1.7604970249310893e-05,
+      "loss": 3.4156,
+      "step": 1291
+    },
+    {
+      "epoch": 0.24856909239574815,
+      "grad_norm": 3.4423805025691445,
+      "learning_rate": 1.7600922023627137e-05,
+      "loss": 3.4623,
+      "step": 1292
+    },
+    {
+      "epoch": 0.24876148333413498,
+      "grad_norm": 3.6904070481965214,
+      "learning_rate": 1.759687084583285e-05,
+      "loss": 3.0947,
+      "step": 1293
+    },
+    {
+      "epoch": 0.24895387427252177,
+      "grad_norm": 3.676783483109481,
+      "learning_rate": 1.759281671750147e-05,
+      "loss": 3.4268,
+      "step": 1294
+    },
+    {
+      "epoch": 0.24914626521090857,
+      "grad_norm": 3.3883545980632395,
+      "learning_rate": 1.7588759640207564e-05,
+      "loss": 3.4759,
+      "step": 1295
+    },
+    {
+      "epoch": 0.24933865614929537,
+      "grad_norm": 3.3597763451945104,
+      "learning_rate": 1.7584699615526857e-05,
+      "loss": 3.3818,
+      "step": 1296
+    },
+    {
+      "epoch": 0.24953104708768217,
+      "grad_norm": 3.6191925098112394,
+      "learning_rate": 1.7580636645036224e-05,
+      "loss": 3.2303,
+      "step": 1297
+    },
+    {
+      "epoch": 0.24972343802606897,
+      "grad_norm": 3.4534553377916524,
+      "learning_rate": 1.757657073031367e-05,
+      "loss": 3.0834,
+      "step": 1298
+    },
+    {
+      "epoch": 0.24991582896445577,
+      "grad_norm": 3.568930500558017,
+      "learning_rate": 1.7572501872938343e-05,
+      "loss": 3.1835,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2501082199028426,
+      "grad_norm": 3.588344406235961,
+      "learning_rate": 1.756843007449055e-05,
+      "loss": 3.2983,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2503006108412294,
+      "grad_norm": 2.571964946372255,
+      "learning_rate": 1.7564355336551727e-05,
+      "loss": 2.6882,
+      "step": 1301
+    },
+    {
+      "epoch": 0.2504930017796162,
+      "grad_norm": 3.922570846410561,
+      "learning_rate": 1.7560277660704455e-05,
+      "loss": 3.3191,
+      "step": 1302
+    },
+    {
+      "epoch": 0.250685392718003,
+      "grad_norm": 3.3549238243751764,
+      "learning_rate": 1.755619704853246e-05,
+      "loss": 3.2112,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2508777836563898,
+      "grad_norm": 2.364419976968673,
+      "learning_rate": 1.7552113501620595e-05,
+      "loss": 2.765,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2510701745947766,
+      "grad_norm": 3.790265777046758,
+      "learning_rate": 1.7548027021554874e-05,
+      "loss": 3.4081,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2512625655331634,
+      "grad_norm": 3.6545237115263,
+      "learning_rate": 1.754393760992243e-05,
+      "loss": 3.2526,
+      "step": 1306
+    },
+    {
+      "epoch": 0.2514549564715502,
+      "grad_norm": 3.6567747804973343,
+      "learning_rate": 1.7539845268311548e-05,
+      "loss": 3.3444,
+      "step": 1307
+    },
+    {
+      "epoch": 0.251647347409937,
+      "grad_norm": 3.666851945696372,
+      "learning_rate": 1.7535749998311645e-05,
+      "loss": 3.489,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2518397383483238,
+      "grad_norm": 3.95375899188879,
+      "learning_rate": 1.753165180151328e-05,
+      "loss": 3.4998,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2520321292867106,
+      "grad_norm": 3.7016441198040435,
+      "learning_rate": 1.752755067950814e-05,
+      "loss": 3.0255,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2522245202250974,
+      "grad_norm": 3.4544836722814325,
+      "learning_rate": 1.752344663388906e-05,
+      "loss": 3.0758,
+      "step": 1311
+    },
+    {
+      "epoch": 0.2524169111634842,
+      "grad_norm": 3.811417392160884,
+      "learning_rate": 1.7519339666249997e-05,
+      "loss": 3.3543,
+      "step": 1312
+    },
+    {
+      "epoch": 0.252609302101871,
+      "grad_norm": 3.5298533389629223,
+      "learning_rate": 1.7515229778186052e-05,
+      "loss": 3.3795,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2528016930402578,
+      "grad_norm": 3.6936894271937817,
+      "learning_rate": 1.7511116971293463e-05,
+      "loss": 3.5007,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2529940839786446,
+      "grad_norm": 3.7061605420794446,
+      "learning_rate": 1.7507001247169587e-05,
+      "loss": 3.1851,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2531864749170314,
+      "grad_norm": 3.520594535638476,
+      "learning_rate": 1.7502882607412933e-05,
+      "loss": 3.1245,
+      "step": 1316
+    },
+    {
+      "epoch": 0.2533788658554182,
+      "grad_norm": 3.49371624541918,
+      "learning_rate": 1.749876105362313e-05,
+      "loss": 3.2557,
+      "step": 1317
+    },
+    {
+      "epoch": 0.253571256793805,
+      "grad_norm": 4.218476849569187,
+      "learning_rate": 1.7494636587400942e-05,
+      "loss": 3.2691,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2537636477321918,
+      "grad_norm": 3.8964453352828277,
+      "learning_rate": 1.749050921034826e-05,
+      "loss": 3.2602,
+      "step": 1319
+    },
+    {
+      "epoch": 0.25395603867057864,
+      "grad_norm": 3.86920497860535,
+      "learning_rate": 1.7486378924068123e-05,
+      "loss": 3.5926,
+      "step": 1320
+    },
+    {
+      "epoch": 0.25414842960896544,
+      "grad_norm": 3.5802151191798637,
+      "learning_rate": 1.748224573016467e-05,
+      "loss": 3.1491,
+      "step": 1321
+    },
+    {
+      "epoch": 0.25434082054735224,
+      "grad_norm": 2.88606597850098,
+      "learning_rate": 1.7478109630243195e-05,
+      "loss": 2.6708,
+      "step": 1322
+    },
+    {
+      "epoch": 0.25453321148573904,
+      "grad_norm": 3.562579750127719,
+      "learning_rate": 1.747397062591011e-05,
+      "loss": 3.131,
+      "step": 1323
+    },
+    {
+      "epoch": 0.25472560242412584,
+      "grad_norm": 3.6541705788672996,
+      "learning_rate": 1.746982871877296e-05,
+      "loss": 3.27,
+      "step": 1324
+    },
+    {
+      "epoch": 0.25491799336251264,
+      "grad_norm": 3.43191326894443,
+      "learning_rate": 1.7465683910440405e-05,
+      "loss": 3.1151,
+      "step": 1325
+    },
+    {
+      "epoch": 0.25511038430089944,
+      "grad_norm": 3.803784405596268,
+      "learning_rate": 1.7461536202522248e-05,
+      "loss": 3.4253,
+      "step": 1326
+    },
+    {
+      "epoch": 0.25530277523928624,
+      "grad_norm": 3.925717183226613,
+      "learning_rate": 1.745738559662941e-05,
+      "loss": 3.0036,
+      "step": 1327
+    },
+    {
+      "epoch": 0.25549516617767304,
+      "grad_norm": 3.692614112627578,
+      "learning_rate": 1.7453232094373936e-05,
+      "loss": 3.4558,
+      "step": 1328
+    },
+    {
+      "epoch": 0.25568755711605984,
+      "grad_norm": 3.530325392977712,
+      "learning_rate": 1.7449075697369005e-05,
+      "loss": 3.2168,
+      "step": 1329
+    },
+    {
+      "epoch": 0.25587994805444664,
+      "grad_norm": 3.6474184781604873,
+      "learning_rate": 1.7444916407228904e-05,
+      "loss": 3.1455,
+      "step": 1330
+    },
+    {
+      "epoch": 0.25607233899283344,
+      "grad_norm": 3.521623245317118,
+      "learning_rate": 1.744075422556906e-05,
+      "loss": 3.1478,
+      "step": 1331
+    },
+    {
+      "epoch": 0.25626472993122024,
+      "grad_norm": 3.581224791772049,
+      "learning_rate": 1.7436589154006014e-05,
+      "loss": 3.0578,
+      "step": 1332
+    },
+    {
+      "epoch": 0.25645712086960704,
+      "grad_norm": 3.706490957756989,
+      "learning_rate": 1.743242119415743e-05,
+      "loss": 3.3317,
+      "step": 1333
+    },
+    {
+      "epoch": 0.25664951180799384,
+      "grad_norm": 3.650433854213416,
+      "learning_rate": 1.7428250347642102e-05,
+      "loss": 3.3207,
+      "step": 1334
+    },
+    {
+      "epoch": 0.25684190274638063,
+      "grad_norm": 3.687779031620586,
+      "learning_rate": 1.7424076616079933e-05,
+      "loss": 3.4331,
+      "step": 1335
+    },
+    {
+      "epoch": 0.25703429368476743,
+      "grad_norm": 3.957389294530363,
+      "learning_rate": 1.7419900001091953e-05,
+      "loss": 3.3008,
+      "step": 1336
+    },
+    {
+      "epoch": 0.25722668462315423,
+      "grad_norm": 3.5509136586307566,
+      "learning_rate": 1.7415720504300314e-05,
+      "loss": 3.4469,
+      "step": 1337
+    },
+    {
+      "epoch": 0.25741907556154103,
+      "grad_norm": 3.8333646316583665,
+      "learning_rate": 1.741153812732828e-05,
+      "loss": 3.1607,
+      "step": 1338
+    },
+    {
+      "epoch": 0.25761146649992783,
+      "grad_norm": 3.3158505176982938,
+      "learning_rate": 1.7407352871800246e-05,
+      "loss": 3.2443,
+      "step": 1339
+    },
+    {
+      "epoch": 0.25780385743831463,
+      "grad_norm": 3.4018273591291885,
+      "learning_rate": 1.7403164739341708e-05,
+      "loss": 3.4034,
+      "step": 1340
+    },
+    {
+      "epoch": 0.25799624837670143,
+      "grad_norm": 3.838895708826907,
+      "learning_rate": 1.739897373157929e-05,
+      "loss": 3.4236,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2581886393150883,
+      "grad_norm": 3.594104330161211,
+      "learning_rate": 1.7394779850140736e-05,
+      "loss": 3.3185,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2583810302534751,
+      "grad_norm": 3.948258119441056,
+      "learning_rate": 1.7390583096654895e-05,
+      "loss": 3.3036,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2585734211918619,
+      "grad_norm": 3.5037212742153265,
+      "learning_rate": 1.7386383472751745e-05,
+      "loss": 3.3843,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2587658121302487,
+      "grad_norm": 3.5742399949250605,
+      "learning_rate": 1.7382180980062365e-05,
+      "loss": 3.2685,
+      "step": 1345
+    },
+    {
+      "epoch": 0.2589582030686355,
+      "grad_norm": 3.617609447138076,
+      "learning_rate": 1.7377975620218954e-05,
+      "loss": 3.5237,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2591505940070223,
+      "grad_norm": 3.8327614034968174,
+      "learning_rate": 1.7373767394854836e-05,
+      "loss": 3.1651,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2593429849454091,
+      "grad_norm": 3.7352154953243595,
+      "learning_rate": 1.7369556305604422e-05,
+      "loss": 3.1257,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2595353758837959,
+      "grad_norm": 3.5599412753885438,
+      "learning_rate": 1.736534235410326e-05,
+      "loss": 3.0828,
+      "step": 1349
+    },
+    {
+      "epoch": 0.2597277668221827,
+      "grad_norm": 3.9679884008749564,
+      "learning_rate": 1.7361125541988e-05,
+      "loss": 3.6008,
+      "step": 1350
+    },
+    {
+      "epoch": 0.2599201577605695,
+      "grad_norm": 3.404290129520339,
+      "learning_rate": 1.7356905870896407e-05,
+      "loss": 3.3143,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2601125486989563,
+      "grad_norm": 2.777201204215896,
+      "learning_rate": 1.735268334246734e-05,
+      "loss": 2.8335,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2603049396373431,
+      "grad_norm": 3.3952472897472816,
+      "learning_rate": 1.7348457958340792e-05,
+      "loss": 3.3718,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2604973305757299,
+      "grad_norm": 3.7505200812004054,
+      "learning_rate": 1.7344229720157846e-05,
+      "loss": 3.4902,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2606897215141167,
+      "grad_norm": 3.514927840794614,
+      "learning_rate": 1.7339998629560705e-05,
+      "loss": 3.0875,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2608821124525035,
+      "grad_norm": 3.457755841254673,
+      "learning_rate": 1.7335764688192676e-05,
+      "loss": 3.4235,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2610745033908903,
+      "grad_norm": 3.7447933685985237,
+      "learning_rate": 1.733152789769817e-05,
+      "loss": 3.2313,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2612668943292771,
+      "grad_norm": 3.7438723081026373,
+      "learning_rate": 1.7327288259722714e-05,
+      "loss": 3.394,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2614592852676639,
+      "grad_norm": 3.734795467514534,
+      "learning_rate": 1.7323045775912927e-05,
+      "loss": 3.3445,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2616516762060507,
+      "grad_norm": 3.6586048593957683,
+      "learning_rate": 1.7318800447916543e-05,
+      "loss": 3.493,
+      "step": 1360
+    },
+    {
+      "epoch": 0.2618440671444375,
+      "grad_norm": 3.678360828968549,
+      "learning_rate": 1.7314552277382403e-05,
+      "loss": 3.4526,
+      "step": 1361
+    },
+    {
+      "epoch": 0.2620364580828243,
+      "grad_norm": 3.2126882975328575,
+      "learning_rate": 1.7310301265960446e-05,
+      "loss": 3.431,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2622288490212111,
+      "grad_norm": 3.5936761536821735,
+      "learning_rate": 1.7306047415301706e-05,
+      "loss": 3.3247,
+      "step": 1363
+    },
+    {
+      "epoch": 0.26242123995959793,
+      "grad_norm": 3.374904419520369,
+      "learning_rate": 1.7301790727058344e-05,
+      "loss": 3.3212,
+      "step": 1364
+    },
+    {
+      "epoch": 0.26261363089798473,
+      "grad_norm": 3.460698498518395,
+      "learning_rate": 1.7297531202883598e-05,
+      "loss": 3.1962,
+      "step": 1365
+    },
+    {
+      "epoch": 0.26280602183637153,
+      "grad_norm": 3.1886374816393164,
+      "learning_rate": 1.7293268844431826e-05,
+      "loss": 3.2244,
+      "step": 1366
+    },
+    {
+      "epoch": 0.26299841277475833,
+      "grad_norm": 3.346721433497498,
+      "learning_rate": 1.7289003653358472e-05,
+      "loss": 3.282,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2631908037131451,
+      "grad_norm": 3.401344133475771,
+      "learning_rate": 1.7284735631320093e-05,
+      "loss": 3.2416,
+      "step": 1368
+    },
+    {
+      "epoch": 0.2633831946515319,
+      "grad_norm": 3.5187467677116504,
+      "learning_rate": 1.7280464779974335e-05,
+      "loss": 3.2422,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2635755855899187,
+      "grad_norm": 3.610297774050357,
+      "learning_rate": 1.7276191100979952e-05,
+      "loss": 3.3095,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2637679765283055,
+      "grad_norm": 3.7586784043110657,
+      "learning_rate": 1.7271914595996784e-05,
+      "loss": 3.3919,
+      "step": 1371
+    },
+    {
+      "epoch": 0.2639603674666923,
+      "grad_norm": 3.451340508564228,
+      "learning_rate": 1.7267635266685782e-05,
+      "loss": 3.3532,
+      "step": 1372
+    },
+    {
+      "epoch": 0.2641527584050791,
+      "grad_norm": 3.7411850858929534,
+      "learning_rate": 1.7263353114708993e-05,
+      "loss": 3.0397,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2643451493434659,
+      "grad_norm": 3.7595973721607288,
+      "learning_rate": 1.7259068141729542e-05,
+      "loss": 3.0782,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2645375402818527,
+      "grad_norm": 2.329936174261849,
+      "learning_rate": 1.7254780349411677e-05,
+      "loss": 2.6692,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2647299312202395,
+      "grad_norm": 3.6619829874437673,
+      "learning_rate": 1.7250489739420718e-05,
+      "loss": 3.4781,
+      "step": 1376
+    },
+    {
+      "epoch": 0.2649223221586263,
+      "grad_norm": 3.629530505280693,
+      "learning_rate": 1.7246196313423095e-05,
+      "loss": 3.2416,
+      "step": 1377
+    },
+    {
+      "epoch": 0.2651147130970131,
+      "grad_norm": 3.7985521340249133,
+      "learning_rate": 1.7241900073086318e-05,
+      "loss": 3.3217,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2653071040353999,
+      "grad_norm": 3.5169270660130634,
+      "learning_rate": 1.7237601020079003e-05,
+      "loss": 3.0685,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2654994949737867,
+      "grad_norm": 3.6991130351735566,
+      "learning_rate": 1.7233299156070852e-05,
+      "loss": 3.3727,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2656918859121735,
+      "grad_norm": 3.7724338104285433,
+      "learning_rate": 1.7228994482732653e-05,
+      "loss": 3.1859,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2658842768505603,
+      "grad_norm": 3.7269054995460174,
+      "learning_rate": 1.72246870017363e-05,
+      "loss": 3.3974,
+      "step": 1382
+    },
+    {
+      "epoch": 0.2660766677889471,
+      "grad_norm": 3.6692432650670104,
+      "learning_rate": 1.7220376714754766e-05,
+      "loss": 3.1276,
+      "step": 1383
+    },
+    {
+      "epoch": 0.2662690587273339,
+      "grad_norm": 3.4506111117455878,
+      "learning_rate": 1.7216063623462112e-05,
+      "loss": 3.3206,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2664614496657207,
+      "grad_norm": 3.532217750815872,
+      "learning_rate": 1.7211747729533504e-05,
+      "loss": 3.2937,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2666538406041076,
+      "grad_norm": 3.4069116088935774,
+      "learning_rate": 1.7207429034645176e-05,
+      "loss": 3.4306,
+      "step": 1386
+    },
+    {
+      "epoch": 0.26684623154249437,
+      "grad_norm": 3.2133043070143237,
+      "learning_rate": 1.720310754047446e-05,
+      "loss": 3.2212,
+      "step": 1387
+    },
+    {
+      "epoch": 0.26703862248088117,
+      "grad_norm": 3.6668916868250285,
+      "learning_rate": 1.719878324869978e-05,
+      "loss": 3.0187,
+      "step": 1388
+    },
+    {
+      "epoch": 0.26723101341926797,
+      "grad_norm": 3.5656195659490884,
+      "learning_rate": 1.7194456161000634e-05,
+      "loss": 3.1672,
+      "step": 1389
+    },
+    {
+      "epoch": 0.26742340435765477,
+      "grad_norm": 3.690432065813498,
+      "learning_rate": 1.719012627905762e-05,
+      "loss": 3.5663,
+      "step": 1390
+    },
+    {
+      "epoch": 0.26761579529604157,
+      "grad_norm": 3.3930417962743156,
+      "learning_rate": 1.718579360455241e-05,
+      "loss": 3.2364,
+      "step": 1391
+    },
+    {
+      "epoch": 0.26780818623442837,
+      "grad_norm": 3.3031264477356723,
+      "learning_rate": 1.7181458139167767e-05,
+      "loss": 3.1957,
+      "step": 1392
+    },
+    {
+      "epoch": 0.26800057717281517,
+      "grad_norm": 3.7406331313640333,
+      "learning_rate": 1.7177119884587536e-05,
+      "loss": 3.2738,
+      "step": 1393
+    },
+    {
+      "epoch": 0.26819296811120197,
+      "grad_norm": 3.6055162323893737,
+      "learning_rate": 1.717277884249664e-05,
+      "loss": 3.2777,
+      "step": 1394
+    },
+    {
+      "epoch": 0.26838535904958877,
+      "grad_norm": 3.3226686839888506,
+      "learning_rate": 1.716843501458109e-05,
+      "loss": 3.1436,
+      "step": 1395
+    },
+    {
+      "epoch": 0.26857774998797557,
+      "grad_norm": 3.3078396522316313,
+      "learning_rate": 1.716408840252799e-05,
+      "loss": 3.496,
+      "step": 1396
+    },
+    {
+      "epoch": 0.26877014092636237,
+      "grad_norm": 3.643893223507906,
+      "learning_rate": 1.7159739008025503e-05,
+      "loss": 3.3685,
+      "step": 1397
+    },
+    {
+      "epoch": 0.26896253186474917,
+      "grad_norm": 3.553705328163313,
+      "learning_rate": 1.7155386832762892e-05,
+      "loss": 3.0234,
+      "step": 1398
+    },
+    {
+      "epoch": 0.26915492280313597,
+      "grad_norm": 3.6690174354385885,
+      "learning_rate": 1.715103187843048e-05,
+      "loss": 3.2307,
+      "step": 1399
+    },
+    {
+      "epoch": 0.26934731374152276,
+      "grad_norm": 3.3997824939970793,
+      "learning_rate": 1.7146674146719688e-05,
+      "loss": 3.539,
+      "step": 1400
+    },
+    {
+      "epoch": 0.26953970467990956,
+      "grad_norm": 3.57259003233204,
+      "learning_rate": 1.7142313639323012e-05,
+      "loss": 3.3893,
+      "step": 1401
+    },
+    {
+      "epoch": 0.26973209561829636,
+      "grad_norm": 3.9569284231039132,
+      "learning_rate": 1.7137950357934017e-05,
+      "loss": 3.3828,
+      "step": 1402
+    },
+    {
+      "epoch": 0.26992448655668316,
+      "grad_norm": 3.366357024677266,
+      "learning_rate": 1.7133584304247354e-05,
+      "loss": 3.2795,
+      "step": 1403
+    },
+    {
+      "epoch": 0.27011687749506996,
+      "grad_norm": 3.161643169842232,
+      "learning_rate": 1.7129215479958747e-05,
+      "loss": 3.3098,
+      "step": 1404
+    },
+    {
+      "epoch": 0.27030926843345676,
+      "grad_norm": 3.7321586748184377,
+      "learning_rate": 1.7124843886765e-05,
+      "loss": 3.1529,
+      "step": 1405
+    },
+    {
+      "epoch": 0.27050165937184356,
+      "grad_norm": 3.8214326124608,
+      "learning_rate": 1.712046952636398e-05,
+      "loss": 3.3189,
+      "step": 1406
+    },
+    {
+      "epoch": 0.27069405031023036,
+      "grad_norm": 2.460352347458236,
+      "learning_rate": 1.7116092400454655e-05,
+      "loss": 2.7022,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2708864412486172,
+      "grad_norm": 3.3937267329942364,
+      "learning_rate": 1.7111712510737035e-05,
+      "loss": 3.1669,
+      "step": 1408
+    },
+    {
+      "epoch": 0.271078832187004,
+      "grad_norm": 3.7432122271008086,
+      "learning_rate": 1.7107329858912226e-05,
+      "loss": 3.3573,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2712712231253908,
+      "grad_norm": 3.3773461551180626,
+      "learning_rate": 1.7102944446682393e-05,
+      "loss": 3.1407,
+      "step": 1410
+    },
+    {
+      "epoch": 0.2714636140637776,
+      "grad_norm": 3.2911270757583493,
+      "learning_rate": 1.709855627575079e-05,
+      "loss": 3.2553,
+      "step": 1411
+    },
+    {
+      "epoch": 0.2716560050021644,
+      "grad_norm": 3.836661602653968,
+      "learning_rate": 1.7094165347821724e-05,
+      "loss": 3.1497,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2718483959405512,
+      "grad_norm": 3.4254907417302265,
+      "learning_rate": 1.7089771664600584e-05,
+      "loss": 3.2883,
+      "step": 1413
+    },
+    {
+      "epoch": 0.272040786878938,
+      "grad_norm": 3.2044581857145236,
+      "learning_rate": 1.708537522779382e-05,
+      "loss": 3.2746,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2722331778173248,
+      "grad_norm": 3.5219155270703966,
+      "learning_rate": 1.7080976039108964e-05,
+      "loss": 3.0116,
+      "step": 1415
+    },
+    {
+      "epoch": 0.2724255687557116,
+      "grad_norm": 3.937972678812608,
+      "learning_rate": 1.7076574100254614e-05,
+      "loss": 3.154,
+      "step": 1416
+    },
+    {
+      "epoch": 0.2726179596940984,
+      "grad_norm": 3.7696474392876818,
+      "learning_rate": 1.707216941294042e-05,
+      "loss": 3.114,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2728103506324852,
+      "grad_norm": 3.6786027092457325,
+      "learning_rate": 1.706776197887712e-05,
+      "loss": 3.3391,
+      "step": 1418
+    },
+    {
+      "epoch": 0.273002741570872,
+      "grad_norm": 3.4845665442023317,
+      "learning_rate": 1.7063351799776514e-05,
+      "loss": 3.2218,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2731951325092588,
+      "grad_norm": 3.474980131114133,
+      "learning_rate": 1.7058938877351456e-05,
+      "loss": 3.2763,
+      "step": 1420
+    },
+    {
+      "epoch": 0.2733875234476456,
+      "grad_norm": 3.788691956153074,
+      "learning_rate": 1.705452321331588e-05,
+      "loss": 3.2226,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2735799143860324,
+      "grad_norm": 3.3610493351224897,
+      "learning_rate": 1.7050104809384774e-05,
+      "loss": 3.0999,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2737723053244192,
+      "grad_norm": 3.467483888398625,
+      "learning_rate": 1.70456836672742e-05,
+      "loss": 3.3412,
+      "step": 1423
+    },
+    {
+      "epoch": 0.273964696262806,
+      "grad_norm": 3.422912026027124,
+      "learning_rate": 1.704125978870128e-05,
+      "loss": 3.4585,
+      "step": 1424
+    },
+    {
+      "epoch": 0.2741570872011928,
+      "grad_norm": 3.8574245639269957,
+      "learning_rate": 1.7036833175384192e-05,
+      "loss": 3.1969,
+      "step": 1425
+    },
+    {
+      "epoch": 0.2743494781395796,
+      "grad_norm": 3.542252507638624,
+      "learning_rate": 1.7032403829042182e-05,
+      "loss": 3.2233,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2745418690779664,
+      "grad_norm": 3.6570713553780254,
+      "learning_rate": 1.7027971751395563e-05,
+      "loss": 3.3025,
+      "step": 1427
+    },
+    {
+      "epoch": 0.2747342600163532,
+      "grad_norm": 3.664154627984966,
+      "learning_rate": 1.7023536944165697e-05,
+      "loss": 3.1183,
+      "step": 1428
+    },
+    {
+      "epoch": 0.27492665095474006,
+      "grad_norm": 2.4225989634179563,
+      "learning_rate": 1.7019099409075014e-05,
+      "loss": 2.7723,
+      "step": 1429
+    },
+    {
+      "epoch": 0.27511904189312686,
+      "grad_norm": 3.798866815956034,
+      "learning_rate": 1.7014659147847005e-05,
+      "loss": 3.1042,
+      "step": 1430
+    },
+    {
+      "epoch": 0.27531143283151366,
+      "grad_norm": 3.5546710053841792,
+      "learning_rate": 1.701021616220621e-05,
+      "loss": 3.155,
+      "step": 1431
+    },
+    {
+      "epoch": 0.27550382376990046,
+      "grad_norm": 3.861759739083181,
+      "learning_rate": 1.7005770453878234e-05,
+      "loss": 3.4184,
+      "step": 1432
+    },
+    {
+      "epoch": 0.27569621470828726,
+      "grad_norm": 3.6547150756722924,
+      "learning_rate": 1.7001322024589742e-05,
+      "loss": 3.1603,
+      "step": 1433
+    },
+    {
+      "epoch": 0.27588860564667406,
+      "grad_norm": 3.4990430921855675,
+      "learning_rate": 1.6996870876068455e-05,
+      "loss": 3.3543,
+      "step": 1434
+    },
+    {
+      "epoch": 0.27608099658506086,
+      "grad_norm": 3.7322749902923884,
+      "learning_rate": 1.6992417010043144e-05,
+      "loss": 3.1793,
+      "step": 1435
+    },
+    {
+      "epoch": 0.27627338752344766,
+      "grad_norm": 3.349715424842203,
+      "learning_rate": 1.6987960428243637e-05,
+      "loss": 3.4522,
+      "step": 1436
+    },
+    {
+      "epoch": 0.27646577846183445,
+      "grad_norm": 3.284580454069076,
+      "learning_rate": 1.6983501132400825e-05,
+      "loss": 3.0136,
+      "step": 1437
+    },
+    {
+      "epoch": 0.27665816940022125,
+      "grad_norm": 3.645100774714857,
+      "learning_rate": 1.6979039124246643e-05,
+      "loss": 3.4225,
+      "step": 1438
+    },
+    {
+      "epoch": 0.27685056033860805,
+      "grad_norm": 3.840044981552036,
+      "learning_rate": 1.6974574405514083e-05,
+      "loss": 3.4384,
+      "step": 1439
+    },
+    {
+      "epoch": 0.27704295127699485,
+      "grad_norm": 3.786387975386778,
+      "learning_rate": 1.6970106977937192e-05,
+      "loss": 3.0139,
+      "step": 1440
+    },
+    {
+      "epoch": 0.27723534221538165,
+      "grad_norm": 3.690835025033368,
+      "learning_rate": 1.696563684325107e-05,
+      "loss": 3.1232,
+      "step": 1441
+    },
+    {
+      "epoch": 0.27742773315376845,
+      "grad_norm": 3.6268062212600136,
+      "learning_rate": 1.6961164003191862e-05,
+      "loss": 3.0671,
+      "step": 1442
+    },
+    {
+      "epoch": 0.27762012409215525,
+      "grad_norm": 3.1452038408154674,
+      "learning_rate": 1.6956688459496767e-05,
+      "loss": 3.2907,
+      "step": 1443
+    },
+    {
+      "epoch": 0.27781251503054205,
+      "grad_norm": 3.638113658245732,
+      "learning_rate": 1.695221021390404e-05,
+      "loss": 3.1046,
+      "step": 1444
+    },
+    {
+      "epoch": 0.27800490596892885,
+      "grad_norm": 3.6432105195207174,
+      "learning_rate": 1.6947729268152972e-05,
+      "loss": 3.3069,
+      "step": 1445
+    },
+    {
+      "epoch": 0.27819729690731565,
+      "grad_norm": 3.5827209310771346,
+      "learning_rate": 1.6943245623983918e-05,
+      "loss": 3.0844,
+      "step": 1446
+    },
+    {
+      "epoch": 0.27838968784570245,
+      "grad_norm": 3.495267604412483,
+      "learning_rate": 1.6938759283138268e-05,
+      "loss": 3.0842,
+      "step": 1447
+    },
+    {
+      "epoch": 0.27858207878408925,
+      "grad_norm": 3.766498187510574,
+      "learning_rate": 1.693427024735847e-05,
+      "loss": 3.2359,
+      "step": 1448
+    },
+    {
+      "epoch": 0.27877446972247605,
+      "grad_norm": 3.472218543925517,
+      "learning_rate": 1.692977851838801e-05,
+      "loss": 3.2895,
+      "step": 1449
+    },
+    {
+      "epoch": 0.27896686066086285,
+      "grad_norm": 3.5453554680622914,
+      "learning_rate": 1.6925284097971427e-05,
+      "loss": 3.2582,
+      "step": 1450
+    },
+    {
+      "epoch": 0.2791592515992497,
+      "grad_norm": 3.5563048115494253,
+      "learning_rate": 1.6920786987854296e-05,
+      "loss": 3.3365,
+      "step": 1451
+    },
+    {
+      "epoch": 0.2793516425376365,
+      "grad_norm": 2.434609747531918,
+      "learning_rate": 1.691628718978325e-05,
+      "loss": 2.7976,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2795440334760233,
+      "grad_norm": 3.6089849186490706,
+      "learning_rate": 1.691178470550596e-05,
+      "loss": 3.2146,
+      "step": 1453
+    },
+    {
+      "epoch": 0.2797364244144101,
+      "grad_norm": 3.679726742831733,
+      "learning_rate": 1.6907279536771127e-05,
+      "loss": 3.1535,
+      "step": 1454
+    },
+    {
+      "epoch": 0.2799288153527969,
+      "grad_norm": 3.5298132770120274,
+      "learning_rate": 1.6902771685328524e-05,
+      "loss": 2.9478,
+      "step": 1455
+    },
+    {
+      "epoch": 0.2801212062911837,
+      "grad_norm": 3.6557696179317287,
+      "learning_rate": 1.6898261152928933e-05,
+      "loss": 3.1255,
+      "step": 1456
+    },
+    {
+      "epoch": 0.2803135972295705,
+      "grad_norm": 3.614945462936983,
+      "learning_rate": 1.6893747941324197e-05,
+      "loss": 3.4056,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2805059881679573,
+      "grad_norm": 3.5308258502102707,
+      "learning_rate": 1.6889232052267203e-05,
+      "loss": 2.9878,
+      "step": 1458
+    },
+    {
+      "epoch": 0.2806983791063441,
+      "grad_norm": 3.623090036124668,
+      "learning_rate": 1.688471348751186e-05,
+      "loss": 3.3373,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2808907700447309,
+      "grad_norm": 3.411056489742332,
+      "learning_rate": 1.688019224881313e-05,
+      "loss": 3.3304,
+      "step": 1460
+    },
+    {
+      "epoch": 0.2810831609831177,
+      "grad_norm": 3.522424787411962,
+      "learning_rate": 1.6875668337927014e-05,
+      "loss": 3.3011,
+      "step": 1461
+    },
+    {
+      "epoch": 0.2812755519215045,
+      "grad_norm": 3.7467611286941205,
+      "learning_rate": 1.6871141756610544e-05,
+      "loss": 3.3018,
+      "step": 1462
+    },
+    {
+      "epoch": 0.2814679428598913,
+      "grad_norm": 3.773796717680108,
+      "learning_rate": 1.6866612506621788e-05,
+      "loss": 3.2402,
+      "step": 1463
+    },
+    {
+      "epoch": 0.2816603337982781,
+      "grad_norm": 3.4272549926091185,
+      "learning_rate": 1.6862080589719863e-05,
+      "loss": 3.0134,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2818527247366649,
+      "grad_norm": 3.4300220503667984,
+      "learning_rate": 1.6857546007664908e-05,
+      "loss": 3.3787,
+      "step": 1465
+    },
+    {
+      "epoch": 0.2820451156750517,
+      "grad_norm": 3.148138652632097,
+      "learning_rate": 1.6853008762218103e-05,
+      "loss": 3.3241,
+      "step": 1466
+    },
+    {
+      "epoch": 0.2822375066134385,
+      "grad_norm": 3.5636438767561023,
+      "learning_rate": 1.684846885514166e-05,
+      "loss": 3.287,
+      "step": 1467
+    },
+    {
+      "epoch": 0.2824298975518253,
+      "grad_norm": 3.3914150493062163,
+      "learning_rate": 1.6843926288198828e-05,
+      "loss": 3.1528,
+      "step": 1468
+    },
+    {
+      "epoch": 0.2826222884902121,
+      "grad_norm": 3.678973116188667,
+      "learning_rate": 1.683938106315389e-05,
+      "loss": 3.247,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2828146794285989,
+      "grad_norm": 3.6642751667179394,
+      "learning_rate": 1.683483318177216e-05,
+      "loss": 3.1977,
+      "step": 1470
+    },
+    {
+      "epoch": 0.2830070703669857,
+      "grad_norm": 3.4771028123882655,
+      "learning_rate": 1.6830282645819974e-05,
+      "loss": 3.5268,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2831994613053725,
+      "grad_norm": 3.4506266572844853,
+      "learning_rate": 1.6825729457064718e-05,
+      "loss": 3.0626,
+      "step": 1472
+    },
+    {
+      "epoch": 0.28339185224375935,
+      "grad_norm": 3.4497363496839184,
+      "learning_rate": 1.6821173617274793e-05,
+      "loss": 3.174,
+      "step": 1473
+    },
+    {
+      "epoch": 0.28358424318214615,
+      "grad_norm": 3.7558767549960597,
+      "learning_rate": 1.6816615128219635e-05,
+      "loss": 3.0126,
+      "step": 1474
+    },
+    {
+      "epoch": 0.28377663412053294,
+      "grad_norm": 3.5348944161515052,
+      "learning_rate": 1.681205399166971e-05,
+      "loss": 3.0757,
+      "step": 1475
+    },
+    {
+      "epoch": 0.28396902505891974,
+      "grad_norm": 3.5638750613002963,
+      "learning_rate": 1.6807490209396506e-05,
+      "loss": 3.1315,
+      "step": 1476
+    },
+    {
+      "epoch": 0.28416141599730654,
+      "grad_norm": 3.613333454263572,
+      "learning_rate": 1.6802923783172553e-05,
+      "loss": 3.09,
+      "step": 1477
+    },
+    {
+      "epoch": 0.28435380693569334,
+      "grad_norm": 3.8361734493933275,
+      "learning_rate": 1.679835471477139e-05,
+      "loss": 3.0553,
+      "step": 1478
+    },
+    {
+      "epoch": 0.28454619787408014,
+      "grad_norm": 2.5348796504533833,
+      "learning_rate": 1.6793783005967593e-05,
+      "loss": 2.8069,
+      "step": 1479
+    },
+    {
+      "epoch": 0.28473858881246694,
+      "grad_norm": 3.079950372653075,
+      "learning_rate": 1.678920865853676e-05,
+      "loss": 3.2574,
+      "step": 1480
+    },
+    {
+      "epoch": 0.28493097975085374,
+      "grad_norm": 3.5002085077181424,
+      "learning_rate": 1.678463167425552e-05,
+      "loss": 3.2824,
+      "step": 1481
+    },
+    {
+      "epoch": 0.28512337068924054,
+      "grad_norm": 3.6108405905788405,
+      "learning_rate": 1.6780052054901512e-05,
+      "loss": 3.3542,
+      "step": 1482
+    },
+    {
+      "epoch": 0.28531576162762734,
+      "grad_norm": 3.465830490670203,
+      "learning_rate": 1.6775469802253416e-05,
+      "loss": 3.0444,
+      "step": 1483
+    },
+    {
+      "epoch": 0.28550815256601414,
+      "grad_norm": 3.5736942760250248,
+      "learning_rate": 1.6770884918090923e-05,
+      "loss": 3.2906,
+      "step": 1484
+    },
+    {
+      "epoch": 0.28570054350440094,
+      "grad_norm": 3.4586966922320705,
+      "learning_rate": 1.6766297404194745e-05,
+      "loss": 3.0477,
+      "step": 1485
+    },
+    {
+      "epoch": 0.28589293444278774,
+      "grad_norm": 3.8370018908959764,
+      "learning_rate": 1.6761707262346624e-05,
+      "loss": 3.2151,
+      "step": 1486
+    },
+    {
+      "epoch": 0.28608532538117454,
+      "grad_norm": 3.631981622614642,
+      "learning_rate": 1.675711449432932e-05,
+      "loss": 2.7968,
+      "step": 1487
+    },
+    {
+      "epoch": 0.28627771631956134,
+      "grad_norm": 3.769821333169058,
+      "learning_rate": 1.6752519101926606e-05,
+      "loss": 3.2092,
+      "step": 1488
+    },
+    {
+      "epoch": 0.28647010725794814,
+      "grad_norm": 3.7246725044600137,
+      "learning_rate": 1.6747921086923284e-05,
+      "loss": 3.2297,
+      "step": 1489
+    },
+    {
+      "epoch": 0.28666249819633494,
+      "grad_norm": 3.715682238342402,
+      "learning_rate": 1.674332045110517e-05,
+      "loss": 3.2476,
+      "step": 1490
+    },
+    {
+      "epoch": 0.28685488913472174,
+      "grad_norm": 3.5072274261756498,
+      "learning_rate": 1.6738717196259092e-05,
+      "loss": 3.2212,
+      "step": 1491
+    },
+    {
+      "epoch": 0.28704728007310853,
+      "grad_norm": 3.783440536129484,
+      "learning_rate": 1.673411132417291e-05,
+      "loss": 3.129,
+      "step": 1492
+    },
+    {
+      "epoch": 0.28723967101149533,
+      "grad_norm": 3.6012837797407773,
+      "learning_rate": 1.672950283663548e-05,
+      "loss": 3.0885,
+      "step": 1493
+    },
+    {
+      "epoch": 0.28743206194988213,
+      "grad_norm": 3.76301911540152,
+      "learning_rate": 1.6724891735436697e-05,
+      "loss": 3.2713,
+      "step": 1494
+    },
+    {
+      "epoch": 0.287624452888269,
+      "grad_norm": 3.5696567412664475,
+      "learning_rate": 1.6720278022367453e-05,
+      "loss": 3.3325,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2878168438266558,
+      "grad_norm": 3.5813136613123797,
+      "learning_rate": 1.6715661699219664e-05,
+      "loss": 3.3595,
+      "step": 1496
+    },
+    {
+      "epoch": 0.2880092347650426,
+      "grad_norm": 4.075717518432014,
+      "learning_rate": 1.6711042767786257e-05,
+      "loss": 3.3768,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2882016257034294,
+      "grad_norm": 3.5007818673495494,
+      "learning_rate": 1.6706421229861168e-05,
+      "loss": 3.0994,
+      "step": 1498
+    },
+    {
+      "epoch": 0.2883940166418162,
+      "grad_norm": 3.3136344959816624,
+      "learning_rate": 1.6701797087239354e-05,
+      "loss": 3.204,
+      "step": 1499
+    },
+    {
+      "epoch": 0.288586407580203,
+      "grad_norm": 3.6006594050013385,
+      "learning_rate": 1.6697170341716772e-05,
+      "loss": 3.3451,
+      "step": 1500
+    },
+    {
+      "epoch": 0.2887787985185898,
+      "grad_norm": 3.3987376618121643,
+      "learning_rate": 1.6692540995090403e-05,
+      "loss": 3.2836,
+      "step": 1501
+    },
+    {
+      "epoch": 0.2889711894569766,
+      "grad_norm": 3.377214207359315,
+      "learning_rate": 1.668790904915823e-05,
+      "loss": 3.0013,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2891635803953634,
+      "grad_norm": 3.6408646344452067,
+      "learning_rate": 1.6683274505719248e-05,
+      "loss": 3.2211,
+      "step": 1503
+    },
+    {
+      "epoch": 0.2893559713337502,
+      "grad_norm": 2.5349521430986317,
+      "learning_rate": 1.6678637366573455e-05,
+      "loss": 2.5866,
+      "step": 1504
+    },
+    {
+      "epoch": 0.289548362272137,
+      "grad_norm": 3.7501420232684572,
+      "learning_rate": 1.667399763352187e-05,
+      "loss": 3.1728,
+      "step": 1505
+    },
+    {
+      "epoch": 0.2897407532105238,
+      "grad_norm": 3.4958932085324643,
+      "learning_rate": 1.666935530836651e-05,
+      "loss": 3.407,
+      "step": 1506
+    },
+    {
+      "epoch": 0.2899331441489106,
+      "grad_norm": 3.46859369955534,
+      "learning_rate": 1.6664710392910396e-05,
+      "loss": 3.1905,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2901255350872974,
+      "grad_norm": 3.285644248171616,
+      "learning_rate": 1.6660062888957564e-05,
+      "loss": 3.1575,
+      "step": 1508
+    },
+    {
+      "epoch": 0.2903179260256842,
+      "grad_norm": 3.5811541672774343,
+      "learning_rate": 1.665541279831305e-05,
+      "loss": 3.3792,
+      "step": 1509
+    },
+    {
+      "epoch": 0.290510316964071,
+      "grad_norm": 3.5234003264000977,
+      "learning_rate": 1.6650760122782898e-05,
+      "loss": 3.0352,
+      "step": 1510
+    },
+    {
+      "epoch": 0.2907027079024578,
+      "grad_norm": 3.683319305991437,
+      "learning_rate": 1.6646104864174147e-05,
+      "loss": 3.2174,
+      "step": 1511
+    },
+    {
+      "epoch": 0.2908950988408446,
+      "grad_norm": 4.098050266710883,
+      "learning_rate": 1.664144702429485e-05,
+      "loss": 3.5731,
+      "step": 1512
+    },
+    {
+      "epoch": 0.2910874897792314,
+      "grad_norm": 3.4611867396546954,
+      "learning_rate": 1.663678660495406e-05,
+      "loss": 3.3748,
+      "step": 1513
+    },
+    {
+      "epoch": 0.2912798807176182,
+      "grad_norm": 3.2923702430273942,
+      "learning_rate": 1.663212360796183e-05,
+      "loss": 3.4586,
+      "step": 1514
+    },
+    {
+      "epoch": 0.291472271656005,
+      "grad_norm": 3.8684812908195783,
+      "learning_rate": 1.662745803512921e-05,
+      "loss": 3.309,
+      "step": 1515
+    },
+    {
+      "epoch": 0.2916646625943918,
+      "grad_norm": 3.351274810157583,
+      "learning_rate": 1.662278988826826e-05,
+      "loss": 3.3531,
+      "step": 1516
+    },
+    {
+      "epoch": 0.29185705353277863,
+      "grad_norm": 3.491964284303135,
+      "learning_rate": 1.6618119169192027e-05,
+      "loss": 3.2207,
+      "step": 1517
+    },
+    {
+      "epoch": 0.29204944447116543,
+      "grad_norm": 3.6650399780181147,
+      "learning_rate": 1.661344587971457e-05,
+      "loss": 3.3292,
+      "step": 1518
+    },
+    {
+      "epoch": 0.29224183540955223,
+      "grad_norm": 3.5699930929962758,
+      "learning_rate": 1.6608770021650945e-05,
+      "loss": 3.3747,
+      "step": 1519
+    },
+    {
+      "epoch": 0.29243422634793903,
+      "grad_norm": 3.722749533070468,
+      "learning_rate": 1.6604091596817193e-05,
+      "loss": 3.1969,
+      "step": 1520
+    },
+    {
+      "epoch": 0.29262661728632583,
+      "grad_norm": 3.8230600079062556,
+      "learning_rate": 1.6599410607030363e-05,
+      "loss": 3.5812,
+      "step": 1521
+    },
+    {
+      "epoch": 0.29281900822471263,
+      "grad_norm": 3.593970298775221,
+      "learning_rate": 1.6594727054108498e-05,
+      "loss": 3.2358,
+      "step": 1522
+    },
+    {
+      "epoch": 0.29301139916309943,
+      "grad_norm": 3.600223245604381,
+      "learning_rate": 1.659004093987064e-05,
+      "loss": 3.1039,
+      "step": 1523
+    },
+    {
+      "epoch": 0.29320379010148623,
+      "grad_norm": 3.6534072197823177,
+      "learning_rate": 1.6585352266136814e-05,
+      "loss": 3.334,
+      "step": 1524
+    },
+    {
+      "epoch": 0.293396181039873,
+      "grad_norm": 3.6877377610878126,
+      "learning_rate": 1.6580661034728055e-05,
+      "loss": 3.2985,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2935885719782598,
+      "grad_norm": 3.3995809793857674,
+      "learning_rate": 1.6575967247466376e-05,
+      "loss": 3.0963,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2937809629166466,
+      "grad_norm": 3.5464141711731916,
+      "learning_rate": 1.657127090617479e-05,
+      "loss": 3.2102,
+      "step": 1527
+    },
+    {
+      "epoch": 0.2939733538550334,
+      "grad_norm": 3.3793738254073813,
+      "learning_rate": 1.656657201267731e-05,
+      "loss": 3.2207,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2941657447934202,
+      "grad_norm": 2.759897145413667,
+      "learning_rate": 1.6561870568798927e-05,
+      "loss": 2.7428,
+      "step": 1529
+    },
+    {
+      "epoch": 0.294358135731807,
+      "grad_norm": 3.910688158615534,
+      "learning_rate": 1.655716657636562e-05,
+      "loss": 3.4266,
+      "step": 1530
+    },
+    {
+      "epoch": 0.2945505266701938,
+      "grad_norm": 3.6695102567962268,
+      "learning_rate": 1.6552460037204382e-05,
+      "loss": 3.3542,
+      "step": 1531
+    },
+    {
+      "epoch": 0.2947429176085806,
+      "grad_norm": 3.422500341976175,
+      "learning_rate": 1.6547750953143168e-05,
+      "loss": 3.1783,
+      "step": 1532
+    },
+    {
+      "epoch": 0.2949353085469674,
+      "grad_norm": 3.422855415391366,
+      "learning_rate": 1.654303932601093e-05,
+      "loss": 3.2451,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2951276994853542,
+      "grad_norm": 3.585907292581935,
+      "learning_rate": 1.6538325157637614e-05,
+      "loss": 3.1472,
+      "step": 1534
+    },
+    {
+      "epoch": 0.295320090423741,
+      "grad_norm": 3.3657659465877465,
+      "learning_rate": 1.653360844985415e-05,
+      "loss": 2.9731,
+      "step": 1535
+    },
+    {
+      "epoch": 0.2955124813621278,
+      "grad_norm": 3.16815436391064,
+      "learning_rate": 1.652888920449245e-05,
+      "loss": 3.1336,
+      "step": 1536
+    },
+    {
+      "epoch": 0.2957048723005146,
+      "grad_norm": 3.445519151961484,
+      "learning_rate": 1.6524167423385414e-05,
+      "loss": 3.1117,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2958972632389014,
+      "grad_norm": 3.4526616461960455,
+      "learning_rate": 1.651944310836693e-05,
+      "loss": 3.1548,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2960896541772883,
+      "grad_norm": 3.460598545336078,
+      "learning_rate": 1.6514716261271866e-05,
+      "loss": 3.0471,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2962820451156751,
+      "grad_norm": 3.6489894837442516,
+      "learning_rate": 1.6509986883936073e-05,
+      "loss": 3.359,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2964744360540619,
+      "grad_norm": 3.9703556307382883,
+      "learning_rate": 1.650525497819639e-05,
+      "loss": 3.241,
+      "step": 1541
+    },
+    {
+      "epoch": 0.2966668269924487,
+      "grad_norm": 3.6259319390759748,
+      "learning_rate": 1.6500520545890634e-05,
+      "loss": 3.2174,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2968592179308355,
+      "grad_norm": 3.7027064771924407,
+      "learning_rate": 1.6495783588857605e-05,
+      "loss": 3.1942,
+      "step": 1543
+    },
+    {
+      "epoch": 0.2970516088692223,
+      "grad_norm": 3.6150431207263596,
+      "learning_rate": 1.649104410893708e-05,
+      "loss": 3.1231,
+      "step": 1544
+    },
+    {
+      "epoch": 0.29724399980760907,
+      "grad_norm": 3.405554710689078,
+      "learning_rate": 1.648630210796982e-05,
+      "loss": 2.9775,
+      "step": 1545
+    },
+    {
+      "epoch": 0.29743639074599587,
+      "grad_norm": 3.3483482260697697,
+      "learning_rate": 1.6481557587797562e-05,
+      "loss": 3.2917,
+      "step": 1546
+    },
+    {
+      "epoch": 0.29762878168438267,
+      "grad_norm": 3.406662653970037,
+      "learning_rate": 1.6476810550263023e-05,
+      "loss": 3.1576,
+      "step": 1547
+    },
+    {
+      "epoch": 0.29782117262276947,
+      "grad_norm": 2.455717760989008,
+      "learning_rate": 1.6472060997209898e-05,
+      "loss": 2.623,
+      "step": 1548
+    },
+    {
+      "epoch": 0.29801356356115627,
+      "grad_norm": 3.5461205909770968,
+      "learning_rate": 1.6467308930482863e-05,
+      "loss": 3.1259,
+      "step": 1549
+    },
+    {
+      "epoch": 0.29820595449954307,
+      "grad_norm": 3.490342732927331,
+      "learning_rate": 1.6462554351927558e-05,
+      "loss": 3.1928,
+      "step": 1550
+    },
+    {
+      "epoch": 0.29839834543792987,
+      "grad_norm": 3.8196669356677497,
+      "learning_rate": 1.6457797263390613e-05,
+      "loss": 2.9367,
+      "step": 1551
+    },
+    {
+      "epoch": 0.29859073637631667,
+      "grad_norm": 3.7338320189630294,
+      "learning_rate": 1.6453037666719624e-05,
+      "loss": 3.4402,
+      "step": 1552
+    },
+    {
+      "epoch": 0.29878312731470347,
+      "grad_norm": 3.2406070129783666,
+      "learning_rate": 1.6448275563763162e-05,
+      "loss": 3.1063,
+      "step": 1553
+    },
+    {
+      "epoch": 0.29897551825309027,
+      "grad_norm": 3.4261196362973196,
+      "learning_rate": 1.644351095637078e-05,
+      "loss": 3.1166,
+      "step": 1554
+    },
+    {
+      "epoch": 0.29916790919147707,
+      "grad_norm": 3.4975790249001264,
+      "learning_rate": 1.6438743846392987e-05,
+      "loss": 3.4015,
+      "step": 1555
+    },
+    {
+      "epoch": 0.29936030012986387,
+      "grad_norm": 3.4165978168898965,
+      "learning_rate": 1.6433974235681274e-05,
+      "loss": 2.8447,
+      "step": 1556
+    },
+    {
+      "epoch": 0.29955269106825066,
+      "grad_norm": 3.7657708922548316,
+      "learning_rate": 1.6429202126088112e-05,
+      "loss": 3.2315,
+      "step": 1557
+    },
+    {
+      "epoch": 0.29974508200663746,
+      "grad_norm": 2.577099846661282,
+      "learning_rate": 1.6424427519466925e-05,
+      "loss": 2.8141,
+      "step": 1558
+    },
+    {
+      "epoch": 0.29993747294502426,
+      "grad_norm": 3.771217095604037,
+      "learning_rate": 1.641965041767212e-05,
+      "loss": 3.1097,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3001298638834111,
+      "grad_norm": 3.4602478073306626,
+      "learning_rate": 1.6414870822559064e-05,
+      "loss": 2.8725,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3003222548217979,
+      "grad_norm": 3.3365287084312185,
+      "learning_rate": 1.6410088735984103e-05,
+      "loss": 3.5036,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3005146457601847,
+      "grad_norm": 3.6170954212472304,
+      "learning_rate": 1.6405304159804534e-05,
+      "loss": 3.1977,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3007070366985715,
+      "grad_norm": 3.6167747850118044,
+      "learning_rate": 1.6400517095878644e-05,
+      "loss": 3.1815,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3008994276369583,
+      "grad_norm": 2.4203085476856314,
+      "learning_rate": 1.6395727546065665e-05,
+      "loss": 2.8131,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3010918185753451,
+      "grad_norm": 4.014913457064844,
+      "learning_rate": 1.6390935512225806e-05,
+      "loss": 3.3048,
+      "step": 1565
+    },
+    {
+      "epoch": 0.3012842095137319,
+      "grad_norm": 3.4384656974390517,
+      "learning_rate": 1.6386140996220232e-05,
+      "loss": 3.2691,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3014766004521187,
+      "grad_norm": 3.5878357317862486,
+      "learning_rate": 1.6381343999911088e-05,
+      "loss": 3.2113,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3016689913905055,
+      "grad_norm": 3.4107664106535536,
+      "learning_rate": 1.6376544525161463e-05,
+      "loss": 3.071,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3018613823288923,
+      "grad_norm": 3.497621480174249,
+      "learning_rate": 1.6371742573835426e-05,
+      "loss": 3.3015,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3020537732672791,
+      "grad_norm": 3.3615493181656113,
+      "learning_rate": 1.636693814779799e-05,
+      "loss": 3.3163,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3022461642056659,
+      "grad_norm": 3.6324906631565512,
+      "learning_rate": 1.6362131248915145e-05,
+      "loss": 3.2186,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3024385551440527,
+      "grad_norm": 3.393371827116578,
+      "learning_rate": 1.6357321879053833e-05,
+      "loss": 3.1092,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3026309460824395,
+      "grad_norm": 3.813105108563652,
+      "learning_rate": 1.6352510040081962e-05,
+      "loss": 3.3657,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3028233370208263,
+      "grad_norm": 3.5726470097269516,
+      "learning_rate": 1.634769573386839e-05,
+      "loss": 3.192,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3030157279592131,
+      "grad_norm": 3.327724761925461,
+      "learning_rate": 1.634287896228294e-05,
+      "loss": 3.3229,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3032081188975999,
+      "grad_norm": 3.642639784028509,
+      "learning_rate": 1.6338059727196386e-05,
+      "loss": 3.4761,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3034005098359867,
+      "grad_norm": 3.4813775087684062,
+      "learning_rate": 1.6333238030480473e-05,
+      "loss": 3.2792,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3035929007743735,
+      "grad_norm": 3.243710284516949,
+      "learning_rate": 1.6328413874007884e-05,
+      "loss": 3.2487,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3037852917127603,
+      "grad_norm": 3.600633022460528,
+      "learning_rate": 1.6323587259652267e-05,
+      "loss": 3.1726,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3039776826511471,
+      "grad_norm": 3.5968686009502284,
+      "learning_rate": 1.6318758189288227e-05,
+      "loss": 3.1016,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3041700735895339,
+      "grad_norm": 2.4505064054643513,
+      "learning_rate": 1.6313926664791316e-05,
+      "loss": 2.7306,
+      "step": 1581
+    },
+    {
+      "epoch": 0.30436246452792076,
+      "grad_norm": 2.2863981448793815,
+      "learning_rate": 1.6309092688038047e-05,
+      "loss": 2.6367,
+      "step": 1582
+    },
+    {
+      "epoch": 0.30455485546630756,
+      "grad_norm": 3.4119549380555716,
+      "learning_rate": 1.6304256260905872e-05,
+      "loss": 3.236,
+      "step": 1583
+    },
+    {
+      "epoch": 0.30474724640469436,
+      "grad_norm": 3.5023296913629904,
+      "learning_rate": 1.6299417385273216e-05,
+      "loss": 3.1621,
+      "step": 1584
+    },
+    {
+      "epoch": 0.30493963734308116,
+      "grad_norm": 3.75707096233396,
+      "learning_rate": 1.629457606301943e-05,
+      "loss": 3.2209,
+      "step": 1585
+    },
+    {
+      "epoch": 0.30513202828146796,
+      "grad_norm": 3.6103850906809805,
+      "learning_rate": 1.6289732296024837e-05,
+      "loss": 3.289,
+      "step": 1586
+    },
+    {
+      "epoch": 0.30532441921985476,
+      "grad_norm": 3.422353091764222,
+      "learning_rate": 1.6284886086170697e-05,
+      "loss": 3.0634,
+      "step": 1587
+    },
+    {
+      "epoch": 0.30551681015824156,
+      "grad_norm": 3.977389380996052,
+      "learning_rate": 1.628003743533922e-05,
+      "loss": 3.4543,
+      "step": 1588
+    },
+    {
+      "epoch": 0.30570920109662836,
+      "grad_norm": 3.41238786742715,
+      "learning_rate": 1.6275186345413566e-05,
+      "loss": 3.1006,
+      "step": 1589
+    },
+    {
+      "epoch": 0.30590159203501516,
+      "grad_norm": 3.5673159746725833,
+      "learning_rate": 1.627033281827785e-05,
+      "loss": 3.3645,
+      "step": 1590
+    },
+    {
+      "epoch": 0.30609398297340196,
+      "grad_norm": 2.5745676906536086,
+      "learning_rate": 1.6265476855817116e-05,
+      "loss": 2.609,
+      "step": 1591
+    },
+    {
+      "epoch": 0.30628637391178876,
+      "grad_norm": 3.3811598366071234,
+      "learning_rate": 1.6260618459917366e-05,
+      "loss": 3.3156,
+      "step": 1592
+    },
+    {
+      "epoch": 0.30647876485017556,
+      "grad_norm": 3.8754185058262056,
+      "learning_rate": 1.6255757632465553e-05,
+      "loss": 3.4196,
+      "step": 1593
+    },
+    {
+      "epoch": 0.30667115578856236,
+      "grad_norm": 3.5654844695911243,
+      "learning_rate": 1.625089437534956e-05,
+      "loss": 3.1621,
+      "step": 1594
+    },
+    {
+      "epoch": 0.30686354672694915,
+      "grad_norm": 3.3084935598379297,
+      "learning_rate": 1.624602869045822e-05,
+      "loss": 2.9735,
+      "step": 1595
+    },
+    {
+      "epoch": 0.30705593766533595,
+      "grad_norm": 3.2787209211952426,
+      "learning_rate": 1.624116057968131e-05,
+      "loss": 3.4469,
+      "step": 1596
+    },
+    {
+      "epoch": 0.30724832860372275,
+      "grad_norm": 3.66488207641828,
+      "learning_rate": 1.6236290044909543e-05,
+      "loss": 3.3515,
+      "step": 1597
+    },
+    {
+      "epoch": 0.30744071954210955,
+      "grad_norm": 3.272003392335416,
+      "learning_rate": 1.6231417088034585e-05,
+      "loss": 3.057,
+      "step": 1598
+    },
+    {
+      "epoch": 0.30763311048049635,
+      "grad_norm": 3.368137459599917,
+      "learning_rate": 1.622654171094904e-05,
+      "loss": 3.1089,
+      "step": 1599
+    },
+    {
+      "epoch": 0.30782550141888315,
+      "grad_norm": 3.6444366985871,
+      "learning_rate": 1.6221663915546437e-05,
+      "loss": 3.259,
+      "step": 1600
+    },
+    {
+      "epoch": 0.30801789235726995,
+      "grad_norm": 3.666295298497803,
+      "learning_rate": 1.6216783703721265e-05,
+      "loss": 3.4235,
+      "step": 1601
+    },
+    {
+      "epoch": 0.30821028329565675,
+      "grad_norm": 3.352977235004946,
+      "learning_rate": 1.6211901077368937e-05,
+      "loss": 2.9344,
+      "step": 1602
+    },
+    {
+      "epoch": 0.30840267423404355,
+      "grad_norm": 3.420704855238811,
+      "learning_rate": 1.620701603838581e-05,
+      "loss": 3.0569,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3085950651724304,
+      "grad_norm": 3.715068999420625,
+      "learning_rate": 1.6202128588669177e-05,
+      "loss": 3.1849,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3087874561108172,
+      "grad_norm": 3.29476921760618,
+      "learning_rate": 1.619723873011727e-05,
+      "loss": 3.2379,
+      "step": 1605
+    },
+    {
+      "epoch": 0.308979847049204,
+      "grad_norm": 3.502526036169402,
+      "learning_rate": 1.6192346464629247e-05,
+      "loss": 3.058,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3091722379875908,
+      "grad_norm": 3.4124847266048324,
+      "learning_rate": 1.6187451794105212e-05,
+      "loss": 3.1995,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3093646289259776,
+      "grad_norm": 3.5314596752570857,
+      "learning_rate": 1.61825547204462e-05,
+      "loss": 3.1404,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3095570198643644,
+      "grad_norm": 3.624585641095926,
+      "learning_rate": 1.6177655245554177e-05,
+      "loss": 3.1591,
+      "step": 1609
+    },
+    {
+      "epoch": 0.3097494108027512,
+      "grad_norm": 3.623885551656112,
+      "learning_rate": 1.617275337133204e-05,
+      "loss": 3.3444,
+      "step": 1610
+    },
+    {
+      "epoch": 0.309941801741138,
+      "grad_norm": 3.6190160482387306,
+      "learning_rate": 1.6167849099683623e-05,
+      "loss": 3.2305,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3101341926795248,
+      "grad_norm": 2.7247365822276106,
+      "learning_rate": 1.6162942432513687e-05,
+      "loss": 2.7932,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3103265836179116,
+      "grad_norm": 2.4225293723343357,
+      "learning_rate": 1.6158033371727924e-05,
+      "loss": 2.6765,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3105189745562984,
+      "grad_norm": 3.484484402610532,
+      "learning_rate": 1.6153121919232962e-05,
+      "loss": 3.0397,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3107113654946852,
+      "grad_norm": 4.252740445386187,
+      "learning_rate": 1.614820807693635e-05,
+      "loss": 2.9333,
+      "step": 1615
+    },
+    {
+      "epoch": 0.310903756433072,
+      "grad_norm": 3.7178970124182467,
+      "learning_rate": 1.6143291846746563e-05,
+      "loss": 3.3653,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3110961473714588,
+      "grad_norm": 3.3344099989162626,
+      "learning_rate": 1.613837323057301e-05,
+      "loss": 3.2459,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3112885383098456,
+      "grad_norm": 3.508443697971295,
+      "learning_rate": 1.6133452230326035e-05,
+      "loss": 3.167,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3114809292482324,
+      "grad_norm": 3.494642887178929,
+      "learning_rate": 1.6128528847916883e-05,
+      "loss": 3.3595,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3116733201866192,
+      "grad_norm": 3.3964624125572653,
+      "learning_rate": 1.6123603085257746e-05,
+      "loss": 3.465,
+      "step": 1620
+    },
+    {
+      "epoch": 0.311865711125006,
+      "grad_norm": 3.6775075778431185,
+      "learning_rate": 1.6118674944261732e-05,
+      "loss": 3.2971,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3120581020633928,
+      "grad_norm": 3.555262120538171,
+      "learning_rate": 1.6113744426842882e-05,
+      "loss": 2.9794,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3122504930017796,
+      "grad_norm": 3.6780540454955473,
+      "learning_rate": 1.6108811534916137e-05,
+      "loss": 3.2681,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3124428839401664,
+      "grad_norm": 3.523731401866378,
+      "learning_rate": 1.6103876270397387e-05,
+      "loss": 3.4475,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3126352748785532,
+      "grad_norm": 3.288935231459593,
+      "learning_rate": 1.609893863520343e-05,
+      "loss": 3.0979,
+      "step": 1625
+    },
+    {
+      "epoch": 0.31282766581694005,
+      "grad_norm": 3.0867527047469676,
+      "learning_rate": 1.609399863125198e-05,
+      "loss": 2.5588,
+      "step": 1626
+    },
+    {
+      "epoch": 0.31302005675532685,
+      "grad_norm": 3.6866101424761992,
+      "learning_rate": 1.6089056260461687e-05,
+      "loss": 3.301,
+      "step": 1627
+    },
+    {
+      "epoch": 0.31321244769371365,
+      "grad_norm": 3.4293583466679696,
+      "learning_rate": 1.6084111524752107e-05,
+      "loss": 3.1011,
+      "step": 1628
+    },
+    {
+      "epoch": 0.31340483863210045,
+      "grad_norm": 3.529785401187913,
+      "learning_rate": 1.607916442604372e-05,
+      "loss": 3.312,
+      "step": 1629
+    },
+    {
+      "epoch": 0.31359722957048725,
+      "grad_norm": 3.5482246922043204,
+      "learning_rate": 1.6074214966257914e-05,
+      "loss": 3.0323,
+      "step": 1630
+    },
+    {
+      "epoch": 0.31378962050887405,
+      "grad_norm": 3.7799796078838956,
+      "learning_rate": 1.6069263147317015e-05,
+      "loss": 3.6017,
+      "step": 1631
+    },
+    {
+      "epoch": 0.31398201144726084,
+      "grad_norm": 2.4410163101093127,
+      "learning_rate": 1.6064308971144236e-05,
+      "loss": 2.7344,
+      "step": 1632
+    },
+    {
+      "epoch": 0.31417440238564764,
+      "grad_norm": 3.554546483551753,
+      "learning_rate": 1.605935243966374e-05,
+      "loss": 3.3262,
+      "step": 1633
+    },
+    {
+      "epoch": 0.31436679332403444,
+      "grad_norm": 3.599858480446572,
+      "learning_rate": 1.6054393554800574e-05,
+      "loss": 3.2955,
+      "step": 1634
+    },
+    {
+      "epoch": 0.31455918426242124,
+      "grad_norm": 3.3049113027678967,
+      "learning_rate": 1.604943231848072e-05,
+      "loss": 3.3644,
+      "step": 1635
+    },
+    {
+      "epoch": 0.31475157520080804,
+      "grad_norm": 3.5167622611198315,
+      "learning_rate": 1.604446873263106e-05,
+      "loss": 2.9433,
+      "step": 1636
+    },
+    {
+      "epoch": 0.31494396613919484,
+      "grad_norm": 3.311558060553367,
+      "learning_rate": 1.6039502799179394e-05,
+      "loss": 3.3628,
+      "step": 1637
+    },
+    {
+      "epoch": 0.31513635707758164,
+      "grad_norm": 3.597718594817423,
+      "learning_rate": 1.6034534520054435e-05,
+      "loss": 3.2363,
+      "step": 1638
+    },
+    {
+      "epoch": 0.31532874801596844,
+      "grad_norm": 3.4747917330811022,
+      "learning_rate": 1.60295638971858e-05,
+      "loss": 2.9847,
+      "step": 1639
+    },
+    {
+      "epoch": 0.31552113895435524,
+      "grad_norm": 3.473249275901693,
+      "learning_rate": 1.602459093250403e-05,
+      "loss": 3.1448,
+      "step": 1640
+    },
+    {
+      "epoch": 0.31571352989274204,
+      "grad_norm": 3.680170141796251,
+      "learning_rate": 1.601961562794056e-05,
+      "loss": 3.2879,
+      "step": 1641
+    },
+    {
+      "epoch": 0.31590592083112884,
+      "grad_norm": 3.3797412243799276,
+      "learning_rate": 1.601463798542775e-05,
+      "loss": 3.3357,
+      "step": 1642
+    },
+    {
+      "epoch": 0.31609831176951564,
+      "grad_norm": 3.199359973980341,
+      "learning_rate": 1.6009658006898848e-05,
+      "loss": 3.2402,
+      "step": 1643
+    },
+    {
+      "epoch": 0.31629070270790244,
+      "grad_norm": 3.608039250946129,
+      "learning_rate": 1.600467569428803e-05,
+      "loss": 3.0981,
+      "step": 1644
+    },
+    {
+      "epoch": 0.31648309364628924,
+      "grad_norm": 3.4131812303738736,
+      "learning_rate": 1.599969104953036e-05,
+      "loss": 3.228,
+      "step": 1645
+    },
+    {
+      "epoch": 0.31667548458467604,
+      "grad_norm": 3.679755567762259,
+      "learning_rate": 1.599470407456182e-05,
+      "loss": 3.0673,
+      "step": 1646
+    },
+    {
+      "epoch": 0.31686787552306284,
+      "grad_norm": 3.4771881109723526,
+      "learning_rate": 1.5989714771319297e-05,
+      "loss": 3.41,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3170602664614497,
+      "grad_norm": 3.5327187329975485,
+      "learning_rate": 1.5984723141740578e-05,
+      "loss": 3.2996,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3172526573998365,
+      "grad_norm": 3.4840679536787906,
+      "learning_rate": 1.597972918776435e-05,
+      "loss": 3.2293,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3174450483382233,
+      "grad_norm": 3.5354875411858053,
+      "learning_rate": 1.5974732911330208e-05,
+      "loss": 3.2138,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3176374392766101,
+      "grad_norm": 3.3517725423587916,
+      "learning_rate": 1.5969734314378654e-05,
+      "loss": 3.1473,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3178298302149969,
+      "grad_norm": 3.4595348185858072,
+      "learning_rate": 1.5964733398851078e-05,
+      "loss": 3.1916,
+      "step": 1652
+    },
+    {
+      "epoch": 0.3180222211533837,
+      "grad_norm": 3.714864951703756,
+      "learning_rate": 1.5959730166689783e-05,
+      "loss": 3.4097,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3182146120917705,
+      "grad_norm": 3.788050650902863,
+      "learning_rate": 1.5954724619837966e-05,
+      "loss": 3.3276,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3184070030301573,
+      "grad_norm": 3.2969262243799347,
+      "learning_rate": 1.5949716760239722e-05,
+      "loss": 3.0822,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3185993939685441,
+      "grad_norm": 3.467421011238162,
+      "learning_rate": 1.5944706589840046e-05,
+      "loss": 3.3691,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3187917849069309,
+      "grad_norm": 3.622731455068241,
+      "learning_rate": 1.5939694110584833e-05,
+      "loss": 3.1477,
+      "step": 1657
+    },
+    {
+      "epoch": 0.3189841758453177,
+      "grad_norm": 3.699378346434176,
+      "learning_rate": 1.593467932442087e-05,
+      "loss": 3.0236,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3191765667837045,
+      "grad_norm": 3.510381860768288,
+      "learning_rate": 1.5929662233295846e-05,
+      "loss": 3.3461,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3193689577220913,
+      "grad_norm": 3.4569310617969142,
+      "learning_rate": 1.5924642839158334e-05,
+      "loss": 3.3728,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3195613486604781,
+      "grad_norm": 3.510761737368162,
+      "learning_rate": 1.591962114395781e-05,
+      "loss": 3.3445,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3197537395988649,
+      "grad_norm": 3.740750443095011,
+      "learning_rate": 1.5914597149644654e-05,
+      "loss": 3.307,
+      "step": 1662
+    },
+    {
+      "epoch": 0.3199461305372517,
+      "grad_norm": 3.2465979116165524,
+      "learning_rate": 1.5909570858170115e-05,
+      "loss": 3.1761,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3201385214756385,
+      "grad_norm": 3.4826461554724455,
+      "learning_rate": 1.5904542271486346e-05,
+      "loss": 3.0003,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3203309124140253,
+      "grad_norm": 3.4353280019617474,
+      "learning_rate": 1.5899511391546403e-05,
+      "loss": 3.6157,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3205233033524121,
+      "grad_norm": 3.3725042894826536,
+      "learning_rate": 1.5894478220304215e-05,
+      "loss": 3.2305,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3207156942907989,
+      "grad_norm": 3.506374281197134,
+      "learning_rate": 1.5889442759714603e-05,
+      "loss": 3.1622,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3209080852291857,
+      "grad_norm": 3.348339679516042,
+      "learning_rate": 1.5884405011733294e-05,
+      "loss": 3.0561,
+      "step": 1668
+    },
+    {
+      "epoch": 0.32110047616757253,
+      "grad_norm": 3.390736324070696,
+      "learning_rate": 1.587936497831688e-05,
+      "loss": 3.0874,
+      "step": 1669
+    },
+    {
+      "epoch": 0.32129286710595933,
+      "grad_norm": 3.2430651401709363,
+      "learning_rate": 1.5874322661422856e-05,
+      "loss": 2.9607,
+      "step": 1670
+    },
+    {
+      "epoch": 0.32148525804434613,
+      "grad_norm": 3.720441522332533,
+      "learning_rate": 1.5869278063009602e-05,
+      "loss": 3.2309,
+      "step": 1671
+    },
+    {
+      "epoch": 0.32167764898273293,
+      "grad_norm": 2.6180124927404913,
+      "learning_rate": 1.586423118503638e-05,
+      "loss": 2.7382,
+      "step": 1672
+    },
+    {
+      "epoch": 0.32187003992111973,
+      "grad_norm": 3.3693525104535382,
+      "learning_rate": 1.585918202946334e-05,
+      "loss": 3.3706,
+      "step": 1673
+    },
+    {
+      "epoch": 0.32206243085950653,
+      "grad_norm": 3.5478069124806084,
+      "learning_rate": 1.5854130598251514e-05,
+      "loss": 3.1662,
+      "step": 1674
+    },
+    {
+      "epoch": 0.32225482179789333,
+      "grad_norm": 3.480891903234074,
+      "learning_rate": 1.5849076893362822e-05,
+      "loss": 3.1238,
+      "step": 1675
+    },
+    {
+      "epoch": 0.32244721273628013,
+      "grad_norm": 3.8089731369358106,
+      "learning_rate": 1.584402091676006e-05,
+      "loss": 3.2648,
+      "step": 1676
+    },
+    {
+      "epoch": 0.32263960367466693,
+      "grad_norm": 3.699687502603764,
+      "learning_rate": 1.5838962670406918e-05,
+      "loss": 3.1309,
+      "step": 1677
+    },
+    {
+      "epoch": 0.32283199461305373,
+      "grad_norm": 3.4032051822159195,
+      "learning_rate": 1.5833902156267956e-05,
+      "loss": 3.3346,
+      "step": 1678
+    },
+    {
+      "epoch": 0.32302438555144053,
+      "grad_norm": 3.57391689930954,
+      "learning_rate": 1.582883937630862e-05,
+      "loss": 3.105,
+      "step": 1679
+    },
+    {
+      "epoch": 0.32321677648982733,
+      "grad_norm": 3.4979591728765858,
+      "learning_rate": 1.5823774332495236e-05,
+      "loss": 3.1294,
+      "step": 1680
+    },
+    {
+      "epoch": 0.32340916742821413,
+      "grad_norm": 3.525295590882185,
+      "learning_rate": 1.581870702679501e-05,
+      "loss": 2.8674,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3236015583666009,
+      "grad_norm": 3.5559873765452243,
+      "learning_rate": 1.581363746117602e-05,
+      "loss": 3.3815,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3237939493049877,
+      "grad_norm": 3.793172905017284,
+      "learning_rate": 1.580856563760724e-05,
+      "loss": 3.3441,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3239863402433745,
+      "grad_norm": 3.4528080374179217,
+      "learning_rate": 1.5803491558058486e-05,
+      "loss": 3.2719,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3241787311817613,
+      "grad_norm": 3.583961179699517,
+      "learning_rate": 1.579841522450049e-05,
+      "loss": 3.2832,
+      "step": 1685
+    },
+    {
+      "epoch": 0.3243711221201481,
+      "grad_norm": 3.6639436192557664,
+      "learning_rate": 1.5793336638904838e-05,
+      "loss": 3.0155,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3245635130585349,
+      "grad_norm": 3.494051558600203,
+      "learning_rate": 1.578825580324399e-05,
+      "loss": 3.3427,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3247559039969217,
+      "grad_norm": 3.5885166298179247,
+      "learning_rate": 1.5783172719491288e-05,
+      "loss": 3.5095,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3249482949353085,
+      "grad_norm": 3.6590793892276876,
+      "learning_rate": 1.577808738962094e-05,
+      "loss": 3.2449,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3251406858736953,
+      "grad_norm": 3.3865783824839224,
+      "learning_rate": 1.577299981560803e-05,
+      "loss": 3.1279,
+      "step": 1690
+    },
+    {
+      "epoch": 0.3253330768120822,
+      "grad_norm": 4.003652002964614,
+      "learning_rate": 1.5767909999428513e-05,
+      "loss": 3.2396,
+      "step": 1691
+    },
+    {
+      "epoch": 0.325525467750469,
+      "grad_norm": 3.4557061944810603,
+      "learning_rate": 1.576281794305922e-05,
+      "loss": 3.3547,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3257178586888558,
+      "grad_norm": 3.790853951078264,
+      "learning_rate": 1.575772364847784e-05,
+      "loss": 3.3894,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3259102496272426,
+      "grad_norm": 3.7950932320488575,
+      "learning_rate": 1.575262711766294e-05,
+      "loss": 3.1314,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3261026405656294,
+      "grad_norm": 3.441893883850741,
+      "learning_rate": 1.5747528352593956e-05,
+      "loss": 3.2828,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3262950315040162,
+      "grad_norm": 3.5035283280301184,
+      "learning_rate": 1.574242735525119e-05,
+      "loss": 3.2214,
+      "step": 1696
+    },
+    {
+      "epoch": 0.326487422442403,
+      "grad_norm": 3.5080695361874215,
+      "learning_rate": 1.5737324127615808e-05,
+      "loss": 3.0676,
+      "step": 1697
+    },
+    {
+      "epoch": 0.3266798133807898,
+      "grad_norm": 3.5100365841820427,
+      "learning_rate": 1.5732218671669847e-05,
+      "loss": 3.0922,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3268722043191766,
+      "grad_norm": 3.6597037588189627,
+      "learning_rate": 1.5727110989396205e-05,
+      "loss": 3.2654,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3270645952575634,
+      "grad_norm": 4.124793032461809,
+      "learning_rate": 1.5722001082778645e-05,
+      "loss": 3.2882,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3272569861959502,
+      "grad_norm": 3.641571658453913,
+      "learning_rate": 1.5716888953801805e-05,
+      "loss": 3.0291,
+      "step": 1701
+    },
+    {
+      "epoch": 0.32744937713433697,
+      "grad_norm": 3.6318253414246415,
+      "learning_rate": 1.5711774604451168e-05,
+      "loss": 3.0699,
+      "step": 1702
+    },
+    {
+      "epoch": 0.32764176807272377,
+      "grad_norm": 3.405925793710819,
+      "learning_rate": 1.5706658036713093e-05,
+      "loss": 2.9523,
+      "step": 1703
+    },
+    {
+      "epoch": 0.32783415901111057,
+      "grad_norm": 3.673214756812535,
+      "learning_rate": 1.5701539252574795e-05,
+      "loss": 3.0987,
+      "step": 1704
+    },
+    {
+      "epoch": 0.32802654994949737,
+      "grad_norm": 3.4841120436116535,
+      "learning_rate": 1.5696418254024344e-05,
+      "loss": 3.0718,
+      "step": 1705
+    },
+    {
+      "epoch": 0.32821894088788417,
+      "grad_norm": 3.793820401400769,
+      "learning_rate": 1.569129504305069e-05,
+      "loss": 2.8576,
+      "step": 1706
+    },
+    {
+      "epoch": 0.32841133182627097,
+      "grad_norm": 3.709681506248668,
+      "learning_rate": 1.568616962164362e-05,
+      "loss": 3.2827,
+      "step": 1707
+    },
+    {
+      "epoch": 0.32860372276465777,
+      "grad_norm": 3.8122565207097083,
+      "learning_rate": 1.5681041991793788e-05,
+      "loss": 3.4336,
+      "step": 1708
+    },
+    {
+      "epoch": 0.32879611370304457,
+      "grad_norm": 3.5178140228820447,
+      "learning_rate": 1.567591215549271e-05,
+      "loss": 3.1817,
+      "step": 1709
+    },
+    {
+      "epoch": 0.32898850464143137,
+      "grad_norm": 3.667342105332199,
+      "learning_rate": 1.567078011473276e-05,
+      "loss": 3.3028,
+      "step": 1710
+    },
+    {
+      "epoch": 0.32918089557981817,
+      "grad_norm": 2.9316747315078167,
+      "learning_rate": 1.5665645871507152e-05,
+      "loss": 2.6734,
+      "step": 1711
+    },
+    {
+      "epoch": 0.32937328651820497,
+      "grad_norm": 3.442844283683188,
+      "learning_rate": 1.5660509427809973e-05,
+      "loss": 2.9597,
+      "step": 1712
+    },
+    {
+      "epoch": 0.3295656774565918,
+      "grad_norm": 3.6166769187056405,
+      "learning_rate": 1.565537078563616e-05,
+      "loss": 3.4108,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3297580683949786,
+      "grad_norm": 3.793945565433624,
+      "learning_rate": 1.56502299469815e-05,
+      "loss": 3.2599,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3299504593333654,
+      "grad_norm": 3.522935803805033,
+      "learning_rate": 1.564508691384264e-05,
+      "loss": 3.0262,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3301428502717522,
+      "grad_norm": 2.642183698188439,
+      "learning_rate": 1.5639941688217063e-05,
+      "loss": 2.7249,
+      "step": 1716
+    },
+    {
+      "epoch": 0.330335241210139,
+      "grad_norm": 3.4683111660302104,
+      "learning_rate": 1.5634794272103126e-05,
+      "loss": 2.8125,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3305276321485258,
+      "grad_norm": 3.788855487203169,
+      "learning_rate": 1.562964466750003e-05,
+      "loss": 3.0293,
+      "step": 1718
+    },
+    {
+      "epoch": 0.3307200230869126,
+      "grad_norm": 3.6309916724998628,
+      "learning_rate": 1.562449287640781e-05,
+      "loss": 3.3696,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3309124140252994,
+      "grad_norm": 3.183225631515213,
+      "learning_rate": 1.5619338900827368e-05,
+      "loss": 3.101,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3311048049636862,
+      "grad_norm": 3.417716809196949,
+      "learning_rate": 1.5614182742760448e-05,
+      "loss": 3.0311,
+      "step": 1721
+    },
+    {
+      "epoch": 0.331297195902073,
+      "grad_norm": 3.550302464197607,
+      "learning_rate": 1.5609024404209643e-05,
+      "loss": 3.3093,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3314895868404598,
+      "grad_norm": 3.38213145173183,
+      "learning_rate": 1.5603863887178393e-05,
+      "loss": 3.2202,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3316819777788466,
+      "grad_norm": 3.3460960775349826,
+      "learning_rate": 1.5598701193670983e-05,
+      "loss": 3.5069,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3318743687172334,
+      "grad_norm": 3.9277274580776926,
+      "learning_rate": 1.559353632569254e-05,
+      "loss": 3.4338,
+      "step": 1725
+    },
+    {
+      "epoch": 0.3320667596556202,
+      "grad_norm": 3.5386711184087676,
+      "learning_rate": 1.5588369285249048e-05,
+      "loss": 3.1835,
+      "step": 1726
+    },
+    {
+      "epoch": 0.332259150594007,
+      "grad_norm": 3.2339005627639152,
+      "learning_rate": 1.5583200074347318e-05,
+      "loss": 3.0276,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3324515415323938,
+      "grad_norm": 2.6556897061543583,
+      "learning_rate": 1.557802869499501e-05,
+      "loss": 2.7875,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3326439324707806,
+      "grad_norm": 3.5465184634410294,
+      "learning_rate": 1.5572855149200637e-05,
+      "loss": 3.2053,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3328363234091674,
+      "grad_norm": 3.4062357721213052,
+      "learning_rate": 1.5567679438973543e-05,
+      "loss": 2.9728,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3330287143475542,
+      "grad_norm": 3.614579725883572,
+      "learning_rate": 1.5562501566323906e-05,
+      "loss": 3.2696,
+      "step": 1731
+    },
+    {
+      "epoch": 0.333221105285941,
+      "grad_norm": 3.496289152295159,
+      "learning_rate": 1.555732153326276e-05,
+      "loss": 2.8414,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3334134962243278,
+      "grad_norm": 3.5269973695288566,
+      "learning_rate": 1.5552139341801965e-05,
+      "loss": 3.4515,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3336058871627146,
+      "grad_norm": 3.7764850984312233,
+      "learning_rate": 1.554695499395423e-05,
+      "loss": 3.2569,
+      "step": 1734
+    },
+    {
+      "epoch": 0.33379827810110146,
+      "grad_norm": 3.4507740301852787,
+      "learning_rate": 1.5541768491733092e-05,
+      "loss": 2.7441,
+      "step": 1735
+    },
+    {
+      "epoch": 0.33399066903948826,
+      "grad_norm": 3.3491723552753823,
+      "learning_rate": 1.5536579837152927e-05,
+      "loss": 2.8991,
+      "step": 1736
+    },
+    {
+      "epoch": 0.33418305997787506,
+      "grad_norm": 3.2977336066920793,
+      "learning_rate": 1.5531389032228955e-05,
+      "loss": 3.1864,
+      "step": 1737
+    },
+    {
+      "epoch": 0.33437545091626186,
+      "grad_norm": 3.4915178014754926,
+      "learning_rate": 1.552619607897722e-05,
+      "loss": 2.9883,
+      "step": 1738
+    },
+    {
+      "epoch": 0.33456784185464866,
+      "grad_norm": 3.5311155364310167,
+      "learning_rate": 1.55210009794146e-05,
+      "loss": 3.2983,
+      "step": 1739
+    },
+    {
+      "epoch": 0.33476023279303546,
+      "grad_norm": 3.7166286444311574,
+      "learning_rate": 1.5515803735558827e-05,
+      "loss": 3.4545,
+      "step": 1740
+    },
+    {
+      "epoch": 0.33495262373142226,
+      "grad_norm": 3.6197878603278384,
+      "learning_rate": 1.5510604349428438e-05,
+      "loss": 3.1824,
+      "step": 1741
+    },
+    {
+      "epoch": 0.33514501466980906,
+      "grad_norm": 3.4962871569692364,
+      "learning_rate": 1.550540282304282e-05,
+      "loss": 3.2304,
+      "step": 1742
+    },
+    {
+      "epoch": 0.33533740560819586,
+      "grad_norm": 3.846159758362262,
+      "learning_rate": 1.550019915842218e-05,
+      "loss": 3.3375,
+      "step": 1743
+    },
+    {
+      "epoch": 0.33552979654658266,
+      "grad_norm": 3.3231079148434217,
+      "learning_rate": 1.549499335758757e-05,
+      "loss": 2.989,
+      "step": 1744
+    },
+    {
+      "epoch": 0.33572218748496946,
+      "grad_norm": 3.1853729335383534,
+      "learning_rate": 1.548978542256086e-05,
+      "loss": 3.3125,
+      "step": 1745
+    },
+    {
+      "epoch": 0.33591457842335626,
+      "grad_norm": 3.5136607373064725,
+      "learning_rate": 1.5484575355364744e-05,
+      "loss": 3.2155,
+      "step": 1746
+    },
+    {
+      "epoch": 0.33610696936174306,
+      "grad_norm": 3.579782409306136,
+      "learning_rate": 1.5479363158022763e-05,
+      "loss": 3.4357,
+      "step": 1747
+    },
+    {
+      "epoch": 0.33629936030012986,
+      "grad_norm": 3.447036283266094,
+      "learning_rate": 1.547414883255927e-05,
+      "loss": 3.0528,
+      "step": 1748
+    },
+    {
+      "epoch": 0.33649175123851666,
+      "grad_norm": 3.5889338384106053,
+      "learning_rate": 1.546893238099945e-05,
+      "loss": 3.1768,
+      "step": 1749
+    },
+    {
+      "epoch": 0.33668414217690346,
+      "grad_norm": 3.526841971139054,
+      "learning_rate": 1.5463713805369312e-05,
+      "loss": 3.3668,
+      "step": 1750
+    },
+    {
+      "epoch": 0.33687653311529026,
+      "grad_norm": 4.176358713866971,
+      "learning_rate": 1.5458493107695688e-05,
+      "loss": 3.3402,
+      "step": 1751
+    },
+    {
+      "epoch": 0.33706892405367705,
+      "grad_norm": 3.6392429551536054,
+      "learning_rate": 1.5453270290006237e-05,
+      "loss": 3.3093,
+      "step": 1752
+    },
+    {
+      "epoch": 0.33726131499206385,
+      "grad_norm": 3.190706469705008,
+      "learning_rate": 1.544804535432945e-05,
+      "loss": 3.0043,
+      "step": 1753
+    },
+    {
+      "epoch": 0.33745370593045065,
+      "grad_norm": 3.7002467437611597,
+      "learning_rate": 1.544281830269462e-05,
+      "loss": 3.1696,
+      "step": 1754
+    },
+    {
+      "epoch": 0.33764609686883745,
+      "grad_norm": 3.692259524495714,
+      "learning_rate": 1.5437589137131882e-05,
+      "loss": 3.1232,
+      "step": 1755
+    },
+    {
+      "epoch": 0.33783848780722425,
+      "grad_norm": 3.65162315549488,
+      "learning_rate": 1.5432357859672177e-05,
+      "loss": 3.403,
+      "step": 1756
+    },
+    {
+      "epoch": 0.3380308787456111,
+      "grad_norm": 3.44019616725529,
+      "learning_rate": 1.542712447234728e-05,
+      "loss": 3.2109,
+      "step": 1757
+    },
+    {
+      "epoch": 0.3382232696839979,
+      "grad_norm": 3.491365222427023,
+      "learning_rate": 1.542188897718977e-05,
+      "loss": 3.0501,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3384156606223847,
+      "grad_norm": 3.6700758673334155,
+      "learning_rate": 1.5416651376233062e-05,
+      "loss": 3.2924,
+      "step": 1759
+    },
+    {
+      "epoch": 0.3386080515607715,
+      "grad_norm": 3.4490365736748148,
+      "learning_rate": 1.5411411671511376e-05,
+      "loss": 3.2067,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3388004424991583,
+      "grad_norm": 3.715711212575054,
+      "learning_rate": 1.5406169865059747e-05,
+      "loss": 3.3599,
+      "step": 1761
+    },
+    {
+      "epoch": 0.3389928334375451,
+      "grad_norm": 3.7503660109286825,
+      "learning_rate": 1.5400925958914045e-05,
+      "loss": 3.3359,
+      "step": 1762
+    },
+    {
+      "epoch": 0.3391852243759319,
+      "grad_norm": 3.7270799462042685,
+      "learning_rate": 1.5395679955110927e-05,
+      "loss": 2.9293,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3393776153143187,
+      "grad_norm": 3.455172435512445,
+      "learning_rate": 1.53904318556879e-05,
+      "loss": 2.9353,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3395700062527055,
+      "grad_norm": 3.697383160473131,
+      "learning_rate": 1.5385181662683244e-05,
+      "loss": 3.2539,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3397623971910923,
+      "grad_norm": 2.511718029455896,
+      "learning_rate": 1.5379929378136088e-05,
+      "loss": 2.6803,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3399547881294791,
+      "grad_norm": 3.7707702847353213,
+      "learning_rate": 1.5374675004086353e-05,
+      "loss": 3.4168,
+      "step": 1767
+    },
+    {
+      "epoch": 0.3401471790678659,
+      "grad_norm": 3.5776114184669905,
+      "learning_rate": 1.5369418542574782e-05,
+      "loss": 3.3784,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3403395700062527,
+      "grad_norm": 3.7401115772772004,
+      "learning_rate": 1.536415999564292e-05,
+      "loss": 3.1185,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3405319609446395,
+      "grad_norm": 3.937369314899265,
+      "learning_rate": 1.5358899365333123e-05,
+      "loss": 3.0784,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3407243518830263,
+      "grad_norm": 2.2176677876942152,
+      "learning_rate": 1.5353636653688563e-05,
+      "loss": 2.5249,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3409167428214131,
+      "grad_norm": 3.7249385139875155,
+      "learning_rate": 1.534837186275322e-05,
+      "loss": 2.9706,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3411091337597999,
+      "grad_norm": 3.512925985237401,
+      "learning_rate": 1.5343104994571877e-05,
+      "loss": 2.9188,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3413015246981867,
+      "grad_norm": 3.354248787339819,
+      "learning_rate": 1.533783605119012e-05,
+      "loss": 3.0741,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3414939156365735,
+      "grad_norm": 3.4384860181128856,
+      "learning_rate": 1.5332565034654344e-05,
+      "loss": 3.214,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3416863065749603,
+      "grad_norm": 3.4914805479909576,
+      "learning_rate": 1.5327291947011763e-05,
+      "loss": 3.1518,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3418786975133471,
+      "grad_norm": 3.3960308209009242,
+      "learning_rate": 1.5322016790310373e-05,
+      "loss": 2.8617,
+      "step": 1777
+    },
+    {
+      "epoch": 0.3420710884517339,
+      "grad_norm": 3.3178088495525135,
+      "learning_rate": 1.5316739566598985e-05,
+      "loss": 3.272,
+      "step": 1778
+    },
+    {
+      "epoch": 0.34226347939012075,
+      "grad_norm": 3.3148496112077708,
+      "learning_rate": 1.531146027792722e-05,
+      "loss": 2.8894,
+      "step": 1779
+    },
+    {
+      "epoch": 0.34245587032850755,
+      "grad_norm": 2.5240471751631253,
+      "learning_rate": 1.530617892634548e-05,
+      "loss": 2.9379,
+      "step": 1780
+    },
+    {
+      "epoch": 0.34264826126689435,
+      "grad_norm": 3.4307866754949834,
+      "learning_rate": 1.5300895513904993e-05,
+      "loss": 2.9829,
+      "step": 1781
+    },
+    {
+      "epoch": 0.34284065220528115,
+      "grad_norm": 3.6935790198693677,
+      "learning_rate": 1.529561004265777e-05,
+      "loss": 3.333,
+      "step": 1782
+    },
+    {
+      "epoch": 0.34303304314366795,
+      "grad_norm": 3.5322162163395596,
+      "learning_rate": 1.5290322514656624e-05,
+      "loss": 3.0274,
+      "step": 1783
+    },
+    {
+      "epoch": 0.34322543408205475,
+      "grad_norm": 3.258889800558704,
+      "learning_rate": 1.5285032931955177e-05,
+      "loss": 3.096,
+      "step": 1784
+    },
+    {
+      "epoch": 0.34341782502044155,
+      "grad_norm": 3.70973267827125,
+      "learning_rate": 1.527974129660784e-05,
+      "loss": 3.2152,
+      "step": 1785
+    },
+    {
+      "epoch": 0.34361021595882835,
+      "grad_norm": 2.34021164690482,
+      "learning_rate": 1.527444761066982e-05,
+      "loss": 2.6028,
+      "step": 1786
+    },
+    {
+      "epoch": 0.34380260689721515,
+      "grad_norm": 3.6939478934949954,
+      "learning_rate": 1.5269151876197127e-05,
+      "loss": 3.127,
+      "step": 1787
+    },
+    {
+      "epoch": 0.34399499783560195,
+      "grad_norm": 3.4748951991922796,
+      "learning_rate": 1.5263854095246557e-05,
+      "loss": 3.2365,
+      "step": 1788
+    },
+    {
+      "epoch": 0.34418738877398874,
+      "grad_norm": 3.5953281210628605,
+      "learning_rate": 1.5258554269875716e-05,
+      "loss": 3.266,
+      "step": 1789
+    },
+    {
+      "epoch": 0.34437977971237554,
+      "grad_norm": 3.219548079590355,
+      "learning_rate": 1.5253252402142989e-05,
+      "loss": 2.9255,
+      "step": 1790
+    },
+    {
+      "epoch": 0.34457217065076234,
+      "grad_norm": 3.9664678206278556,
+      "learning_rate": 1.5247948494107566e-05,
+      "loss": 3.107,
+      "step": 1791
+    },
+    {
+      "epoch": 0.34476456158914914,
+      "grad_norm": 3.3194420021692044,
+      "learning_rate": 1.5242642547829416e-05,
+      "loss": 3.2731,
+      "step": 1792
+    },
+    {
+      "epoch": 0.34495695252753594,
+      "grad_norm": 3.44662057144085,
+      "learning_rate": 1.523733456536931e-05,
+      "loss": 3.1519,
+      "step": 1793
+    },
+    {
+      "epoch": 0.34514934346592274,
+      "grad_norm": 3.7036687569174327,
+      "learning_rate": 1.5232024548788813e-05,
+      "loss": 3.2251,
+      "step": 1794
+    },
+    {
+      "epoch": 0.34534173440430954,
+      "grad_norm": 3.3388397446348157,
+      "learning_rate": 1.5226712500150267e-05,
+      "loss": 3.1603,
+      "step": 1795
+    },
+    {
+      "epoch": 0.34553412534269634,
+      "grad_norm": 3.52911474208417,
+      "learning_rate": 1.5221398421516816e-05,
+      "loss": 3.3794,
+      "step": 1796
+    },
+    {
+      "epoch": 0.34572651628108314,
+      "grad_norm": 3.2633618997097735,
+      "learning_rate": 1.5216082314952383e-05,
+      "loss": 3.3483,
+      "step": 1797
+    },
+    {
+      "epoch": 0.34591890721946994,
+      "grad_norm": 2.5050879340750254,
+      "learning_rate": 1.521076418252168e-05,
+      "loss": 2.7921,
+      "step": 1798
+    },
+    {
+      "epoch": 0.34611129815785674,
+      "grad_norm": 2.2686528282854677,
+      "learning_rate": 1.5205444026290218e-05,
+      "loss": 2.7386,
+      "step": 1799
+    },
+    {
+      "epoch": 0.3463036890962436,
+      "grad_norm": 3.52204703431708,
+      "learning_rate": 1.5200121848324276e-05,
+      "loss": 3.2682,
+      "step": 1800
+    },
+    {
+      "epoch": 0.3464960800346304,
+      "grad_norm": 3.7312365378108754,
+      "learning_rate": 1.5194797650690926e-05,
+      "loss": 3.126,
+      "step": 1801
+    },
+    {
+      "epoch": 0.3466884709730172,
+      "grad_norm": 3.4109761772129956,
+      "learning_rate": 1.5189471435458032e-05,
+      "loss": 3.1409,
+      "step": 1802
+    },
+    {
+      "epoch": 0.346880861911404,
+      "grad_norm": 2.362434767672615,
+      "learning_rate": 1.5184143204694231e-05,
+      "loss": 2.568,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3470732528497908,
+      "grad_norm": 3.620888966871579,
+      "learning_rate": 1.5178812960468945e-05,
+      "loss": 2.9843,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3472656437881776,
+      "grad_norm": 3.481712361921981,
+      "learning_rate": 1.5173480704852379e-05,
+      "loss": 3.1461,
+      "step": 1805
+    },
+    {
+      "epoch": 0.3474580347265644,
+      "grad_norm": 3.685940809726664,
+      "learning_rate": 1.5168146439915525e-05,
+      "loss": 3.0737,
+      "step": 1806
+    },
+    {
+      "epoch": 0.3476504256649512,
+      "grad_norm": 3.3231083355024884,
+      "learning_rate": 1.5162810167730144e-05,
+      "loss": 2.9308,
+      "step": 1807
+    },
+    {
+      "epoch": 0.347842816603338,
+      "grad_norm": 3.656836872022973,
+      "learning_rate": 1.5157471890368785e-05,
+      "loss": 3.2648,
+      "step": 1808
+    },
+    {
+      "epoch": 0.3480352075417248,
+      "grad_norm": 3.6216765469386356,
+      "learning_rate": 1.5152131609904773e-05,
+      "loss": 3.0771,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3482275984801116,
+      "grad_norm": 3.4786748916994568,
+      "learning_rate": 1.5146789328412213e-05,
+      "loss": 3.3483,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3484199894184984,
+      "grad_norm": 2.2256127988755114,
+      "learning_rate": 1.5141445047965984e-05,
+      "loss": 2.6317,
+      "step": 1811
+    },
+    {
+      "epoch": 0.3486123803568852,
+      "grad_norm": 3.5405751750007326,
+      "learning_rate": 1.5136098770641741e-05,
+      "loss": 3.3087,
+      "step": 1812
+    },
+    {
+      "epoch": 0.348804771295272,
+      "grad_norm": 3.502896696567136,
+      "learning_rate": 1.513075049851592e-05,
+      "loss": 3.4361,
+      "step": 1813
+    },
+    {
+      "epoch": 0.3489971622336588,
+      "grad_norm": 3.3833352761848055,
+      "learning_rate": 1.5125400233665728e-05,
+      "loss": 3.0914,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3491895531720456,
+      "grad_norm": 3.429185250670187,
+      "learning_rate": 1.5120047978169146e-05,
+      "loss": 3.1662,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3493819441104324,
+      "grad_norm": 3.642862927307115,
+      "learning_rate": 1.5114693734104926e-05,
+      "loss": 3.1115,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3495743350488192,
+      "grad_norm": 3.3809768122619803,
+      "learning_rate": 1.5109337503552594e-05,
+      "loss": 3.1539,
+      "step": 1817
+    },
+    {
+      "epoch": 0.349766725987206,
+      "grad_norm": 3.551372155378786,
+      "learning_rate": 1.5103979288592454e-05,
+      "loss": 2.7959,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3499591169255928,
+      "grad_norm": 3.399362444826478,
+      "learning_rate": 1.5098619091305571e-05,
+      "loss": 3.3229,
+      "step": 1819
+    },
+    {
+      "epoch": 0.3501515078639796,
+      "grad_norm": 3.3075707757992765,
+      "learning_rate": 1.5093256913773786e-05,
+      "loss": 3.1244,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3503438988023664,
+      "grad_norm": 3.6920423406441243,
+      "learning_rate": 1.50878927580797e-05,
+      "loss": 3.0295,
+      "step": 1821
+    },
+    {
+      "epoch": 0.35053628974075324,
+      "grad_norm": 3.58323987928435,
+      "learning_rate": 1.5082526626306698e-05,
+      "loss": 3.1654,
+      "step": 1822
+    },
+    {
+      "epoch": 0.35072868067914004,
+      "grad_norm": 3.4843289617677624,
+      "learning_rate": 1.5077158520538921e-05,
+      "loss": 3.3279,
+      "step": 1823
+    },
+    {
+      "epoch": 0.35092107161752684,
+      "grad_norm": 3.7032639746916343,
+      "learning_rate": 1.5071788442861277e-05,
+      "loss": 2.7806,
+      "step": 1824
+    },
+    {
+      "epoch": 0.35111346255591364,
+      "grad_norm": 3.428468560940912,
+      "learning_rate": 1.5066416395359444e-05,
+      "loss": 3.1247,
+      "step": 1825
+    },
+    {
+      "epoch": 0.35130585349430044,
+      "grad_norm": 3.469613477789403,
+      "learning_rate": 1.5061042380119864e-05,
+      "loss": 3.3326,
+      "step": 1826
+    },
+    {
+      "epoch": 0.35149824443268723,
+      "grad_norm": 3.3752859078328354,
+      "learning_rate": 1.5055666399229743e-05,
+      "loss": 3.0406,
+      "step": 1827
+    },
+    {
+      "epoch": 0.35169063537107403,
+      "grad_norm": 3.492555733490673,
+      "learning_rate": 1.5050288454777047e-05,
+      "loss": 3.2204,
+      "step": 1828
+    },
+    {
+      "epoch": 0.35188302630946083,
+      "grad_norm": 3.2485425934761833,
+      "learning_rate": 1.504490854885051e-05,
+      "loss": 3.1761,
+      "step": 1829
+    },
+    {
+      "epoch": 0.35207541724784763,
+      "grad_norm": 3.5469965249770548,
+      "learning_rate": 1.5039526683539627e-05,
+      "loss": 3.3824,
+      "step": 1830
+    },
+    {
+      "epoch": 0.35226780818623443,
+      "grad_norm": 3.3887819654971993,
+      "learning_rate": 1.5034142860934649e-05,
+      "loss": 3.3888,
+      "step": 1831
+    },
+    {
+      "epoch": 0.35246019912462123,
+      "grad_norm": 3.3108721618714108,
+      "learning_rate": 1.5028757083126594e-05,
+      "loss": 3.0871,
+      "step": 1832
+    },
+    {
+      "epoch": 0.35265259006300803,
+      "grad_norm": 3.5808335764244563,
+      "learning_rate": 1.5023369352207229e-05,
+      "loss": 3.2235,
+      "step": 1833
+    },
+    {
+      "epoch": 0.35284498100139483,
+      "grad_norm": 3.535387137321828,
+      "learning_rate": 1.5017979670269096e-05,
+      "loss": 3.3013,
+      "step": 1834
+    },
+    {
+      "epoch": 0.35303737193978163,
+      "grad_norm": 3.5209816548972204,
+      "learning_rate": 1.501258803940548e-05,
+      "loss": 2.681,
+      "step": 1835
+    },
+    {
+      "epoch": 0.35322976287816843,
+      "grad_norm": 3.5702937981207836,
+      "learning_rate": 1.500719446171043e-05,
+      "loss": 2.9824,
+      "step": 1836
+    },
+    {
+      "epoch": 0.35342215381655523,
+      "grad_norm": 3.461341074284008,
+      "learning_rate": 1.500179893927875e-05,
+      "loss": 3.2059,
+      "step": 1837
+    },
+    {
+      "epoch": 0.35361454475494203,
+      "grad_norm": 3.59290768660366,
+      "learning_rate": 1.4996401474205997e-05,
+      "loss": 3.0401,
+      "step": 1838
+    },
+    {
+      "epoch": 0.35380693569332883,
+      "grad_norm": 3.5890023889280105,
+      "learning_rate": 1.4991002068588484e-05,
+      "loss": 3.1298,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3539993266317156,
+      "grad_norm": 3.4482173034411687,
+      "learning_rate": 1.4985600724523282e-05,
+      "loss": 3.1267,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3541917175701024,
+      "grad_norm": 3.8635390419403084,
+      "learning_rate": 1.4980197444108205e-05,
+      "loss": 3.2582,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3543841085084892,
+      "grad_norm": 3.322720018699568,
+      "learning_rate": 1.4974792229441826e-05,
+      "loss": 2.8626,
+      "step": 1842
+    },
+    {
+      "epoch": 0.354576499446876,
+      "grad_norm": 3.4568087940080012,
+      "learning_rate": 1.4969385082623473e-05,
+      "loss": 3.1829,
+      "step": 1843
+    },
+    {
+      "epoch": 0.3547688903852629,
+      "grad_norm": 3.430346230442413,
+      "learning_rate": 1.4963976005753216e-05,
+      "loss": 3.2223,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3549612813236497,
+      "grad_norm": 2.585828997734722,
+      "learning_rate": 1.4958565000931877e-05,
+      "loss": 2.5599,
+      "step": 1845
+    },
+    {
+      "epoch": 0.3551536722620365,
+      "grad_norm": 3.5901141886067274,
+      "learning_rate": 1.4953152070261027e-05,
+      "loss": 2.8589,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3553460632004233,
+      "grad_norm": 3.695397322242156,
+      "learning_rate": 1.494773721584299e-05,
+      "loss": 3.105,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3555384541388101,
+      "grad_norm": 3.4625887613167583,
+      "learning_rate": 1.4942320439780833e-05,
+      "loss": 3.0458,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3557308450771969,
+      "grad_norm": 3.584268005412205,
+      "learning_rate": 1.4936901744178367e-05,
+      "loss": 3.3042,
+      "step": 1849
+    },
+    {
+      "epoch": 0.3559232360155837,
+      "grad_norm": 3.6235793527872895,
+      "learning_rate": 1.4931481131140149e-05,
+      "loss": 3.4071,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3561156269539705,
+      "grad_norm": 3.4519286065112027,
+      "learning_rate": 1.4926058602771484e-05,
+      "loss": 3.1467,
+      "step": 1851
+    },
+    {
+      "epoch": 0.3563080178923573,
+      "grad_norm": 3.5397249174218754,
+      "learning_rate": 1.4920634161178424e-05,
+      "loss": 3.2701,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3565004088307441,
+      "grad_norm": 3.8970127133664834,
+      "learning_rate": 1.4915207808467756e-05,
+      "loss": 3.3746,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3566927997691309,
+      "grad_norm": 3.342983700514847,
+      "learning_rate": 1.4909779546747011e-05,
+      "loss": 3.2622,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3568851907075177,
+      "grad_norm": 3.132006088631519,
+      "learning_rate": 1.4904349378124467e-05,
+      "loss": 3.0875,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3570775816459045,
+      "grad_norm": 3.4939705107349597,
+      "learning_rate": 1.489891730470914e-05,
+      "loss": 3.2817,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3572699725842913,
+      "grad_norm": 3.5130006647890513,
+      "learning_rate": 1.4893483328610778e-05,
+      "loss": 2.9582,
+      "step": 1857
+    },
+    {
+      "epoch": 0.3574623635226781,
+      "grad_norm": 3.4204390701573253,
+      "learning_rate": 1.488804745193988e-05,
+      "loss": 3.0812,
+      "step": 1858
+    },
+    {
+      "epoch": 0.35765475446106487,
+      "grad_norm": 3.426904917904968,
+      "learning_rate": 1.4882609676807675e-05,
+      "loss": 3.2533,
+      "step": 1859
+    },
+    {
+      "epoch": 0.35784714539945167,
+      "grad_norm": 2.581529219988701,
+      "learning_rate": 1.4877170005326136e-05,
+      "loss": 2.7159,
+      "step": 1860
+    },
+    {
+      "epoch": 0.35803953633783847,
+      "grad_norm": 3.271407544991519,
+      "learning_rate": 1.4871728439607967e-05,
+      "loss": 3.317,
+      "step": 1861
+    },
+    {
+      "epoch": 0.35823192727622527,
+      "grad_norm": 3.2473433168279366,
+      "learning_rate": 1.4866284981766607e-05,
+      "loss": 3.083,
+      "step": 1862
+    },
+    {
+      "epoch": 0.35842431821461207,
+      "grad_norm": 3.275470688582748,
+      "learning_rate": 1.4860839633916236e-05,
+      "loss": 3.3424,
+      "step": 1863
+    },
+    {
+      "epoch": 0.35861670915299887,
+      "grad_norm": 3.3493473758079895,
+      "learning_rate": 1.4855392398171762e-05,
+      "loss": 3.0401,
+      "step": 1864
+    },
+    {
+      "epoch": 0.35880910009138567,
+      "grad_norm": 3.424110006747634,
+      "learning_rate": 1.484994327664883e-05,
+      "loss": 2.9851,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3590014910297725,
+      "grad_norm": 3.509481250350719,
+      "learning_rate": 1.4844492271463814e-05,
+      "loss": 3.2347,
+      "step": 1866
+    },
+    {
+      "epoch": 0.3591938819681593,
+      "grad_norm": 3.80895377717247,
+      "learning_rate": 1.4839039384733821e-05,
+      "loss": 3.5319,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3593862729065461,
+      "grad_norm": 2.4547087495049595,
+      "learning_rate": 1.4833584618576695e-05,
+      "loss": 2.7291,
+      "step": 1868
+    },
+    {
+      "epoch": 0.3595786638449329,
+      "grad_norm": 3.0153471718007263,
+      "learning_rate": 1.4828127975111e-05,
+      "loss": 3.1906,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3597710547833197,
+      "grad_norm": 3.3804436533259516,
+      "learning_rate": 1.4822669456456031e-05,
+      "loss": 2.9129,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3599634457217065,
+      "grad_norm": 3.6051039791719774,
+      "learning_rate": 1.4817209064731819e-05,
+      "loss": 3.4621,
+      "step": 1871
+    },
+    {
+      "epoch": 0.3601558366600933,
+      "grad_norm": 3.4997874836842895,
+      "learning_rate": 1.4811746802059115e-05,
+      "loss": 3.3697,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3603482275984801,
+      "grad_norm": 3.5104527813665753,
+      "learning_rate": 1.48062826705594e-05,
+      "loss": 3.1432,
+      "step": 1873
+    },
+    {
+      "epoch": 0.3605406185368669,
+      "grad_norm": 3.676889824984,
+      "learning_rate": 1.4800816672354876e-05,
+      "loss": 3.165,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3607330094752537,
+      "grad_norm": 3.4955333286327472,
+      "learning_rate": 1.4795348809568477e-05,
+      "loss": 2.9155,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3609254004136405,
+      "grad_norm": 3.3899755811812553,
+      "learning_rate": 1.4789879084323858e-05,
+      "loss": 3.338,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3611177913520273,
+      "grad_norm": 3.4212703573168373,
+      "learning_rate": 1.4784407498745394e-05,
+      "loss": 3.1751,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3613101822904141,
+      "grad_norm": 3.4985666167567206,
+      "learning_rate": 1.477893405495819e-05,
+      "loss": 3.1374,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3615025732288009,
+      "grad_norm": 3.581488096672409,
+      "learning_rate": 1.4773458755088068e-05,
+      "loss": 3.4771,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3616949641671877,
+      "grad_norm": 3.4990369513329878,
+      "learning_rate": 1.4767981601261567e-05,
+      "loss": 2.9355,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3618873551055745,
+      "grad_norm": 3.465191629043056,
+      "learning_rate": 1.4762502595605957e-05,
+      "loss": 2.993,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3620797460439613,
+      "grad_norm": 3.298900513486685,
+      "learning_rate": 1.4757021740249213e-05,
+      "loss": 3.3703,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3622721369823481,
+      "grad_norm": 3.4946947772605013,
+      "learning_rate": 1.4751539037320044e-05,
+      "loss": 3.1781,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3624645279207349,
+      "grad_norm": 3.7345863407221684,
+      "learning_rate": 1.4746054488947863e-05,
+      "loss": 3.3444,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3626569188591217,
+      "grad_norm": 2.4374476023988687,
+      "learning_rate": 1.4740568097262811e-05,
+      "loss": 2.8204,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3628493097975085,
+      "grad_norm": 3.7201882313082124,
+      "learning_rate": 1.473507986439573e-05,
+      "loss": 3.144,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3630417007358953,
+      "grad_norm": 3.6373036063315345,
+      "learning_rate": 1.4729589792478193e-05,
+      "loss": 3.0385,
+      "step": 1887
+    },
+    {
+      "epoch": 0.36323409167428217,
+      "grad_norm": 3.506063165101945,
+      "learning_rate": 1.4724097883642482e-05,
+      "loss": 3.0846,
+      "step": 1888
+    },
+    {
+      "epoch": 0.36342648261266897,
+      "grad_norm": 3.3859983853510194,
+      "learning_rate": 1.4718604140021588e-05,
+      "loss": 3.3618,
+      "step": 1889
+    },
+    {
+      "epoch": 0.36361887355105577,
+      "grad_norm": 3.592685130616504,
+      "learning_rate": 1.471310856374922e-05,
+      "loss": 3.2674,
+      "step": 1890
+    },
+    {
+      "epoch": 0.36381126448944257,
+      "grad_norm": 3.5774360107907297,
+      "learning_rate": 1.470761115695979e-05,
+      "loss": 3.0173,
+      "step": 1891
+    },
+    {
+      "epoch": 0.36400365542782936,
+      "grad_norm": 3.9134490995316735,
+      "learning_rate": 1.4702111921788437e-05,
+      "loss": 3.0772,
+      "step": 1892
+    },
+    {
+      "epoch": 0.36419604636621616,
+      "grad_norm": 3.5926187193077923,
+      "learning_rate": 1.4696610860370997e-05,
+      "loss": 3.0944,
+      "step": 1893
+    },
+    {
+      "epoch": 0.36438843730460296,
+      "grad_norm": 2.5741964119824265,
+      "learning_rate": 1.4691107974844015e-05,
+      "loss": 2.7252,
+      "step": 1894
+    },
+    {
+      "epoch": 0.36458082824298976,
+      "grad_norm": 3.579067047005191,
+      "learning_rate": 1.468560326734475e-05,
+      "loss": 3.2983,
+      "step": 1895
+    },
+    {
+      "epoch": 0.36477321918137656,
+      "grad_norm": 3.6319356100169364,
+      "learning_rate": 1.4680096740011172e-05,
+      "loss": 3.4243,
+      "step": 1896
+    },
+    {
+      "epoch": 0.36496561011976336,
+      "grad_norm": 3.5620822923759836,
+      "learning_rate": 1.4674588394981948e-05,
+      "loss": 3.4634,
+      "step": 1897
+    },
+    {
+      "epoch": 0.36515800105815016,
+      "grad_norm": 3.2607493680392117,
+      "learning_rate": 1.4669078234396454e-05,
+      "loss": 3.134,
+      "step": 1898
+    },
+    {
+      "epoch": 0.36535039199653696,
+      "grad_norm": 3.6214405990914065,
+      "learning_rate": 1.4663566260394775e-05,
+      "loss": 3.2419,
+      "step": 1899
+    },
+    {
+      "epoch": 0.36554278293492376,
+      "grad_norm": 3.1719161657382866,
+      "learning_rate": 1.4658052475117704e-05,
+      "loss": 3.2406,
+      "step": 1900
+    },
+    {
+      "epoch": 0.36573517387331056,
+      "grad_norm": 3.318072959178684,
+      "learning_rate": 1.4652536880706723e-05,
+      "loss": 3.3493,
+      "step": 1901
+    },
+    {
+      "epoch": 0.36592756481169736,
+      "grad_norm": 3.3748366934050633,
+      "learning_rate": 1.4647019479304028e-05,
+      "loss": 3.3096,
+      "step": 1902
+    },
+    {
+      "epoch": 0.36611995575008416,
+      "grad_norm": 3.281902907509956,
+      "learning_rate": 1.4641500273052516e-05,
+      "loss": 3.1991,
+      "step": 1903
+    },
+    {
+      "epoch": 0.36631234668847096,
+      "grad_norm": 3.6673349277009306,
+      "learning_rate": 1.463597926409578e-05,
+      "loss": 3.4007,
+      "step": 1904
+    },
+    {
+      "epoch": 0.36650473762685776,
+      "grad_norm": 3.577091744166845,
+      "learning_rate": 1.4630456454578122e-05,
+      "loss": 3.3403,
+      "step": 1905
+    },
+    {
+      "epoch": 0.36669712856524456,
+      "grad_norm": 3.2672228906302223,
+      "learning_rate": 1.462493184664453e-05,
+      "loss": 3.0992,
+      "step": 1906
+    },
+    {
+      "epoch": 0.36688951950363136,
+      "grad_norm": 3.4943999256881804,
+      "learning_rate": 1.4619405442440702e-05,
+      "loss": 3.1297,
+      "step": 1907
+    },
+    {
+      "epoch": 0.36708191044201816,
+      "grad_norm": 3.5713511626815233,
+      "learning_rate": 1.4613877244113033e-05,
+      "loss": 3.3694,
+      "step": 1908
+    },
+    {
+      "epoch": 0.367274301380405,
+      "grad_norm": 3.5959005875688246,
+      "learning_rate": 1.4608347253808605e-05,
+      "loss": 3.0726,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3674666923187918,
+      "grad_norm": 2.264055605160871,
+      "learning_rate": 1.460281547367521e-05,
+      "loss": 2.7693,
+      "step": 1910
+    },
+    {
+      "epoch": 0.3676590832571786,
+      "grad_norm": 3.4027986029788746,
+      "learning_rate": 1.4597281905861318e-05,
+      "loss": 3.2923,
+      "step": 1911
+    },
+    {
+      "epoch": 0.3678514741955654,
+      "grad_norm": 3.057021791371183,
+      "learning_rate": 1.4591746552516109e-05,
+      "loss": 3.0783,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3680438651339522,
+      "grad_norm": 3.526800594992673,
+      "learning_rate": 1.4586209415789452e-05,
+      "loss": 3.4761,
+      "step": 1913
+    },
+    {
+      "epoch": 0.368236256072339,
+      "grad_norm": 2.1225343099378136,
+      "learning_rate": 1.4580670497831904e-05,
+      "loss": 2.506,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3684286470107258,
+      "grad_norm": 3.347358462371007,
+      "learning_rate": 1.4575129800794718e-05,
+      "loss": 3.0133,
+      "step": 1915
+    },
+    {
+      "epoch": 0.3686210379491126,
+      "grad_norm": 3.402265490871244,
+      "learning_rate": 1.4569587326829834e-05,
+      "loss": 3.1662,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3688134288874994,
+      "grad_norm": 3.525261312280619,
+      "learning_rate": 1.4564043078089891e-05,
+      "loss": 3.0971,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3690058198258862,
+      "grad_norm": 3.427077404404568,
+      "learning_rate": 1.4558497056728205e-05,
+      "loss": 2.9417,
+      "step": 1918
+    },
+    {
+      "epoch": 0.369198210764273,
+      "grad_norm": 3.3291386249852137,
+      "learning_rate": 1.4552949264898795e-05,
+      "loss": 3.0289,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3693906017026598,
+      "grad_norm": 3.3113071733697472,
+      "learning_rate": 1.4547399704756348e-05,
+      "loss": 3.1464,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3695829926410466,
+      "grad_norm": 3.3938842648754806,
+      "learning_rate": 1.4541848378456255e-05,
+      "loss": 3.1303,
+      "step": 1921
+    },
+    {
+      "epoch": 0.3697753835794334,
+      "grad_norm": 3.661129129239712,
+      "learning_rate": 1.4536295288154594e-05,
+      "loss": 3.2156,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3699677745178202,
+      "grad_norm": 2.4852312166088826,
+      "learning_rate": 1.4530740436008111e-05,
+      "loss": 2.7724,
+      "step": 1923
+    },
+    {
+      "epoch": 0.370160165456207,
+      "grad_norm": 3.655906859730094,
+      "learning_rate": 1.452518382417425e-05,
+      "loss": 3.1555,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3703525563945938,
+      "grad_norm": 2.2170859543024433,
+      "learning_rate": 1.4519625454811135e-05,
+      "loss": 2.5273,
+      "step": 1925
+    },
+    {
+      "epoch": 0.3705449473329806,
+      "grad_norm": 3.558105332182304,
+      "learning_rate": 1.4514065330077575e-05,
+      "loss": 3.3372,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3707373382713674,
+      "grad_norm": 3.3641363362474483,
+      "learning_rate": 1.4508503452133053e-05,
+      "loss": 3.0946,
+      "step": 1927
+    },
+    {
+      "epoch": 0.3709297292097542,
+      "grad_norm": 3.638197080299161,
+      "learning_rate": 1.4502939823137744e-05,
+      "loss": 3.2429,
+      "step": 1928
+    },
+    {
+      "epoch": 0.371122120148141,
+      "grad_norm": 3.4207377219562223,
+      "learning_rate": 1.4497374445252496e-05,
+      "loss": 3.1509,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3713145110865278,
+      "grad_norm": 3.4569167943001586,
+      "learning_rate": 1.4491807320638835e-05,
+      "loss": 3.2759,
+      "step": 1930
+    },
+    {
+      "epoch": 0.37150690202491465,
+      "grad_norm": 3.400208486966514,
+      "learning_rate": 1.4486238451458972e-05,
+      "loss": 3.2857,
+      "step": 1931
+    },
+    {
+      "epoch": 0.37169929296330145,
+      "grad_norm": 3.511864883276557,
+      "learning_rate": 1.4480667839875786e-05,
+      "loss": 3.2135,
+      "step": 1932
+    },
+    {
+      "epoch": 0.37189168390168825,
+      "grad_norm": 2.8406825882376636,
+      "learning_rate": 1.4475095488052843e-05,
+      "loss": 2.6993,
+      "step": 1933
+    },
+    {
+      "epoch": 0.37208407484007505,
+      "grad_norm": 3.440613162002084,
+      "learning_rate": 1.4469521398154381e-05,
+      "loss": 3.027,
+      "step": 1934
+    },
+    {
+      "epoch": 0.37227646577846185,
+      "grad_norm": 3.4600331481042836,
+      "learning_rate": 1.4463945572345308e-05,
+      "loss": 3.3224,
+      "step": 1935
+    },
+    {
+      "epoch": 0.37246885671684865,
+      "grad_norm": 3.6390273151033274,
+      "learning_rate": 1.4458368012791213e-05,
+      "loss": 3.1625,
+      "step": 1936
+    },
+    {
+      "epoch": 0.37266124765523545,
+      "grad_norm": 3.3970066803614176,
+      "learning_rate": 1.4452788721658355e-05,
+      "loss": 3.2897,
+      "step": 1937
+    },
+    {
+      "epoch": 0.37285363859362225,
+      "grad_norm": 3.3366723710099655,
+      "learning_rate": 1.4447207701113669e-05,
+      "loss": 2.9948,
+      "step": 1938
+    },
+    {
+      "epoch": 0.37304602953200905,
+      "grad_norm": 3.572181167113255,
+      "learning_rate": 1.4441624953324755e-05,
+      "loss": 2.9394,
+      "step": 1939
+    },
+    {
+      "epoch": 0.37323842047039585,
+      "grad_norm": 3.650477974218113,
+      "learning_rate": 1.4436040480459891e-05,
+      "loss": 3.4199,
+      "step": 1940
+    },
+    {
+      "epoch": 0.37343081140878265,
+      "grad_norm": 3.5389565004275614,
+      "learning_rate": 1.443045428468802e-05,
+      "loss": 3.1467,
+      "step": 1941
+    },
+    {
+      "epoch": 0.37362320234716945,
+      "grad_norm": 3.100860226156007,
+      "learning_rate": 1.4424866368178761e-05,
+      "loss": 2.8052,
+      "step": 1942
+    },
+    {
+      "epoch": 0.37381559328555625,
+      "grad_norm": 3.4847560840359533,
+      "learning_rate": 1.441927673310239e-05,
+      "loss": 2.9271,
+      "step": 1943
+    },
+    {
+      "epoch": 0.37400798422394305,
+      "grad_norm": 3.456460596562106,
+      "learning_rate": 1.4413685381629855e-05,
+      "loss": 3.3052,
+      "step": 1944
+    },
+    {
+      "epoch": 0.37420037516232985,
+      "grad_norm": 3.5577046043843246,
+      "learning_rate": 1.440809231593278e-05,
+      "loss": 3.1793,
+      "step": 1945
+    },
+    {
+      "epoch": 0.37439276610071665,
+      "grad_norm": 3.7488043564314903,
+      "learning_rate": 1.4402497538183444e-05,
+      "loss": 3.3852,
+      "step": 1946
+    },
+    {
+      "epoch": 0.37458515703910344,
+      "grad_norm": 3.3752157579894133,
+      "learning_rate": 1.4396901050554794e-05,
+      "loss": 3.0879,
+      "step": 1947
+    },
+    {
+      "epoch": 0.37477754797749024,
+      "grad_norm": 3.4667131394047006,
+      "learning_rate": 1.4391302855220442e-05,
+      "loss": 3.1559,
+      "step": 1948
+    },
+    {
+      "epoch": 0.37496993891587704,
+      "grad_norm": 3.596208343999809,
+      "learning_rate": 1.4385702954354662e-05,
+      "loss": 3.2955,
+      "step": 1949
+    },
+    {
+      "epoch": 0.37516232985426384,
+      "grad_norm": 3.646891573395568,
+      "learning_rate": 1.438010135013239e-05,
+      "loss": 2.9025,
+      "step": 1950
+    },
+    {
+      "epoch": 0.37535472079265064,
+      "grad_norm": 3.5020669071841004,
+      "learning_rate": 1.4374498044729225e-05,
+      "loss": 3.3462,
+      "step": 1951
+    },
+    {
+      "epoch": 0.37554711173103744,
+      "grad_norm": 3.4408136523264483,
+      "learning_rate": 1.4368893040321428e-05,
+      "loss": 3.2295,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3757395026694243,
+      "grad_norm": 3.3627420476592644,
+      "learning_rate": 1.4363286339085915e-05,
+      "loss": 2.986,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3759318936078111,
+      "grad_norm": 3.4871628750872157,
+      "learning_rate": 1.435767794320027e-05,
+      "loss": 3.2424,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3761242845461979,
+      "grad_norm": 3.626244482284534,
+      "learning_rate": 1.4352067854842724e-05,
+      "loss": 3.4717,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3763166754845847,
+      "grad_norm": 3.470942130068345,
+      "learning_rate": 1.434645607619217e-05,
+      "loss": 3.3247,
+      "step": 1956
+    },
+    {
+      "epoch": 0.3765090664229715,
+      "grad_norm": 3.4647447217038643,
+      "learning_rate": 1.434084260942816e-05,
+      "loss": 3.3392,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3767014573613583,
+      "grad_norm": 3.532421476867847,
+      "learning_rate": 1.4335227456730902e-05,
+      "loss": 3.1568,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3768938482997451,
+      "grad_norm": 3.3844012079027372,
+      "learning_rate": 1.4329610620281253e-05,
+      "loss": 3.1918,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3770862392381319,
+      "grad_norm": 2.4412559757639314,
+      "learning_rate": 1.4323992102260733e-05,
+      "loss": 2.6861,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3772786301765187,
+      "grad_norm": 3.3674626637168448,
+      "learning_rate": 1.4318371904851502e-05,
+      "loss": 3.1817,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3774710211149055,
+      "grad_norm": 3.5097355400228776,
+      "learning_rate": 1.4312750030236382e-05,
+      "loss": 3.3564,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3776634120532923,
+      "grad_norm": 3.377720936823551,
+      "learning_rate": 1.4307126480598852e-05,
+      "loss": 3.2856,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3778558029916791,
+      "grad_norm": 3.4883179963206437,
+      "learning_rate": 1.4301501258123024e-05,
+      "loss": 2.9496,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3780481939300659,
+      "grad_norm": 2.3746540306737782,
+      "learning_rate": 1.4295874364993672e-05,
+      "loss": 2.4578,
+      "step": 1965
+    },
+    {
+      "epoch": 0.3782405848684527,
+      "grad_norm": 2.359834962786505,
+      "learning_rate": 1.4290245803396221e-05,
+      "loss": 2.5317,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3784329758068395,
+      "grad_norm": 3.2299004560184685,
+      "learning_rate": 1.4284615575516737e-05,
+      "loss": 3.1458,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3786253667452263,
+      "grad_norm": 3.2781931207396826,
+      "learning_rate": 1.4278983683541934e-05,
+      "loss": 3.2533,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3788177576836131,
+      "grad_norm": 3.11230299018791,
+      "learning_rate": 1.4273350129659173e-05,
+      "loss": 2.8984,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3790101486219999,
+      "grad_norm": 3.5352582821546705,
+      "learning_rate": 1.4267714916056465e-05,
+      "loss": 3.2361,
+      "step": 1970
+    },
+    {
+      "epoch": 0.3792025395603867,
+      "grad_norm": 3.542371954046535,
+      "learning_rate": 1.426207804492246e-05,
+      "loss": 3.1897,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3793949304987735,
+      "grad_norm": 3.383901964738278,
+      "learning_rate": 1.4256439518446456e-05,
+      "loss": 3.445,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3795873214371603,
+      "grad_norm": 3.325091184069247,
+      "learning_rate": 1.4250799338818388e-05,
+      "loss": 3.1005,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3797797123755471,
+      "grad_norm": 3.6552032116120823,
+      "learning_rate": 1.424515750822884e-05,
+      "loss": 3.2586,
+      "step": 1974
+    },
+    {
+      "epoch": 0.37997210331393394,
+      "grad_norm": 3.2142839188326557,
+      "learning_rate": 1.4239514028869032e-05,
+      "loss": 3.0621,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38016449425232074,
+      "grad_norm": 3.4408350373815066,
+      "learning_rate": 1.4233868902930827e-05,
+      "loss": 3.081,
+      "step": 1976
+    },
+    {
+      "epoch": 0.38035688519070754,
+      "grad_norm": 3.3576198834783706,
+      "learning_rate": 1.4228222132606729e-05,
+      "loss": 3.0037,
+      "step": 1977
+    },
+    {
+      "epoch": 0.38054927612909434,
+      "grad_norm": 3.4448784968577275,
+      "learning_rate": 1.4222573720089874e-05,
+      "loss": 3.134,
+      "step": 1978
+    },
+    {
+      "epoch": 0.38074166706748114,
+      "grad_norm": 3.4457588100059495,
+      "learning_rate": 1.4216923667574042e-05,
+      "loss": 3.1244,
+      "step": 1979
+    },
+    {
+      "epoch": 0.38093405800586794,
+      "grad_norm": 3.46550175734307,
+      "learning_rate": 1.4211271977253653e-05,
+      "loss": 2.9856,
+      "step": 1980
+    },
+    {
+      "epoch": 0.38112644894425474,
+      "grad_norm": 3.4815692292976768,
+      "learning_rate": 1.4205618651323753e-05,
+      "loss": 3.2535,
+      "step": 1981
+    },
+    {
+      "epoch": 0.38131883988264154,
+      "grad_norm": 3.5842085366175893,
+      "learning_rate": 1.4199963691980027e-05,
+      "loss": 3.3297,
+      "step": 1982
+    },
+    {
+      "epoch": 0.38151123082102834,
+      "grad_norm": 3.4803795302442384,
+      "learning_rate": 1.4194307101418805e-05,
+      "loss": 3.2283,
+      "step": 1983
+    },
+    {
+      "epoch": 0.38170362175941513,
+      "grad_norm": 3.3370158149893414,
+      "learning_rate": 1.4188648881837033e-05,
+      "loss": 2.9251,
+      "step": 1984
+    },
+    {
+      "epoch": 0.38189601269780193,
+      "grad_norm": 3.1990742929653755,
+      "learning_rate": 1.4182989035432299e-05,
+      "loss": 3.0395,
+      "step": 1985
+    },
+    {
+      "epoch": 0.38208840363618873,
+      "grad_norm": 3.576567082238796,
+      "learning_rate": 1.4177327564402825e-05,
+      "loss": 2.9913,
+      "step": 1986
+    },
+    {
+      "epoch": 0.38228079457457553,
+      "grad_norm": 3.629030753808809,
+      "learning_rate": 1.4171664470947464e-05,
+      "loss": 3.4605,
+      "step": 1987
+    },
+    {
+      "epoch": 0.38247318551296233,
+      "grad_norm": 3.3194546104137697,
+      "learning_rate": 1.416599975726569e-05,
+      "loss": 3.245,
+      "step": 1988
+    },
+    {
+      "epoch": 0.38266557645134913,
+      "grad_norm": 3.295231696177719,
+      "learning_rate": 1.4160333425557616e-05,
+      "loss": 2.9915,
+      "step": 1989
+    },
+    {
+      "epoch": 0.38285796738973593,
+      "grad_norm": 3.4056720161413003,
+      "learning_rate": 1.4154665478023977e-05,
+      "loss": 3.1688,
+      "step": 1990
+    },
+    {
+      "epoch": 0.38305035832812273,
+      "grad_norm": 3.329229369911821,
+      "learning_rate": 1.4148995916866139e-05,
+      "loss": 3.1234,
+      "step": 1991
+    },
+    {
+      "epoch": 0.38324274926650953,
+      "grad_norm": 3.4345116378122458,
+      "learning_rate": 1.41433247442861e-05,
+      "loss": 3.1724,
+      "step": 1992
+    },
+    {
+      "epoch": 0.38343514020489633,
+      "grad_norm": 3.365205646444566,
+      "learning_rate": 1.4137651962486472e-05,
+      "loss": 3.3042,
+      "step": 1993
+    },
+    {
+      "epoch": 0.38362753114328313,
+      "grad_norm": 3.466988457981494,
+      "learning_rate": 1.4131977573670499e-05,
+      "loss": 3.0782,
+      "step": 1994
+    },
+    {
+      "epoch": 0.38381992208166993,
+      "grad_norm": 3.3894069173440133,
+      "learning_rate": 1.412630158004205e-05,
+      "loss": 3.1649,
+      "step": 1995
+    },
+    {
+      "epoch": 0.38401231302005673,
+      "grad_norm": 3.6800234592834995,
+      "learning_rate": 1.4120623983805617e-05,
+      "loss": 3.2312,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3842047039584436,
+      "grad_norm": 3.7446980209806457,
+      "learning_rate": 1.4114944787166307e-05,
+      "loss": 3.2739,
+      "step": 1997
+    },
+    {
+      "epoch": 0.3843970948968304,
+      "grad_norm": 3.494617887054876,
+      "learning_rate": 1.4109263992329858e-05,
+      "loss": 2.7943,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3845894858352172,
+      "grad_norm": 3.4920686172652045,
+      "learning_rate": 1.4103581601502629e-05,
+      "loss": 3.2895,
+      "step": 1999
+    },
+    {
+      "epoch": 0.384781876773604,
+      "grad_norm": 3.2275087811823107,
+      "learning_rate": 1.409789761689159e-05,
+      "loss": 3.1916,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3849742677119908,
+      "grad_norm": 3.473259479585605,
+      "learning_rate": 1.4092212040704336e-05,
+      "loss": 3.1353,
+      "step": 2001
+    },
+    {
+      "epoch": 0.3851666586503776,
+      "grad_norm": 3.3819259200880114,
+      "learning_rate": 1.408652487514908e-05,
+      "loss": 3.1288,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3853590495887644,
+      "grad_norm": 3.3543365709544775,
+      "learning_rate": 1.408083612243465e-05,
+      "loss": 2.8883,
+      "step": 2003
+    },
+    {
+      "epoch": 0.3855514405271512,
+      "grad_norm": 3.494223901805367,
+      "learning_rate": 1.4075145784770496e-05,
+      "loss": 3.14,
+      "step": 2004
+    },
+    {
+      "epoch": 0.385743831465538,
+      "grad_norm": 3.462616735032719,
+      "learning_rate": 1.4069453864366678e-05,
+      "loss": 3.3076,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3859362224039248,
+      "grad_norm": 3.1951670100986225,
+      "learning_rate": 1.4063760363433867e-05,
+      "loss": 2.9827,
+      "step": 2006
+    },
+    {
+      "epoch": 0.3861286133423116,
+      "grad_norm": 2.946477127305788,
+      "learning_rate": 1.405806528418336e-05,
+      "loss": 2.813,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3863210042806984,
+      "grad_norm": 3.41886052960662,
+      "learning_rate": 1.4052368628827057e-05,
+      "loss": 2.9027,
+      "step": 2008
+    },
+    {
+      "epoch": 0.3865133952190852,
+      "grad_norm": 3.6958881702688005,
+      "learning_rate": 1.4046670399577478e-05,
+      "loss": 2.9395,
+      "step": 2009
+    },
+    {
+      "epoch": 0.386705786157472,
+      "grad_norm": 3.7593105040200787,
+      "learning_rate": 1.4040970598647742e-05,
+      "loss": 2.9874,
+      "step": 2010
+    },
+    {
+      "epoch": 0.3868981770958588,
+      "grad_norm": 3.2262112019896634,
+      "learning_rate": 1.4035269228251589e-05,
+      "loss": 3.0805,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3870905680342456,
+      "grad_norm": 2.2828862754811237,
+      "learning_rate": 1.4029566290603368e-05,
+      "loss": 2.601,
+      "step": 2012
+    },
+    {
+      "epoch": 0.3872829589726324,
+      "grad_norm": 3.4718640693461156,
+      "learning_rate": 1.4023861787918031e-05,
+      "loss": 3.1922,
+      "step": 2013
+    },
+    {
+      "epoch": 0.3874753499110192,
+      "grad_norm": 4.255777460606054,
+      "learning_rate": 1.4018155722411144e-05,
+      "loss": 3.1451,
+      "step": 2014
+    },
+    {
+      "epoch": 0.387667740849406,
+      "grad_norm": 3.364283583169558,
+      "learning_rate": 1.4012448096298874e-05,
+      "loss": 3.1952,
+      "step": 2015
+    },
+    {
+      "epoch": 0.3878601317877928,
+      "grad_norm": 3.6516363458404157,
+      "learning_rate": 1.4006738911798001e-05,
+      "loss": 3.279,
+      "step": 2016
+    },
+    {
+      "epoch": 0.38805252272617957,
+      "grad_norm": 3.1960354546793086,
+      "learning_rate": 1.40010281711259e-05,
+      "loss": 3.1947,
+      "step": 2017
+    },
+    {
+      "epoch": 0.38824491366456637,
+      "grad_norm": 3.6733554140846065,
+      "learning_rate": 1.3995315876500565e-05,
+      "loss": 3.2375,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3884373046029532,
+      "grad_norm": 3.574622775918724,
+      "learning_rate": 1.3989602030140581e-05,
+      "loss": 3.1636,
+      "step": 2019
+    },
+    {
+      "epoch": 0.38862969554134,
+      "grad_norm": 3.4114844415531516,
+      "learning_rate": 1.398388663426514e-05,
+      "loss": 3.25,
+      "step": 2020
+    },
+    {
+      "epoch": 0.3888220864797268,
+      "grad_norm": 2.706001750031994,
+      "learning_rate": 1.3978169691094037e-05,
+      "loss": 2.6583,
+      "step": 2021
+    },
+    {
+      "epoch": 0.3890144774181136,
+      "grad_norm": 3.868066934619549,
+      "learning_rate": 1.3972451202847665e-05,
+      "loss": 3.0252,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3892068683565004,
+      "grad_norm": 3.5972433226720457,
+      "learning_rate": 1.3966731171747024e-05,
+      "loss": 3.0242,
+      "step": 2023
+    },
+    {
+      "epoch": 0.3893992592948872,
+      "grad_norm": 3.975646980362214,
+      "learning_rate": 1.3961009600013702e-05,
+      "loss": 3.4566,
+      "step": 2024
+    },
+    {
+      "epoch": 0.389591650233274,
+      "grad_norm": 3.610079601753352,
+      "learning_rate": 1.3955286489869894e-05,
+      "loss": 3.3837,
+      "step": 2025
+    },
+    {
+      "epoch": 0.3897840411716608,
+      "grad_norm": 3.680583760458778,
+      "learning_rate": 1.394956184353839e-05,
+      "loss": 3.2423,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3899764321100476,
+      "grad_norm": 3.3657121616795167,
+      "learning_rate": 1.3943835663242577e-05,
+      "loss": 3.1055,
+      "step": 2027
+    },
+    {
+      "epoch": 0.3901688230484344,
+      "grad_norm": 3.348167302464716,
+      "learning_rate": 1.3938107951206438e-05,
+      "loss": 3.0315,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3903612139868212,
+      "grad_norm": 3.445372305459741,
+      "learning_rate": 1.3932378709654548e-05,
+      "loss": 3.2144,
+      "step": 2029
+    },
+    {
+      "epoch": 0.390553604925208,
+      "grad_norm": 3.723477045519266,
+      "learning_rate": 1.3926647940812081e-05,
+      "loss": 3.2013,
+      "step": 2030
+    },
+    {
+      "epoch": 0.3907459958635948,
+      "grad_norm": 3.218437650758881,
+      "learning_rate": 1.39209156469048e-05,
+      "loss": 2.8342,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3909383868019816,
+      "grad_norm": 2.3201283135937523,
+      "learning_rate": 1.3915181830159061e-05,
+      "loss": 2.6986,
+      "step": 2032
+    },
+    {
+      "epoch": 0.3911307777403684,
+      "grad_norm": 3.3487851294786015,
+      "learning_rate": 1.3909446492801819e-05,
+      "loss": 2.8674,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3913231686787552,
+      "grad_norm": 3.5458477662872157,
+      "learning_rate": 1.3903709637060605e-05,
+      "loss": 3.3159,
+      "step": 2034
+    },
+    {
+      "epoch": 0.391515559617142,
+      "grad_norm": 2.3162073187154277,
+      "learning_rate": 1.3897971265163546e-05,
+      "loss": 2.6252,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3917079505555288,
+      "grad_norm": 3.325398481291256,
+      "learning_rate": 1.3892231379339369e-05,
+      "loss": 3.1072,
+      "step": 2036
+    },
+    {
+      "epoch": 0.3919003414939156,
+      "grad_norm": 3.9287954469233046,
+      "learning_rate": 1.3886489981817375e-05,
+      "loss": 3.4273,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3920927324323024,
+      "grad_norm": 2.529984907360552,
+      "learning_rate": 1.3880747074827454e-05,
+      "loss": 2.6837,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3922851233706892,
+      "grad_norm": 3.4536969827265525,
+      "learning_rate": 1.3875002660600085e-05,
+      "loss": 3.3914,
+      "step": 2039
+    },
+    {
+      "epoch": 0.39247751430907607,
+      "grad_norm": 2.282218066971841,
+      "learning_rate": 1.386925674136634e-05,
+      "loss": 2.5789,
+      "step": 2040
+    },
+    {
+      "epoch": 0.39266990524746287,
+      "grad_norm": 3.657757302192482,
+      "learning_rate": 1.3863509319357857e-05,
+      "loss": 3.2303,
+      "step": 2041
+    },
+    {
+      "epoch": 0.39286229618584967,
+      "grad_norm": 3.4200364365183193,
+      "learning_rate": 1.3857760396806876e-05,
+      "loss": 3.3962,
+      "step": 2042
+    },
+    {
+      "epoch": 0.39305468712423647,
+      "grad_norm": 3.588882376907582,
+      "learning_rate": 1.3852009975946209e-05,
+      "loss": 2.9492,
+      "step": 2043
+    },
+    {
+      "epoch": 0.39324707806262327,
+      "grad_norm": 3.5153603543207512,
+      "learning_rate": 1.3846258059009252e-05,
+      "loss": 3.072,
+      "step": 2044
+    },
+    {
+      "epoch": 0.39343946900101007,
+      "grad_norm": 3.29939542922612,
+      "learning_rate": 1.384050464822999e-05,
+      "loss": 3.3665,
+      "step": 2045
+    },
+    {
+      "epoch": 0.39363185993939687,
+      "grad_norm": 3.3246859640742215,
+      "learning_rate": 1.383474974584297e-05,
+      "loss": 3.2605,
+      "step": 2046
+    },
+    {
+      "epoch": 0.39382425087778367,
+      "grad_norm": 3.3725476848394043,
+      "learning_rate": 1.3828993354083342e-05,
+      "loss": 2.9846,
+      "step": 2047
+    },
+    {
+      "epoch": 0.39401664181617047,
+      "grad_norm": 3.783500686429089,
+      "learning_rate": 1.3823235475186816e-05,
+      "loss": 3.2835,
+      "step": 2048
+    },
+    {
+      "epoch": 0.39420903275455726,
+      "grad_norm": 3.2374396441471567,
+      "learning_rate": 1.3817476111389685e-05,
+      "loss": 2.991,
+      "step": 2049
+    },
+    {
+      "epoch": 0.39440142369294406,
+      "grad_norm": 3.241896388038147,
+      "learning_rate": 1.3811715264928824e-05,
+      "loss": 3.1014,
+      "step": 2050
+    },
+    {
+      "epoch": 0.39459381463133086,
+      "grad_norm": 3.521453318200088,
+      "learning_rate": 1.3805952938041674e-05,
+      "loss": 3.2219,
+      "step": 2051
+    },
+    {
+      "epoch": 0.39478620556971766,
+      "grad_norm": 3.636668691359571,
+      "learning_rate": 1.3800189132966257e-05,
+      "loss": 2.991,
+      "step": 2052
+    },
+    {
+      "epoch": 0.39497859650810446,
+      "grad_norm": 2.407814342705515,
+      "learning_rate": 1.3794423851941174e-05,
+      "loss": 2.6809,
+      "step": 2053
+    },
+    {
+      "epoch": 0.39517098744649126,
+      "grad_norm": 3.565157789818108,
+      "learning_rate": 1.378865709720559e-05,
+      "loss": 3.0142,
+      "step": 2054
+    },
+    {
+      "epoch": 0.39536337838487806,
+      "grad_norm": 3.5478236003010215,
+      "learning_rate": 1.3782888870999245e-05,
+      "loss": 3.0513,
+      "step": 2055
+    },
+    {
+      "epoch": 0.39555576932326486,
+      "grad_norm": 3.5542535447158263,
+      "learning_rate": 1.377711917556245e-05,
+      "loss": 2.9382,
+      "step": 2056
+    },
+    {
+      "epoch": 0.39574816026165166,
+      "grad_norm": 3.596817502020254,
+      "learning_rate": 1.3771348013136096e-05,
+      "loss": 3.1664,
+      "step": 2057
+    },
+    {
+      "epoch": 0.39594055120003846,
+      "grad_norm": 3.6031953831768715,
+      "learning_rate": 1.3765575385961627e-05,
+      "loss": 3.3687,
+      "step": 2058
+    },
+    {
+      "epoch": 0.39613294213842526,
+      "grad_norm": 3.4386425168416306,
+      "learning_rate": 1.3759801296281072e-05,
+      "loss": 3.0718,
+      "step": 2059
+    },
+    {
+      "epoch": 0.39632533307681206,
+      "grad_norm": 3.6191161396339098,
+      "learning_rate": 1.3754025746337014e-05,
+      "loss": 2.9788,
+      "step": 2060
+    },
+    {
+      "epoch": 0.39651772401519886,
+      "grad_norm": 3.606307815009061,
+      "learning_rate": 1.3748248738372616e-05,
+      "loss": 3.1689,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3967101149535857,
+      "grad_norm": 3.6823179717809817,
+      "learning_rate": 1.3742470274631599e-05,
+      "loss": 2.9251,
+      "step": 2062
+    },
+    {
+      "epoch": 0.3969025058919725,
+      "grad_norm": 3.4310077418925076,
+      "learning_rate": 1.3736690357358253e-05,
+      "loss": 3.4945,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3970948968303593,
+      "grad_norm": 3.6778410698724477,
+      "learning_rate": 1.3730908988797427e-05,
+      "loss": 2.9861,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3972872877687461,
+      "grad_norm": 3.6141897791914848,
+      "learning_rate": 1.3725126171194543e-05,
+      "loss": 3.1943,
+      "step": 2065
+    },
+    {
+      "epoch": 0.3974796787071329,
+      "grad_norm": 3.437957847805881,
+      "learning_rate": 1.371934190679558e-05,
+      "loss": 3.1495,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3976720696455197,
+      "grad_norm": 3.6275867172945198,
+      "learning_rate": 1.3713556197847076e-05,
+      "loss": 3.2752,
+      "step": 2067
+    },
+    {
+      "epoch": 0.3978644605839065,
+      "grad_norm": 3.7338046358494545,
+      "learning_rate": 1.3707769046596136e-05,
+      "loss": 3.2455,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3980568515222933,
+      "grad_norm": 3.5815535267287983,
+      "learning_rate": 1.3701980455290425e-05,
+      "loss": 3.0461,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3982492424606801,
+      "grad_norm": 3.1509028337122857,
+      "learning_rate": 1.3696190426178162e-05,
+      "loss": 2.8846,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3984416333990669,
+      "grad_norm": 3.6691334715906354,
+      "learning_rate": 1.3690398961508128e-05,
+      "loss": 2.9658,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3986340243374537,
+      "grad_norm": 3.3229197263004946,
+      "learning_rate": 1.3684606063529662e-05,
+      "loss": 3.0518,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3988264152758405,
+      "grad_norm": 3.6699597778416417,
+      "learning_rate": 1.3678811734492659e-05,
+      "loss": 2.9805,
+      "step": 2073
+    },
+    {
+      "epoch": 0.3990188062142273,
+      "grad_norm": 3.5864591039605664,
+      "learning_rate": 1.367301597664757e-05,
+      "loss": 3.1029,
+      "step": 2074
+    },
+    {
+      "epoch": 0.3992111971526141,
+      "grad_norm": 3.5857064236870526,
+      "learning_rate": 1.36672187922454e-05,
+      "loss": 3.155,
+      "step": 2075
+    },
+    {
+      "epoch": 0.3994035880910009,
+      "grad_norm": 3.5484741864338303,
+      "learning_rate": 1.3661420183537705e-05,
+      "loss": 3.2023,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3995959790293877,
+      "grad_norm": 3.4000442123209176,
+      "learning_rate": 1.3655620152776605e-05,
+      "loss": 3.2501,
+      "step": 2077
+    },
+    {
+      "epoch": 0.3997883699677745,
+      "grad_norm": 2.670441980090611,
+      "learning_rate": 1.364981870221476e-05,
+      "loss": 2.8778,
+      "step": 2078
+    },
+    {
+      "epoch": 0.3999807609061613,
+      "grad_norm": 3.604031564318976,
+      "learning_rate": 1.364401583410539e-05,
+      "loss": 3.0069,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4001731518445481,
+      "grad_norm": 3.715232848518105,
+      "learning_rate": 1.3638211550702256e-05,
+      "loss": 3.1529,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4003655427829349,
+      "grad_norm": 3.5190740256674893,
+      "learning_rate": 1.363240585425968e-05,
+      "loss": 3.1355,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4005579337213217,
+      "grad_norm": 3.7124642290633707,
+      "learning_rate": 1.362659874703253e-05,
+      "loss": 3.1836,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4007503246597085,
+      "grad_norm": 3.42144444492785,
+      "learning_rate": 1.3620790231276213e-05,
+      "loss": 3.3576,
+      "step": 2083
+    },
+    {
+      "epoch": 0.40094271559809536,
+      "grad_norm": 3.383334391680382,
+      "learning_rate": 1.3614980309246692e-05,
+      "loss": 3.3091,
+      "step": 2084
+    },
+    {
+      "epoch": 0.40113510653648216,
+      "grad_norm": 3.5753980624342003,
+      "learning_rate": 1.3609168983200474e-05,
+      "loss": 3.1538,
+      "step": 2085
+    },
+    {
+      "epoch": 0.40132749747486895,
+      "grad_norm": 3.5233109138151932,
+      "learning_rate": 1.3603356255394613e-05,
+      "loss": 3.3091,
+      "step": 2086
+    },
+    {
+      "epoch": 0.40151988841325575,
+      "grad_norm": 3.5174920558660423,
+      "learning_rate": 1.3597542128086702e-05,
+      "loss": 3.1997,
+      "step": 2087
+    },
+    {
+      "epoch": 0.40171227935164255,
+      "grad_norm": 3.3166862532727728,
+      "learning_rate": 1.3591726603534885e-05,
+      "loss": 3.1905,
+      "step": 2088
+    },
+    {
+      "epoch": 0.40190467029002935,
+      "grad_norm": 3.812945730233859,
+      "learning_rate": 1.3585909683997842e-05,
+      "loss": 3.0135,
+      "step": 2089
+    },
+    {
+      "epoch": 0.40209706122841615,
+      "grad_norm": 3.4321953891999732,
+      "learning_rate": 1.3580091371734798e-05,
+      "loss": 3.2064,
+      "step": 2090
+    },
+    {
+      "epoch": 0.40228945216680295,
+      "grad_norm": 3.5086358123750916,
+      "learning_rate": 1.357427166900552e-05,
+      "loss": 3.3759,
+      "step": 2091
+    },
+    {
+      "epoch": 0.40248184310518975,
+      "grad_norm": 3.2129766910243966,
+      "learning_rate": 1.3568450578070309e-05,
+      "loss": 2.9111,
+      "step": 2092
+    },
+    {
+      "epoch": 0.40267423404357655,
+      "grad_norm": 3.6154381137603675,
+      "learning_rate": 1.3562628101190015e-05,
+      "loss": 3.4368,
+      "step": 2093
+    },
+    {
+      "epoch": 0.40286662498196335,
+      "grad_norm": 3.217431954097432,
+      "learning_rate": 1.3556804240626019e-05,
+      "loss": 3.0705,
+      "step": 2094
+    },
+    {
+      "epoch": 0.40305901592035015,
+      "grad_norm": 3.4767077300394926,
+      "learning_rate": 1.3550978998640241e-05,
+      "loss": 3.2318,
+      "step": 2095
+    },
+    {
+      "epoch": 0.40325140685873695,
+      "grad_norm": 3.2134339181869285,
+      "learning_rate": 1.3545152377495136e-05,
+      "loss": 2.8509,
+      "step": 2096
+    },
+    {
+      "epoch": 0.40344379779712375,
+      "grad_norm": 3.3555191890007023,
+      "learning_rate": 1.3539324379453698e-05,
+      "loss": 3.2456,
+      "step": 2097
+    },
+    {
+      "epoch": 0.40363618873551055,
+      "grad_norm": 3.3434008727222637,
+      "learning_rate": 1.3533495006779455e-05,
+      "loss": 2.9463,
+      "step": 2098
+    },
+    {
+      "epoch": 0.40382857967389735,
+      "grad_norm": 3.175782714725847,
+      "learning_rate": 1.3527664261736471e-05,
+      "loss": 2.9253,
+      "step": 2099
+    },
+    {
+      "epoch": 0.40402097061228415,
+      "grad_norm": 3.5307040509699603,
+      "learning_rate": 1.3521832146589335e-05,
+      "loss": 3.0072,
+      "step": 2100
+    },
+    {
+      "epoch": 0.40421336155067095,
+      "grad_norm": 3.3586285951392125,
+      "learning_rate": 1.3515998663603174e-05,
+      "loss": 3.2923,
+      "step": 2101
+    },
+    {
+      "epoch": 0.40440575248905775,
+      "grad_norm": 3.772258634392537,
+      "learning_rate": 1.3510163815043647e-05,
+      "loss": 3.0047,
+      "step": 2102
+    },
+    {
+      "epoch": 0.40459814342744455,
+      "grad_norm": 3.2767255648319518,
+      "learning_rate": 1.3504327603176943e-05,
+      "loss": 2.908,
+      "step": 2103
+    },
+    {
+      "epoch": 0.40479053436583134,
+      "grad_norm": 3.366091746533064,
+      "learning_rate": 1.3498490030269782e-05,
+      "loss": 3.2655,
+      "step": 2104
+    },
+    {
+      "epoch": 0.40498292530421814,
+      "grad_norm": 3.659413655257838,
+      "learning_rate": 1.3492651098589398e-05,
+      "loss": 3.0115,
+      "step": 2105
+    },
+    {
+      "epoch": 0.405175316242605,
+      "grad_norm": 3.397027945073309,
+      "learning_rate": 1.3486810810403578e-05,
+      "loss": 3.1731,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4053677071809918,
+      "grad_norm": 3.3617917947619436,
+      "learning_rate": 1.348096916798062e-05,
+      "loss": 2.8361,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4055600981193786,
+      "grad_norm": 3.3669598901031796,
+      "learning_rate": 1.3475126173589343e-05,
+      "loss": 2.8503,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4057524890577654,
+      "grad_norm": 3.3491919047643237,
+      "learning_rate": 1.3469281829499107e-05,
+      "loss": 3.1276,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4059448799961522,
+      "grad_norm": 3.525462671398618,
+      "learning_rate": 1.3463436137979786e-05,
+      "loss": 3.1803,
+      "step": 2110
+    },
+    {
+      "epoch": 0.406137270934539,
+      "grad_norm": 2.281559662269352,
+      "learning_rate": 1.3457589101301776e-05,
+      "loss": 2.5778,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4063296618729258,
+      "grad_norm": 3.544440551437748,
+      "learning_rate": 1.3451740721736005e-05,
+      "loss": 2.9977,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4065220528113126,
+      "grad_norm": 4.033936231355366,
+      "learning_rate": 1.3445891001553905e-05,
+      "loss": 3.4171,
+      "step": 2113
+    },
+    {
+      "epoch": 0.4067144437496994,
+      "grad_norm": 2.5059310364127976,
+      "learning_rate": 1.3440039943027452e-05,
+      "loss": 2.6453,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4069068346880862,
+      "grad_norm": 3.6864246381295884,
+      "learning_rate": 1.3434187548429126e-05,
+      "loss": 3.1938,
+      "step": 2115
+    },
+    {
+      "epoch": 0.407099225626473,
+      "grad_norm": 3.4408774510362177,
+      "learning_rate": 1.3428333820031922e-05,
+      "loss": 3.3295,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4072916165648598,
+      "grad_norm": 3.301419375322878,
+      "learning_rate": 1.3422478760109371e-05,
+      "loss": 3.0778,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4074840075032466,
+      "grad_norm": 3.382610750350635,
+      "learning_rate": 1.3416622370935507e-05,
+      "loss": 3.2979,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4076763984416334,
+      "grad_norm": 3.7535183075770515,
+      "learning_rate": 1.3410764654784885e-05,
+      "loss": 3.1641,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4078687893800202,
+      "grad_norm": 3.159403596359466,
+      "learning_rate": 1.3404905613932573e-05,
+      "loss": 2.9908,
+      "step": 2120
+    },
+    {
+      "epoch": 0.408061180318407,
+      "grad_norm": 3.3362053068301867,
+      "learning_rate": 1.3399045250654152e-05,
+      "loss": 3.216,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4082535712567938,
+      "grad_norm": 3.584904926702761,
+      "learning_rate": 1.3393183567225724e-05,
+      "loss": 3.171,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4084459621951806,
+      "grad_norm": 3.283588778097424,
+      "learning_rate": 1.3387320565923901e-05,
+      "loss": 3.1701,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4086383531335674,
+      "grad_norm": 3.313447663566547,
+      "learning_rate": 1.33814562490258e-05,
+      "loss": 3.2143,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4088307440719542,
+      "grad_norm": 3.338415532082853,
+      "learning_rate": 1.3375590618809056e-05,
+      "loss": 3.2875,
+      "step": 2125
+    },
+    {
+      "epoch": 0.409023135010341,
+      "grad_norm": 3.6294112034644166,
+      "learning_rate": 1.3369723677551813e-05,
+      "loss": 2.942,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4092155259487278,
+      "grad_norm": 3.253169961570375,
+      "learning_rate": 1.3363855427532724e-05,
+      "loss": 3.0601,
+      "step": 2127
+    },
+    {
+      "epoch": 0.40940791688711464,
+      "grad_norm": 3.6627869961968105,
+      "learning_rate": 1.3357985871030948e-05,
+      "loss": 3.6689,
+      "step": 2128
+    },
+    {
+      "epoch": 0.40960030782550144,
+      "grad_norm": 3.2381656281400693,
+      "learning_rate": 1.3352115010326155e-05,
+      "loss": 3.3999,
+      "step": 2129
+    },
+    {
+      "epoch": 0.40979269876388824,
+      "grad_norm": 3.153344735477907,
+      "learning_rate": 1.3346242847698516e-05,
+      "loss": 3.1017,
+      "step": 2130
+    },
+    {
+      "epoch": 0.40998508970227504,
+      "grad_norm": 3.373502613422817,
+      "learning_rate": 1.3340369385428713e-05,
+      "loss": 3.2074,
+      "step": 2131
+    },
+    {
+      "epoch": 0.41017748064066184,
+      "grad_norm": 3.2647308662746295,
+      "learning_rate": 1.3334494625797936e-05,
+      "loss": 3.2013,
+      "step": 2132
+    },
+    {
+      "epoch": 0.41036987157904864,
+      "grad_norm": 3.3097249048369064,
+      "learning_rate": 1.3328618571087867e-05,
+      "loss": 3.0639,
+      "step": 2133
+    },
+    {
+      "epoch": 0.41056226251743544,
+      "grad_norm": 3.1465412776458184,
+      "learning_rate": 1.33227412235807e-05,
+      "loss": 3.2631,
+      "step": 2134
+    },
+    {
+      "epoch": 0.41075465345582224,
+      "grad_norm": 3.571520275115539,
+      "learning_rate": 1.3316862585559132e-05,
+      "loss": 3.011,
+      "step": 2135
+    },
+    {
+      "epoch": 0.41094704439420904,
+      "grad_norm": 3.416299880803797,
+      "learning_rate": 1.3310982659306352e-05,
+      "loss": 3.12,
+      "step": 2136
+    },
+    {
+      "epoch": 0.41113943533259584,
+      "grad_norm": 3.3481630302866234,
+      "learning_rate": 1.3305101447106064e-05,
+      "loss": 3.2088,
+      "step": 2137
+    },
+    {
+      "epoch": 0.41133182627098264,
+      "grad_norm": 3.41155329245002,
+      "learning_rate": 1.3299218951242456e-05,
+      "loss": 3.3765,
+      "step": 2138
+    },
+    {
+      "epoch": 0.41152421720936944,
+      "grad_norm": 3.325075150560553,
+      "learning_rate": 1.3293335174000226e-05,
+      "loss": 3.1402,
+      "step": 2139
+    },
+    {
+      "epoch": 0.41171660814775624,
+      "grad_norm": 3.2621503728225525,
+      "learning_rate": 1.328745011766456e-05,
+      "loss": 3.2066,
+      "step": 2140
+    },
+    {
+      "epoch": 0.41190899908614304,
+      "grad_norm": 3.350251730000048,
+      "learning_rate": 1.3281563784521154e-05,
+      "loss": 3.0963,
+      "step": 2141
+    },
+    {
+      "epoch": 0.41210139002452983,
+      "grad_norm": 2.293057349740328,
+      "learning_rate": 1.3275676176856185e-05,
+      "loss": 2.6446,
+      "step": 2142
+    },
+    {
+      "epoch": 0.41229378096291663,
+      "grad_norm": 3.3168279091745188,
+      "learning_rate": 1.3269787296956333e-05,
+      "loss": 3.1678,
+      "step": 2143
+    },
+    {
+      "epoch": 0.41248617190130343,
+      "grad_norm": 3.3851103475911617,
+      "learning_rate": 1.3263897147108778e-05,
+      "loss": 3.2709,
+      "step": 2144
+    },
+    {
+      "epoch": 0.41267856283969023,
+      "grad_norm": 3.4217002717981164,
+      "learning_rate": 1.3258005729601178e-05,
+      "loss": 2.9665,
+      "step": 2145
+    },
+    {
+      "epoch": 0.41287095377807703,
+      "grad_norm": 3.504747043291732,
+      "learning_rate": 1.3252113046721692e-05,
+      "loss": 3.072,
+      "step": 2146
+    },
+    {
+      "epoch": 0.41306334471646383,
+      "grad_norm": 3.5055439703555855,
+      "learning_rate": 1.3246219100758974e-05,
+      "loss": 3.2175,
+      "step": 2147
+    },
+    {
+      "epoch": 0.41325573565485063,
+      "grad_norm": 3.4805052407156545,
+      "learning_rate": 1.3240323894002166e-05,
+      "loss": 3.0189,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4134481265932375,
+      "grad_norm": 3.367569080394226,
+      "learning_rate": 1.3234427428740895e-05,
+      "loss": 3.2973,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4136405175316243,
+      "grad_norm": 3.5370819231161352,
+      "learning_rate": 1.3228529707265279e-05,
+      "loss": 3.3138,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4138329084700111,
+      "grad_norm": 3.4696544745333338,
+      "learning_rate": 1.322263073186593e-05,
+      "loss": 3.1721,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4140252994083979,
+      "grad_norm": 3.382810265616876,
+      "learning_rate": 1.3216730504833938e-05,
+      "loss": 3.0354,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4142176903467847,
+      "grad_norm": 3.3307785788226707,
+      "learning_rate": 1.3210829028460883e-05,
+      "loss": 3.0956,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4144100812851715,
+      "grad_norm": 2.4919353324936355,
+      "learning_rate": 1.3204926305038832e-05,
+      "loss": 2.6992,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4146024722235583,
+      "grad_norm": 3.582896648854483,
+      "learning_rate": 1.3199022336860335e-05,
+      "loss": 3.0701,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4147948631619451,
+      "grad_norm": 3.119245626658648,
+      "learning_rate": 1.3193117126218425e-05,
+      "loss": 3.1005,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4149872541003319,
+      "grad_norm": 3.50736599889488,
+      "learning_rate": 1.3187210675406617e-05,
+      "loss": 3.3544,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4151796450387187,
+      "grad_norm": 3.1062029107951554,
+      "learning_rate": 1.318130298671891e-05,
+      "loss": 3.0886,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4153720359771055,
+      "grad_norm": 3.4418237872153536,
+      "learning_rate": 1.3175394062449777e-05,
+      "loss": 3.1805,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4155644269154923,
+      "grad_norm": 3.7875888496706454,
+      "learning_rate": 1.3169483904894185e-05,
+      "loss": 3.3641,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4157568178538791,
+      "grad_norm": 3.2481983134740156,
+      "learning_rate": 1.3163572516347565e-05,
+      "loss": 2.9743,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4159492087922659,
+      "grad_norm": 3.5017089498060714,
+      "learning_rate": 1.3157659899105835e-05,
+      "loss": 3.1225,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4161415997306527,
+      "grad_norm": 3.6906708682772154,
+      "learning_rate": 1.315174605546538e-05,
+      "loss": 3.1757,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4163339906690395,
+      "grad_norm": 3.380838066959464,
+      "learning_rate": 1.3145830987723081e-05,
+      "loss": 3.0739,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4165263816074263,
+      "grad_norm": 2.5916040120696797,
+      "learning_rate": 1.3139914698176273e-05,
+      "loss": 2.6721,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4167187725458131,
+      "grad_norm": 3.676125037051297,
+      "learning_rate": 1.3133997189122777e-05,
+      "loss": 3.0806,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4169111634841999,
+      "grad_norm": 3.5246132717658862,
+      "learning_rate": 1.3128078462860887e-05,
+      "loss": 3.1465,
+      "step": 2167
+    },
+    {
+      "epoch": 0.4171035544225867,
+      "grad_norm": 3.4394300141318555,
+      "learning_rate": 1.3122158521689367e-05,
+      "loss": 2.955,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4172959453609735,
+      "grad_norm": 3.4840318767745915,
+      "learning_rate": 1.3116237367907454e-05,
+      "loss": 3.28,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4174883362993603,
+      "grad_norm": 3.5876986795288097,
+      "learning_rate": 1.3110315003814855e-05,
+      "loss": 3.3447,
+      "step": 2170
+    },
+    {
+      "epoch": 0.41768072723774713,
+      "grad_norm": 3.6382992570449266,
+      "learning_rate": 1.3104391431711748e-05,
+      "loss": 3.0883,
+      "step": 2171
+    },
+    {
+      "epoch": 0.41787311817613393,
+      "grad_norm": 3.4155742177836226,
+      "learning_rate": 1.309846665389878e-05,
+      "loss": 3.0377,
+      "step": 2172
+    },
+    {
+      "epoch": 0.41806550911452073,
+      "grad_norm": 3.2365796486640757,
+      "learning_rate": 1.309254067267707e-05,
+      "loss": 3.1024,
+      "step": 2173
+    },
+    {
+      "epoch": 0.4182579000529075,
+      "grad_norm": 3.3507660287158236,
+      "learning_rate": 1.3086613490348198e-05,
+      "loss": 3.1165,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4184502909912943,
+      "grad_norm": 3.6325203191701476,
+      "learning_rate": 1.3080685109214208e-05,
+      "loss": 3.2824,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4186426819296811,
+      "grad_norm": 3.48768683989848,
+      "learning_rate": 1.3074755531577628e-05,
+      "loss": 2.9523,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4188350728680679,
+      "grad_norm": 3.431078789054427,
+      "learning_rate": 1.3068824759741428e-05,
+      "loss": 3.3753,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4190274638064547,
+      "grad_norm": 3.7677359119059752,
+      "learning_rate": 1.306289279600905e-05,
+      "loss": 3.1019,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4192198547448415,
+      "grad_norm": 3.6511358186964737,
+      "learning_rate": 1.3056959642684404e-05,
+      "loss": 3.182,
+      "step": 2179
+    },
+    {
+      "epoch": 0.4194122456832283,
+      "grad_norm": 3.4353680275058647,
+      "learning_rate": 1.305102530207186e-05,
+      "loss": 3.2593,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4196046366216151,
+      "grad_norm": 3.302870120059589,
+      "learning_rate": 1.3045089776476246e-05,
+      "loss": 2.8391,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4197970275600019,
+      "grad_norm": 3.28017619107968,
+      "learning_rate": 1.3039153068202853e-05,
+      "loss": 3.2379,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4199894184983887,
+      "grad_norm": 3.361234201384691,
+      "learning_rate": 1.3033215179557424e-05,
+      "loss": 3.1706,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4201818094367755,
+      "grad_norm": 3.4435373526755253,
+      "learning_rate": 1.3027276112846172e-05,
+      "loss": 2.8805,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4203742003751623,
+      "grad_norm": 3.4255750602431516,
+      "learning_rate": 1.3021335870375763e-05,
+      "loss": 3.0988,
+      "step": 2185
+    },
+    {
+      "epoch": 0.4205665913135491,
+      "grad_norm": 3.5489795010537515,
+      "learning_rate": 1.3015394454453316e-05,
+      "loss": 3.2057,
+      "step": 2186
+    },
+    {
+      "epoch": 0.4207589822519359,
+      "grad_norm": 3.4277442155481195,
+      "learning_rate": 1.3009451867386411e-05,
+      "loss": 3.1508,
+      "step": 2187
+    },
+    {
+      "epoch": 0.4209513731903227,
+      "grad_norm": 3.3854894327391567,
+      "learning_rate": 1.3003508111483077e-05,
+      "loss": 3.3408,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4211437641287095,
+      "grad_norm": 3.433120957489098,
+      "learning_rate": 1.29975631890518e-05,
+      "loss": 3.286,
+      "step": 2189
+    },
+    {
+      "epoch": 0.4213361550670963,
+      "grad_norm": 2.312814113351686,
+      "learning_rate": 1.2991617102401524e-05,
+      "loss": 2.5232,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4215285460054831,
+      "grad_norm": 3.2830124846678435,
+      "learning_rate": 1.2985669853841635e-05,
+      "loss": 3.2032,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4217209369438699,
+      "grad_norm": 3.3727631567742606,
+      "learning_rate": 1.297972144568198e-05,
+      "loss": 3.1739,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4219133278822568,
+      "grad_norm": 3.2009516435986765,
+      "learning_rate": 1.2973771880232853e-05,
+      "loss": 3.0987,
+      "step": 2193
+    },
+    {
+      "epoch": 0.42210571882064357,
+      "grad_norm": 3.3352512404156194,
+      "learning_rate": 1.2967821159804994e-05,
+      "loss": 3.229,
+      "step": 2194
+    },
+    {
+      "epoch": 0.42229810975903037,
+      "grad_norm": 3.294408057665286,
+      "learning_rate": 1.2961869286709594e-05,
+      "loss": 3.228,
+      "step": 2195
+    },
+    {
+      "epoch": 0.42249050069741717,
+      "grad_norm": 3.342004506325307,
+      "learning_rate": 1.295591626325829e-05,
+      "loss": 3.2083,
+      "step": 2196
+    },
+    {
+      "epoch": 0.42268289163580397,
+      "grad_norm": 3.5415658331197304,
+      "learning_rate": 1.2949962091763174e-05,
+      "loss": 2.9707,
+      "step": 2197
+    },
+    {
+      "epoch": 0.42287528257419077,
+      "grad_norm": 3.4435923217949047,
+      "learning_rate": 1.2944006774536773e-05,
+      "loss": 3.1345,
+      "step": 2198
+    },
+    {
+      "epoch": 0.42306767351257757,
+      "grad_norm": 3.39665513977399,
+      "learning_rate": 1.2938050313892062e-05,
+      "loss": 3.1419,
+      "step": 2199
+    },
+    {
+      "epoch": 0.42326006445096437,
+      "grad_norm": 3.3817730525482794,
+      "learning_rate": 1.2932092712142468e-05,
+      "loss": 2.8812,
+      "step": 2200
+    },
+    {
+      "epoch": 0.42345245538935117,
+      "grad_norm": 3.2952866314231803,
+      "learning_rate": 1.292613397160185e-05,
+      "loss": 3.1436,
+      "step": 2201
+    },
+    {
+      "epoch": 0.42364484632773797,
+      "grad_norm": 3.4926243582941248,
+      "learning_rate": 1.2920174094584514e-05,
+      "loss": 3.0118,
+      "step": 2202
+    },
+    {
+      "epoch": 0.42383723726612477,
+      "grad_norm": 3.5165115218235017,
+      "learning_rate": 1.2914213083405211e-05,
+      "loss": 3.3614,
+      "step": 2203
+    },
+    {
+      "epoch": 0.42402962820451157,
+      "grad_norm": 3.426954081321928,
+      "learning_rate": 1.2908250940379124e-05,
+      "loss": 2.8381,
+      "step": 2204
+    },
+    {
+      "epoch": 0.42422201914289837,
+      "grad_norm": 3.7136310354448825,
+      "learning_rate": 1.2902287667821885e-05,
+      "loss": 3.2615,
+      "step": 2205
+    },
+    {
+      "epoch": 0.42441441008128516,
+      "grad_norm": 3.5057239041893253,
+      "learning_rate": 1.289632326804956e-05,
+      "loss": 2.9994,
+      "step": 2206
+    },
+    {
+      "epoch": 0.42460680101967196,
+      "grad_norm": 3.475580131277571,
+      "learning_rate": 1.2890357743378649e-05,
+      "loss": 3.167,
+      "step": 2207
+    },
+    {
+      "epoch": 0.42479919195805876,
+      "grad_norm": 3.612075633173737,
+      "learning_rate": 1.2884391096126098e-05,
+      "loss": 3.1644,
+      "step": 2208
+    },
+    {
+      "epoch": 0.42499158289644556,
+      "grad_norm": 3.524275110221941,
+      "learning_rate": 1.2878423328609281e-05,
+      "loss": 3.2366,
+      "step": 2209
+    },
+    {
+      "epoch": 0.42518397383483236,
+      "grad_norm": 3.4612298399186305,
+      "learning_rate": 1.2872454443146015e-05,
+      "loss": 3.1864,
+      "step": 2210
+    },
+    {
+      "epoch": 0.42537636477321916,
+      "grad_norm": 3.5810322998326907,
+      "learning_rate": 1.286648444205454e-05,
+      "loss": 3.0439,
+      "step": 2211
+    },
+    {
+      "epoch": 0.42556875571160596,
+      "grad_norm": 3.4500567855735427,
+      "learning_rate": 1.2860513327653537e-05,
+      "loss": 3.2014,
+      "step": 2212
+    },
+    {
+      "epoch": 0.42576114664999276,
+      "grad_norm": 3.371766769369605,
+      "learning_rate": 1.2854541102262119e-05,
+      "loss": 3.3139,
+      "step": 2213
+    },
+    {
+      "epoch": 0.42595353758837956,
+      "grad_norm": 3.473873304230959,
+      "learning_rate": 1.284856776819983e-05,
+      "loss": 3.2011,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4261459285267664,
+      "grad_norm": 3.3733197817350615,
+      "learning_rate": 1.2842593327786649e-05,
+      "loss": 3.0956,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4263383194651532,
+      "grad_norm": 3.2630124398266935,
+      "learning_rate": 1.2836617783342968e-05,
+      "loss": 3.1741,
+      "step": 2216
+    },
+    {
+      "epoch": 0.42653071040354,
+      "grad_norm": 3.2139143225321485,
+      "learning_rate": 1.2830641137189628e-05,
+      "loss": 3.1624,
+      "step": 2217
+    },
+    {
+      "epoch": 0.4267231013419268,
+      "grad_norm": 3.5749302703108317,
+      "learning_rate": 1.282466339164789e-05,
+      "loss": 3.2098,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4269154922803136,
+      "grad_norm": 3.7137269140137388,
+      "learning_rate": 1.2818684549039437e-05,
+      "loss": 3.1501,
+      "step": 2219
+    },
+    {
+      "epoch": 0.4271078832187004,
+      "grad_norm": 3.535741572228054,
+      "learning_rate": 1.2812704611686386e-05,
+      "loss": 3.3057,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4273002741570872,
+      "grad_norm": 3.4006929371739676,
+      "learning_rate": 1.2806723581911274e-05,
+      "loss": 3.155,
+      "step": 2221
+    },
+    {
+      "epoch": 0.427492665095474,
+      "grad_norm": 3.3881278195539517,
+      "learning_rate": 1.2800741462037065e-05,
+      "loss": 2.99,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4276850560338608,
+      "grad_norm": 3.383567910612626,
+      "learning_rate": 1.2794758254387147e-05,
+      "loss": 3.1397,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4278774469722476,
+      "grad_norm": 3.396944652021193,
+      "learning_rate": 1.2788773961285323e-05,
+      "loss": 3.0118,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4280698379106344,
+      "grad_norm": 3.696267804506679,
+      "learning_rate": 1.2782788585055829e-05,
+      "loss": 3.1628,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4282622288490212,
+      "grad_norm": 3.5168421863357047,
+      "learning_rate": 1.2776802128023317e-05,
+      "loss": 3.2335,
+      "step": 2226
+    },
+    {
+      "epoch": 0.428454619787408,
+      "grad_norm": 3.6289742238905798,
+      "learning_rate": 1.2770814592512853e-05,
+      "loss": 3.398,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4286470107257948,
+      "grad_norm": 3.47417074190454,
+      "learning_rate": 1.2764825980849931e-05,
+      "loss": 3.1057,
+      "step": 2228
+    },
+    {
+      "epoch": 0.4288394016641816,
+      "grad_norm": 3.6165763551788324,
+      "learning_rate": 1.2758836295360455e-05,
+      "loss": 3.3258,
+      "step": 2229
+    },
+    {
+      "epoch": 0.4290317926025684,
+      "grad_norm": 3.390255578825238,
+      "learning_rate": 1.2752845538370752e-05,
+      "loss": 3.2029,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4292241835409552,
+      "grad_norm": 3.637442030850725,
+      "learning_rate": 1.2746853712207567e-05,
+      "loss": 2.9322,
+      "step": 2231
+    },
+    {
+      "epoch": 0.429416574479342,
+      "grad_norm": 3.69909647895041,
+      "learning_rate": 1.274086081919805e-05,
+      "loss": 3.1355,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4296089654177288,
+      "grad_norm": 2.3611121509004205,
+      "learning_rate": 1.273486686166977e-05,
+      "loss": 2.6996,
+      "step": 2233
+    },
+    {
+      "epoch": 0.4298013563561156,
+      "grad_norm": 2.508938437266314,
+      "learning_rate": 1.2728871841950719e-05,
+      "loss": 2.7795,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4299937472945024,
+      "grad_norm": 3.3840301302629547,
+      "learning_rate": 1.2722875762369288e-05,
+      "loss": 3.3026,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4301861382328892,
+      "grad_norm": 3.297027749335064,
+      "learning_rate": 1.2716878625254287e-05,
+      "loss": 2.7595,
+      "step": 2236
+    },
+    {
+      "epoch": 0.43037852917127606,
+      "grad_norm": 3.5782080779793004,
+      "learning_rate": 1.2710880432934934e-05,
+      "loss": 3.2477,
+      "step": 2237
+    },
+    {
+      "epoch": 0.43057092010966286,
+      "grad_norm": 3.25093879474937,
+      "learning_rate": 1.270488118774086e-05,
+      "loss": 3.1277,
+      "step": 2238
+    },
+    {
+      "epoch": 0.43076331104804966,
+      "grad_norm": 3.3149724076627174,
+      "learning_rate": 1.26988808920021e-05,
+      "loss": 3.1722,
+      "step": 2239
+    },
+    {
+      "epoch": 0.43095570198643646,
+      "grad_norm": 3.4380210886190414,
+      "learning_rate": 1.26928795480491e-05,
+      "loss": 2.6781,
+      "step": 2240
+    },
+    {
+      "epoch": 0.43114809292482326,
+      "grad_norm": 3.4294123358872337,
+      "learning_rate": 1.2686877158212715e-05,
+      "loss": 2.9407,
+      "step": 2241
+    },
+    {
+      "epoch": 0.43134048386321006,
+      "grad_norm": 3.163029136369284,
+      "learning_rate": 1.26808737248242e-05,
+      "loss": 3.0217,
+      "step": 2242
+    },
+    {
+      "epoch": 0.43153287480159686,
+      "grad_norm": 3.349366287505634,
+      "learning_rate": 1.2674869250215225e-05,
+      "loss": 3.1381,
+      "step": 2243
+    },
+    {
+      "epoch": 0.43172526573998365,
+      "grad_norm": 2.5113762868043144,
+      "learning_rate": 1.2668863736717855e-05,
+      "loss": 2.7428,
+      "step": 2244
+    },
+    {
+      "epoch": 0.43191765667837045,
+      "grad_norm": 2.3914106494587255,
+      "learning_rate": 1.2662857186664558e-05,
+      "loss": 2.8037,
+      "step": 2245
+    },
+    {
+      "epoch": 0.43211004761675725,
+      "grad_norm": 3.620943708660638,
+      "learning_rate": 1.2656849602388222e-05,
+      "loss": 3.0846,
+      "step": 2246
+    },
+    {
+      "epoch": 0.43230243855514405,
+      "grad_norm": 3.804360475654683,
+      "learning_rate": 1.2650840986222111e-05,
+      "loss": 3.2374,
+      "step": 2247
+    },
+    {
+      "epoch": 0.43249482949353085,
+      "grad_norm": 3.6595721018284997,
+      "learning_rate": 1.2644831340499906e-05,
+      "loss": 2.9764,
+      "step": 2248
+    },
+    {
+      "epoch": 0.43268722043191765,
+      "grad_norm": 3.8145801470965064,
+      "learning_rate": 1.2638820667555685e-05,
+      "loss": 3.3005,
+      "step": 2249
+    },
+    {
+      "epoch": 0.43287961137030445,
+      "grad_norm": 3.7831132022953775,
+      "learning_rate": 1.2632808969723927e-05,
+      "loss": 3.4949,
+      "step": 2250
+    },
+    {
+      "epoch": 0.43307200230869125,
+      "grad_norm": 2.3179101291216058,
+      "learning_rate": 1.26267962493395e-05,
+      "loss": 2.593,
+      "step": 2251
+    },
+    {
+      "epoch": 0.43326439324707805,
+      "grad_norm": 3.4963855695459825,
+      "learning_rate": 1.2620782508737678e-05,
+      "loss": 2.9457,
+      "step": 2252
+    },
+    {
+      "epoch": 0.43345678418546485,
+      "grad_norm": 3.8246620463608894,
+      "learning_rate": 1.2614767750254129e-05,
+      "loss": 3.2607,
+      "step": 2253
+    },
+    {
+      "epoch": 0.43364917512385165,
+      "grad_norm": 3.4094780629794585,
+      "learning_rate": 1.2608751976224916e-05,
+      "loss": 3.2569,
+      "step": 2254
+    },
+    {
+      "epoch": 0.43384156606223845,
+      "grad_norm": 3.413649258464563,
+      "learning_rate": 1.2602735188986498e-05,
+      "loss": 3.3933,
+      "step": 2255
+    },
+    {
+      "epoch": 0.43403395700062525,
+      "grad_norm": 3.5989922071355425,
+      "learning_rate": 1.2596717390875721e-05,
+      "loss": 2.9521,
+      "step": 2256
+    },
+    {
+      "epoch": 0.43422634793901205,
+      "grad_norm": 3.4149781225905547,
+      "learning_rate": 1.2590698584229834e-05,
+      "loss": 3.3387,
+      "step": 2257
+    },
+    {
+      "epoch": 0.4344187388773989,
+      "grad_norm": 3.3560908869870762,
+      "learning_rate": 1.2584678771386467e-05,
+      "loss": 2.9359,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4346111298157857,
+      "grad_norm": 3.2948740263444103,
+      "learning_rate": 1.2578657954683651e-05,
+      "loss": 3.1217,
+      "step": 2259
+    },
+    {
+      "epoch": 0.4348035207541725,
+      "grad_norm": 3.5986700008152237,
+      "learning_rate": 1.2572636136459799e-05,
+      "loss": 3.0803,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4349959116925593,
+      "grad_norm": 3.636954716582883,
+      "learning_rate": 1.2566613319053713e-05,
+      "loss": 3.2193,
+      "step": 2261
+    },
+    {
+      "epoch": 0.4351883026309461,
+      "grad_norm": 2.4390551640356697,
+      "learning_rate": 1.2560589504804592e-05,
+      "loss": 2.8194,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4353806935693329,
+      "grad_norm": 3.7412496378548004,
+      "learning_rate": 1.2554564696052011e-05,
+      "loss": 2.9868,
+      "step": 2263
+    },
+    {
+      "epoch": 0.4355730845077197,
+      "grad_norm": 3.5734847714654063,
+      "learning_rate": 1.2548538895135942e-05,
+      "loss": 2.974,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4357654754461065,
+      "grad_norm": 3.251491121797685,
+      "learning_rate": 1.254251210439673e-05,
+      "loss": 3.0847,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4359578663844933,
+      "grad_norm": 3.5153631402564005,
+      "learning_rate": 1.2536484326175114e-05,
+      "loss": 3.3424,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4361502573228801,
+      "grad_norm": 3.4692498240124334,
+      "learning_rate": 1.2530455562812214e-05,
+      "loss": 3.2726,
+      "step": 2267
+    },
+    {
+      "epoch": 0.4363426482612669,
+      "grad_norm": 3.978704334991836,
+      "learning_rate": 1.252442581664953e-05,
+      "loss": 3.1268,
+      "step": 2268
+    },
+    {
+      "epoch": 0.4365350391996537,
+      "grad_norm": 3.542487606943788,
+      "learning_rate": 1.2518395090028952e-05,
+      "loss": 3.0713,
+      "step": 2269
+    },
+    {
+      "epoch": 0.4367274301380405,
+      "grad_norm": 3.456514322918797,
+      "learning_rate": 1.2512363385292739e-05,
+      "loss": 3.0296,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4369198210764273,
+      "grad_norm": 2.3505328283112985,
+      "learning_rate": 1.2506330704783533e-05,
+      "loss": 2.6598,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4371122120148141,
+      "grad_norm": 3.226986990946427,
+      "learning_rate": 1.2500297050844367e-05,
+      "loss": 3.1412,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4373046029532009,
+      "grad_norm": 3.682678017868266,
+      "learning_rate": 1.2494262425818637e-05,
+      "loss": 3.1513,
+      "step": 2273
+    },
+    {
+      "epoch": 0.4374969938915877,
+      "grad_norm": 3.320005800646741,
+      "learning_rate": 1.2488226832050116e-05,
+      "loss": 2.7846,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4376893848299745,
+      "grad_norm": 3.4728092705540705,
+      "learning_rate": 1.2482190271882973e-05,
+      "loss": 3.2221,
+      "step": 2275
+    },
+    {
+      "epoch": 0.4378817757683613,
+      "grad_norm": 5.0785416439559254,
+      "learning_rate": 1.2476152747661727e-05,
+      "loss": 3.0954,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4380741667067481,
+      "grad_norm": 3.607020349923809,
+      "learning_rate": 1.2470114261731288e-05,
+      "loss": 2.7474,
+      "step": 2277
+    },
+    {
+      "epoch": 0.4382665576451349,
+      "grad_norm": 3.5185756434316544,
+      "learning_rate": 1.246407481643693e-05,
+      "loss": 3.0817,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4384589485835217,
+      "grad_norm": 3.5548817800560304,
+      "learning_rate": 1.245803441412431e-05,
+      "loss": 3.1432,
+      "step": 2279
+    },
+    {
+      "epoch": 0.43865133952190855,
+      "grad_norm": 3.7776481042044807,
+      "learning_rate": 1.2451993057139445e-05,
+      "loss": 3.4563,
+      "step": 2280
+    },
+    {
+      "epoch": 0.43884373046029534,
+      "grad_norm": 3.2505221589571307,
+      "learning_rate": 1.2445950747828732e-05,
+      "loss": 2.8677,
+      "step": 2281
+    },
+    {
+      "epoch": 0.43903612139868214,
+      "grad_norm": 3.3989925660939244,
+      "learning_rate": 1.2439907488538934e-05,
+      "loss": 3.3672,
+      "step": 2282
+    },
+    {
+      "epoch": 0.43922851233706894,
+      "grad_norm": 3.2194219894624854,
+      "learning_rate": 1.243386328161718e-05,
+      "loss": 3.014,
+      "step": 2283
+    },
+    {
+      "epoch": 0.43942090327545574,
+      "grad_norm": 3.7871929259527692,
+      "learning_rate": 1.2427818129410975e-05,
+      "loss": 3.3041,
+      "step": 2284
+    },
+    {
+      "epoch": 0.43961329421384254,
+      "grad_norm": 3.5735273874992575,
+      "learning_rate": 1.2421772034268187e-05,
+      "loss": 2.99,
+      "step": 2285
+    },
+    {
+      "epoch": 0.43980568515222934,
+      "grad_norm": 3.7079249152087246,
+      "learning_rate": 1.2415724998537042e-05,
+      "loss": 3.047,
+      "step": 2286
+    },
+    {
+      "epoch": 0.43999807609061614,
+      "grad_norm": 3.4110811748142087,
+      "learning_rate": 1.2409677024566145e-05,
+      "loss": 3.1686,
+      "step": 2287
+    },
+    {
+      "epoch": 0.44019046702900294,
+      "grad_norm": 3.0947197216124014,
+      "learning_rate": 1.240362811470446e-05,
+      "loss": 3.1774,
+      "step": 2288
+    },
+    {
+      "epoch": 0.44038285796738974,
+      "grad_norm": 3.1272332494068316,
+      "learning_rate": 1.2397578271301312e-05,
+      "loss": 3.215,
+      "step": 2289
+    },
+    {
+      "epoch": 0.44057524890577654,
+      "grad_norm": 3.4137404385951196,
+      "learning_rate": 1.2391527496706389e-05,
+      "loss": 2.9449,
+      "step": 2290
+    },
+    {
+      "epoch": 0.44076763984416334,
+      "grad_norm": 3.3416770924260786,
+      "learning_rate": 1.2385475793269744e-05,
+      "loss": 3.0356,
+      "step": 2291
+    },
+    {
+      "epoch": 0.44096003078255014,
+      "grad_norm": 3.4108538427188573,
+      "learning_rate": 1.2379423163341791e-05,
+      "loss": 3.15,
+      "step": 2292
+    },
+    {
+      "epoch": 0.44115242172093694,
+      "grad_norm": 3.2870505453347394,
+      "learning_rate": 1.2373369609273299e-05,
+      "loss": 3.246,
+      "step": 2293
+    },
+    {
+      "epoch": 0.44134481265932374,
+      "grad_norm": 3.5293991153876214,
+      "learning_rate": 1.2367315133415396e-05,
+      "loss": 3.3957,
+      "step": 2294
+    },
+    {
+      "epoch": 0.44153720359771054,
+      "grad_norm": 2.46756972576714,
+      "learning_rate": 1.2361259738119575e-05,
+      "loss": 2.6941,
+      "step": 2295
+    },
+    {
+      "epoch": 0.44172959453609734,
+      "grad_norm": 3.2651182944031416,
+      "learning_rate": 1.2355203425737683e-05,
+      "loss": 3.0076,
+      "step": 2296
+    },
+    {
+      "epoch": 0.44192198547448414,
+      "grad_norm": 3.3584262356789534,
+      "learning_rate": 1.2349146198621917e-05,
+      "loss": 2.9671,
+      "step": 2297
+    },
+    {
+      "epoch": 0.44211437641287094,
+      "grad_norm": 3.5240582773817675,
+      "learning_rate": 1.2343088059124839e-05,
+      "loss": 3.3132,
+      "step": 2298
+    },
+    {
+      "epoch": 0.44230676735125773,
+      "grad_norm": 3.6282809742039692,
+      "learning_rate": 1.2337029009599357e-05,
+      "loss": 3.2497,
+      "step": 2299
+    },
+    {
+      "epoch": 0.44249915828964453,
+      "grad_norm": 3.737508341326293,
+      "learning_rate": 1.2330969052398735e-05,
+      "loss": 3.0586,
+      "step": 2300
+    },
+    {
+      "epoch": 0.44269154922803133,
+      "grad_norm": 3.4484328597684106,
+      "learning_rate": 1.2324908189876597e-05,
+      "loss": 3.2266,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4428839401664182,
+      "grad_norm": 3.90411243120043,
+      "learning_rate": 1.2318846424386907e-05,
+      "loss": 3.0872,
+      "step": 2302
+    },
+    {
+      "epoch": 0.443076331104805,
+      "grad_norm": 3.480430063002451,
+      "learning_rate": 1.2312783758283981e-05,
+      "loss": 3.0268,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4432687220431918,
+      "grad_norm": 3.365168133019966,
+      "learning_rate": 1.23067201939225e-05,
+      "loss": 3.1749,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4434611129815786,
+      "grad_norm": 3.683048409513594,
+      "learning_rate": 1.2300655733657475e-05,
+      "loss": 3.2456,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4436535039199654,
+      "grad_norm": 2.483844763833368,
+      "learning_rate": 1.2294590379844268e-05,
+      "loss": 2.7184,
+      "step": 2306
+    },
+    {
+      "epoch": 0.4438458948583522,
+      "grad_norm": 3.7081243315300534,
+      "learning_rate": 1.2288524134838602e-05,
+      "loss": 3.1922,
+      "step": 2307
+    },
+    {
+      "epoch": 0.444038285796739,
+      "grad_norm": 2.225843852812723,
+      "learning_rate": 1.2282457000996533e-05,
+      "loss": 2.6237,
+      "step": 2308
+    },
+    {
+      "epoch": 0.4442306767351258,
+      "grad_norm": 3.2715693951996982,
+      "learning_rate": 1.2276388980674465e-05,
+      "loss": 2.7793,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4444230676735126,
+      "grad_norm": 3.4169102268260594,
+      "learning_rate": 1.227032007622915e-05,
+      "loss": 3.1017,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4446154586118994,
+      "grad_norm": 3.408952919852035,
+      "learning_rate": 1.2264250290017675e-05,
+      "loss": 2.9627,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4448078495502862,
+      "grad_norm": 3.2041491344563107,
+      "learning_rate": 1.2258179624397477e-05,
+      "loss": 3.2045,
+      "step": 2312
+    },
+    {
+      "epoch": 0.445000240488673,
+      "grad_norm": 3.5874023271498356,
+      "learning_rate": 1.2252108081726337e-05,
+      "loss": 3.2133,
+      "step": 2313
+    },
+    {
+      "epoch": 0.4451926314270598,
+      "grad_norm": 3.2136633728025363,
+      "learning_rate": 1.224603566436237e-05,
+      "loss": 3.0184,
+      "step": 2314
+    },
+    {
+      "epoch": 0.4453850223654466,
+      "grad_norm": 3.718257219816222,
+      "learning_rate": 1.2239962374664029e-05,
+      "loss": 3.2525,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4455774133038334,
+      "grad_norm": 2.541528997861376,
+      "learning_rate": 1.2233888214990113e-05,
+      "loss": 2.6687,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4457698042422202,
+      "grad_norm": 3.826704736323138,
+      "learning_rate": 1.2227813187699757e-05,
+      "loss": 3.2661,
+      "step": 2317
+    },
+    {
+      "epoch": 0.445962195180607,
+      "grad_norm": 3.1845942092540045,
+      "learning_rate": 1.222173729515243e-05,
+      "loss": 3.1119,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4461545861189938,
+      "grad_norm": 3.820575318533883,
+      "learning_rate": 1.2215660539707936e-05,
+      "loss": 3.2416,
+      "step": 2319
+    },
+    {
+      "epoch": 0.4463469770573806,
+      "grad_norm": 3.3898456119241764,
+      "learning_rate": 1.2209582923726424e-05,
+      "loss": 3.1896,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4465393679957674,
+      "grad_norm": 3.0281680586049684,
+      "learning_rate": 1.2203504449568361e-05,
+      "loss": 3.1774,
+      "step": 2321
+    },
+    {
+      "epoch": 0.4467317589341542,
+      "grad_norm": 3.6562612269330614,
+      "learning_rate": 1.2197425119594563e-05,
+      "loss": 2.8808,
+      "step": 2322
+    },
+    {
+      "epoch": 0.446924149872541,
+      "grad_norm": 3.5268684641083468,
+      "learning_rate": 1.219134493616617e-05,
+      "loss": 3.0889,
+      "step": 2323
+    },
+    {
+      "epoch": 0.44711654081092783,
+      "grad_norm": 3.650451343159138,
+      "learning_rate": 1.2185263901644653e-05,
+      "loss": 3.2881,
+      "step": 2324
+    },
+    {
+      "epoch": 0.44730893174931463,
+      "grad_norm": 3.486597320773926,
+      "learning_rate": 1.217918201839182e-05,
+      "loss": 3.3121,
+      "step": 2325
+    },
+    {
+      "epoch": 0.44750132268770143,
+      "grad_norm": 3.4412123242451345,
+      "learning_rate": 1.2173099288769799e-05,
+      "loss": 3.131,
+      "step": 2326
+    },
+    {
+      "epoch": 0.44769371362608823,
+      "grad_norm": 3.6378533246579563,
+      "learning_rate": 1.2167015715141057e-05,
+      "loss": 3.3464,
+      "step": 2327
+    },
+    {
+      "epoch": 0.44788610456447503,
+      "grad_norm": 3.513045166719936,
+      "learning_rate": 1.216093129986838e-05,
+      "loss": 3.0457,
+      "step": 2328
+    },
+    {
+      "epoch": 0.44807849550286183,
+      "grad_norm": 3.382636289770163,
+      "learning_rate": 1.2154846045314885e-05,
+      "loss": 3.0419,
+      "step": 2329
+    },
+    {
+      "epoch": 0.44827088644124863,
+      "grad_norm": 3.4937035637340945,
+      "learning_rate": 1.214875995384402e-05,
+      "loss": 3.0254,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4484632773796354,
+      "grad_norm": 3.4021604822211673,
+      "learning_rate": 1.214267302781955e-05,
+      "loss": 3.0653,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4486556683180222,
+      "grad_norm": 3.4153140363527346,
+      "learning_rate": 1.2136585269605558e-05,
+      "loss": 3.1686,
+      "step": 2332
+    },
+    {
+      "epoch": 0.448848059256409,
+      "grad_norm": 3.4155222276544057,
+      "learning_rate": 1.2130496681566475e-05,
+      "loss": 3.2288,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4490404501947958,
+      "grad_norm": 3.4641524331108347,
+      "learning_rate": 1.212440726606703e-05,
+      "loss": 3.0134,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4492328411331826,
+      "grad_norm": 2.473150667512845,
+      "learning_rate": 1.211831702547228e-05,
+      "loss": 2.5668,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4494252320715694,
+      "grad_norm": 2.4786207467797885,
+      "learning_rate": 1.2112225962147605e-05,
+      "loss": 2.7645,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4496176230099562,
+      "grad_norm": 3.884012121713513,
+      "learning_rate": 1.210613407845871e-05,
+      "loss": 3.2269,
+      "step": 2337
+    },
+    {
+      "epoch": 0.449810013948343,
+      "grad_norm": 3.6637201303465585,
+      "learning_rate": 1.2100041376771605e-05,
+      "loss": 3.122,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4500024048867298,
+      "grad_norm": 3.3464575000086,
+      "learning_rate": 1.209394785945263e-05,
+      "loss": 2.8385,
+      "step": 2339
+    },
+    {
+      "epoch": 0.4501947958251166,
+      "grad_norm": 3.4315255077116382,
+      "learning_rate": 1.2087853528868432e-05,
+      "loss": 3.2085,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4503871867635034,
+      "grad_norm": 3.3415079294175776,
+      "learning_rate": 1.2081758387385982e-05,
+      "loss": 3.1443,
+      "step": 2341
+    },
+    {
+      "epoch": 0.4505795777018902,
+      "grad_norm": 3.3697791405166346,
+      "learning_rate": 1.2075662437372567e-05,
+      "loss": 3.1345,
+      "step": 2342
+    },
+    {
+      "epoch": 0.450771968640277,
+      "grad_norm": 3.5975627704126034,
+      "learning_rate": 1.2069565681195776e-05,
+      "loss": 3.1688,
+      "step": 2343
+    },
+    {
+      "epoch": 0.4509643595786638,
+      "grad_norm": 3.574192915641791,
+      "learning_rate": 1.206346812122352e-05,
+      "loss": 3.233,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4511567505170506,
+      "grad_norm": 3.2650533795219054,
+      "learning_rate": 1.2057369759824025e-05,
+      "loss": 2.9733,
+      "step": 2345
+    },
+    {
+      "epoch": 0.4513491414554375,
+      "grad_norm": 3.336539387399029,
+      "learning_rate": 1.2051270599365825e-05,
+      "loss": 2.9838,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4515415323938243,
+      "grad_norm": 3.5182334103936004,
+      "learning_rate": 1.2045170642217756e-05,
+      "loss": 3.1424,
+      "step": 2347
+    },
+    {
+      "epoch": 0.4517339233322111,
+      "grad_norm": 3.3222790142736476,
+      "learning_rate": 1.2039069890748978e-05,
+      "loss": 3.0202,
+      "step": 2348
+    },
+    {
+      "epoch": 0.4519263142705979,
+      "grad_norm": 3.2899373612069387,
+      "learning_rate": 1.2032968347328952e-05,
+      "loss": 3.0477,
+      "step": 2349
+    },
+    {
+      "epoch": 0.4521187052089847,
+      "grad_norm": 3.6222008420481755,
+      "learning_rate": 1.2026866014327446e-05,
+      "loss": 3.3767,
+      "step": 2350
+    },
+    {
+      "epoch": 0.45231109614737147,
+      "grad_norm": 3.4677051950754016,
+      "learning_rate": 1.2020762894114535e-05,
+      "loss": 3.0229,
+      "step": 2351
+    },
+    {
+      "epoch": 0.45250348708575827,
+      "grad_norm": 3.8558556286598034,
+      "learning_rate": 1.20146589890606e-05,
+      "loss": 3.0887,
+      "step": 2352
+    },
+    {
+      "epoch": 0.45269587802414507,
+      "grad_norm": 3.274456797022817,
+      "learning_rate": 1.2008554301536328e-05,
+      "loss": 2.9302,
+      "step": 2353
+    },
+    {
+      "epoch": 0.45288826896253187,
+      "grad_norm": 3.754679942021001,
+      "learning_rate": 1.2002448833912712e-05,
+      "loss": 3.4723,
+      "step": 2354
+    },
+    {
+      "epoch": 0.45308065990091867,
+      "grad_norm": 3.382256099286495,
+      "learning_rate": 1.1996342588561042e-05,
+      "loss": 3.2005,
+      "step": 2355
+    },
+    {
+      "epoch": 0.45327305083930547,
+      "grad_norm": 3.325252600108224,
+      "learning_rate": 1.1990235567852917e-05,
+      "loss": 2.9654,
+      "step": 2356
+    },
+    {
+      "epoch": 0.45346544177769227,
+      "grad_norm": 3.4146796699333666,
+      "learning_rate": 1.1984127774160226e-05,
+      "loss": 3.0141,
+      "step": 2357
+    },
+    {
+      "epoch": 0.45365783271607907,
+      "grad_norm": 3.620780266972449,
+      "learning_rate": 1.1978019209855174e-05,
+      "loss": 2.948,
+      "step": 2358
+    },
+    {
+      "epoch": 0.45385022365446587,
+      "grad_norm": 2.371695927066451,
+      "learning_rate": 1.1971909877310253e-05,
+      "loss": 2.6123,
+      "step": 2359
+    },
+    {
+      "epoch": 0.45404261459285267,
+      "grad_norm": 3.6564357778850503,
+      "learning_rate": 1.1965799778898258e-05,
+      "loss": 3.4698,
+      "step": 2360
+    },
+    {
+      "epoch": 0.45423500553123947,
+      "grad_norm": 3.4164387101472506,
+      "learning_rate": 1.1959688916992279e-05,
+      "loss": 3.2581,
+      "step": 2361
+    },
+    {
+      "epoch": 0.45442739646962627,
+      "grad_norm": 3.5282415861632366,
+      "learning_rate": 1.1953577293965707e-05,
+      "loss": 2.7933,
+      "step": 2362
+    },
+    {
+      "epoch": 0.45461978740801307,
+      "grad_norm": 3.7018821952865535,
+      "learning_rate": 1.1947464912192228e-05,
+      "loss": 3.0725,
+      "step": 2363
+    },
+    {
+      "epoch": 0.45481217834639986,
+      "grad_norm": 3.4313773650140584,
+      "learning_rate": 1.1941351774045815e-05,
+      "loss": 3.1738,
+      "step": 2364
+    },
+    {
+      "epoch": 0.45500456928478666,
+      "grad_norm": 3.6823863893046043,
+      "learning_rate": 1.1935237881900743e-05,
+      "loss": 3.3601,
+      "step": 2365
+    },
+    {
+      "epoch": 0.45519696022317346,
+      "grad_norm": 3.331725629559208,
+      "learning_rate": 1.1929123238131579e-05,
+      "loss": 3.2247,
+      "step": 2366
+    },
+    {
+      "epoch": 0.45538935116156026,
+      "grad_norm": 3.4284671418575483,
+      "learning_rate": 1.1923007845113178e-05,
+      "loss": 3.2939,
+      "step": 2367
+    },
+    {
+      "epoch": 0.4555817420999471,
+      "grad_norm": 3.3483819307242637,
+      "learning_rate": 1.1916891705220689e-05,
+      "loss": 3.1763,
+      "step": 2368
+    },
+    {
+      "epoch": 0.4557741330383339,
+      "grad_norm": 3.1666015331482384,
+      "learning_rate": 1.191077482082955e-05,
+      "loss": 3.4183,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4559665239767207,
+      "grad_norm": 3.5045293520275247,
+      "learning_rate": 1.1904657194315486e-05,
+      "loss": 3.0197,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4561589149151075,
+      "grad_norm": 3.531873246954225,
+      "learning_rate": 1.1898538828054517e-05,
+      "loss": 2.8459,
+      "step": 2371
+    },
+    {
+      "epoch": 0.4563513058534943,
+      "grad_norm": 3.096213079233776,
+      "learning_rate": 1.1892419724422946e-05,
+      "loss": 2.9806,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4565436967918811,
+      "grad_norm": 3.524633673346819,
+      "learning_rate": 1.1886299885797357e-05,
+      "loss": 3.0664,
+      "step": 2373
+    },
+    {
+      "epoch": 0.4567360877302679,
+      "grad_norm": 3.5561957305118286,
+      "learning_rate": 1.1880179314554629e-05,
+      "loss": 3.1389,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4569284786686547,
+      "grad_norm": 3.543777938028806,
+      "learning_rate": 1.1874058013071923e-05,
+      "loss": 3.2627,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4571208696070415,
+      "grad_norm": 3.6808662956373004,
+      "learning_rate": 1.1867935983726676e-05,
+      "loss": 3.2753,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4573132605454283,
+      "grad_norm": 3.5016878077217712,
+      "learning_rate": 1.186181322889662e-05,
+      "loss": 3.2417,
+      "step": 2377
+    },
+    {
+      "epoch": 0.4575056514838151,
+      "grad_norm": 3.283940427613109,
+      "learning_rate": 1.1855689750959759e-05,
+      "loss": 3.3414,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4576980424222019,
+      "grad_norm": 3.353972205404856,
+      "learning_rate": 1.1849565552294379e-05,
+      "loss": 3.128,
+      "step": 2379
+    },
+    {
+      "epoch": 0.4578904333605887,
+      "grad_norm": 3.369570652115817,
+      "learning_rate": 1.1843440635279056e-05,
+      "loss": 3.1906,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4580828242989755,
+      "grad_norm": 3.4561972108132437,
+      "learning_rate": 1.1837315002292629e-05,
+      "loss": 3.2326,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4582752152373623,
+      "grad_norm": 3.4112647729463603,
+      "learning_rate": 1.1831188655714225e-05,
+      "loss": 3.0018,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4584676061757491,
+      "grad_norm": 3.5337549418030374,
+      "learning_rate": 1.182506159792325e-05,
+      "loss": 3.2777,
+      "step": 2383
+    },
+    {
+      "epoch": 0.4586599971141359,
+      "grad_norm": 3.3065985994082565,
+      "learning_rate": 1.1818933831299381e-05,
+      "loss": 2.9555,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4588523880525227,
+      "grad_norm": 3.391034948458295,
+      "learning_rate": 1.1812805358222571e-05,
+      "loss": 2.9424,
+      "step": 2385
+    },
+    {
+      "epoch": 0.4590447789909095,
+      "grad_norm": 3.7507952615341638,
+      "learning_rate": 1.180667618107305e-05,
+      "loss": 3.2795,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4592371699292963,
+      "grad_norm": 3.7320304185405955,
+      "learning_rate": 1.1800546302231317e-05,
+      "loss": 3.2928,
+      "step": 2387
+    },
+    {
+      "epoch": 0.4594295608676831,
+      "grad_norm": 3.365786085140123,
+      "learning_rate": 1.1794415724078147e-05,
+      "loss": 3.0695,
+      "step": 2388
+    },
+    {
+      "epoch": 0.45962195180606996,
+      "grad_norm": 3.2399473597059285,
+      "learning_rate": 1.1788284448994588e-05,
+      "loss": 3.0078,
+      "step": 2389
+    },
+    {
+      "epoch": 0.45981434274445676,
+      "grad_norm": 3.5543638309587133,
+      "learning_rate": 1.1782152479361956e-05,
+      "loss": 3.0488,
+      "step": 2390
+    },
+    {
+      "epoch": 0.46000673368284356,
+      "grad_norm": 3.3575355541314638,
+      "learning_rate": 1.1776019817561834e-05,
+      "loss": 2.93,
+      "step": 2391
+    },
+    {
+      "epoch": 0.46019912462123036,
+      "grad_norm": 3.38363297265605,
+      "learning_rate": 1.1769886465976086e-05,
+      "loss": 3.0129,
+      "step": 2392
+    },
+    {
+      "epoch": 0.46039151555961716,
+      "grad_norm": 3.3019208603812404,
+      "learning_rate": 1.1763752426986823e-05,
+      "loss": 3.1575,
+      "step": 2393
+    },
+    {
+      "epoch": 0.46058390649800396,
+      "grad_norm": 3.6654181732311213,
+      "learning_rate": 1.1757617702976443e-05,
+      "loss": 3.1141,
+      "step": 2394
+    },
+    {
+      "epoch": 0.46077629743639076,
+      "grad_norm": 3.554762721521145,
+      "learning_rate": 1.17514822963276e-05,
+      "loss": 3.1678,
+      "step": 2395
+    },
+    {
+      "epoch": 0.46096868837477756,
+      "grad_norm": 3.8427957317037267,
+      "learning_rate": 1.1745346209423216e-05,
+      "loss": 2.9478,
+      "step": 2396
+    },
+    {
+      "epoch": 0.46116107931316436,
+      "grad_norm": 2.175669939211205,
+      "learning_rate": 1.1739209444646479e-05,
+      "loss": 2.542,
+      "step": 2397
+    },
+    {
+      "epoch": 0.46135347025155116,
+      "grad_norm": 3.5786561269920205,
+      "learning_rate": 1.1733072004380827e-05,
+      "loss": 3.2601,
+      "step": 2398
+    },
+    {
+      "epoch": 0.46154586118993796,
+      "grad_norm": 3.319843248967918,
+      "learning_rate": 1.1726933891009985e-05,
+      "loss": 2.8159,
+      "step": 2399
+    },
+    {
+      "epoch": 0.46173825212832476,
+      "grad_norm": 3.4748953635755186,
+      "learning_rate": 1.1720795106917917e-05,
+      "loss": 3.1661,
+      "step": 2400
+    },
+    {
+      "epoch": 0.46193064306671155,
+      "grad_norm": 3.457139153682475,
+      "learning_rate": 1.171465565448886e-05,
+      "loss": 3.1236,
+      "step": 2401
+    },
+    {
+      "epoch": 0.46212303400509835,
+      "grad_norm": 3.3852523801494,
+      "learning_rate": 1.1708515536107299e-05,
+      "loss": 3.1359,
+      "step": 2402
+    },
+    {
+      "epoch": 0.46231542494348515,
+      "grad_norm": 3.521037001506194,
+      "learning_rate": 1.1702374754157998e-05,
+      "loss": 2.9654,
+      "step": 2403
+    },
+    {
+      "epoch": 0.46250781588187195,
+      "grad_norm": 3.351991507054885,
+      "learning_rate": 1.1696233311025957e-05,
+      "loss": 3.2643,
+      "step": 2404
+    },
+    {
+      "epoch": 0.46270020682025875,
+      "grad_norm": 3.2767127208438476,
+      "learning_rate": 1.1690091209096441e-05,
+      "loss": 3.0494,
+      "step": 2405
+    },
+    {
+      "epoch": 0.46289259775864555,
+      "grad_norm": 3.2247984028070884,
+      "learning_rate": 1.1683948450754976e-05,
+      "loss": 3.1196,
+      "step": 2406
+    },
+    {
+      "epoch": 0.46308498869703235,
+      "grad_norm": 3.5704605800720826,
+      "learning_rate": 1.1677805038387337e-05,
+      "loss": 3.3615,
+      "step": 2407
+    },
+    {
+      "epoch": 0.46327737963541915,
+      "grad_norm": 3.162086638727146,
+      "learning_rate": 1.1671660974379554e-05,
+      "loss": 3.008,
+      "step": 2408
+    },
+    {
+      "epoch": 0.46346977057380595,
+      "grad_norm": 3.2285397835481047,
+      "learning_rate": 1.1665516261117914e-05,
+      "loss": 2.8455,
+      "step": 2409
+    },
+    {
+      "epoch": 0.46366216151219275,
+      "grad_norm": 3.4067752451580233,
+      "learning_rate": 1.1659370900988946e-05,
+      "loss": 3.1879,
+      "step": 2410
+    },
+    {
+      "epoch": 0.4638545524505796,
+      "grad_norm": 3.8156697848617034,
+      "learning_rate": 1.165322489637944e-05,
+      "loss": 3.3181,
+      "step": 2411
+    },
+    {
+      "epoch": 0.4640469433889664,
+      "grad_norm": 3.5759692012642623,
+      "learning_rate": 1.164707824967644e-05,
+      "loss": 2.992,
+      "step": 2412
+    },
+    {
+      "epoch": 0.4642393343273532,
+      "grad_norm": 3.3533526549069257,
+      "learning_rate": 1.1640930963267226e-05,
+      "loss": 3.0814,
+      "step": 2413
+    },
+    {
+      "epoch": 0.46443172526574,
+      "grad_norm": 3.4539947097816373,
+      "learning_rate": 1.1634783039539328e-05,
+      "loss": 3.2797,
+      "step": 2414
+    },
+    {
+      "epoch": 0.4646241162041268,
+      "grad_norm": 2.457889903208856,
+      "learning_rate": 1.162863448088054e-05,
+      "loss": 2.7666,
+      "step": 2415
+    },
+    {
+      "epoch": 0.4648165071425136,
+      "grad_norm": 2.2734009446907764,
+      "learning_rate": 1.1622485289678886e-05,
+      "loss": 2.7254,
+      "step": 2416
+    },
+    {
+      "epoch": 0.4650088980809004,
+      "grad_norm": 3.529261233762811,
+      "learning_rate": 1.1616335468322641e-05,
+      "loss": 3.2972,
+      "step": 2417
+    },
+    {
+      "epoch": 0.4652012890192872,
+      "grad_norm": 3.771831772069779,
+      "learning_rate": 1.1610185019200324e-05,
+      "loss": 2.9129,
+      "step": 2418
+    },
+    {
+      "epoch": 0.465393679957674,
+      "grad_norm": 3.8277307768260056,
+      "learning_rate": 1.1604033944700701e-05,
+      "loss": 3.1983,
+      "step": 2419
+    },
+    {
+      "epoch": 0.4655860708960608,
+      "grad_norm": 3.668989559061783,
+      "learning_rate": 1.1597882247212776e-05,
+      "loss": 3.2403,
+      "step": 2420
+    },
+    {
+      "epoch": 0.4657784618344476,
+      "grad_norm": 2.242314849086445,
+      "learning_rate": 1.15917299291258e-05,
+      "loss": 2.6348,
+      "step": 2421
+    },
+    {
+      "epoch": 0.4659708527728344,
+      "grad_norm": 3.691587399568446,
+      "learning_rate": 1.1585576992829261e-05,
+      "loss": 3.2281,
+      "step": 2422
+    },
+    {
+      "epoch": 0.4661632437112212,
+      "grad_norm": 3.612196940056623,
+      "learning_rate": 1.1579423440712887e-05,
+      "loss": 3.0818,
+      "step": 2423
+    },
+    {
+      "epoch": 0.466355634649608,
+      "grad_norm": 3.415740370222079,
+      "learning_rate": 1.1573269275166652e-05,
+      "loss": 3.1281,
+      "step": 2424
+    },
+    {
+      "epoch": 0.4665480255879948,
+      "grad_norm": 3.4150210765069207,
+      "learning_rate": 1.1567114498580758e-05,
+      "loss": 3.376,
+      "step": 2425
+    },
+    {
+      "epoch": 0.4667404165263816,
+      "grad_norm": 3.5287594168201664,
+      "learning_rate": 1.1560959113345649e-05,
+      "loss": 3.1139,
+      "step": 2426
+    },
+    {
+      "epoch": 0.4669328074647684,
+      "grad_norm": 3.3292974836315516,
+      "learning_rate": 1.1554803121852005e-05,
+      "loss": 3.0436,
+      "step": 2427
+    },
+    {
+      "epoch": 0.4671251984031552,
+      "grad_norm": 3.3734578045883663,
+      "learning_rate": 1.1548646526490749e-05,
+      "loss": 3.1125,
+      "step": 2428
+    },
+    {
+      "epoch": 0.467317589341542,
+      "grad_norm": 3.4555019434037364,
+      "learning_rate": 1.1542489329653024e-05,
+      "loss": 3.2007,
+      "step": 2429
+    },
+    {
+      "epoch": 0.4675099802799288,
+      "grad_norm": 3.452248821671693,
+      "learning_rate": 1.153633153373022e-05,
+      "loss": 3.2241,
+      "step": 2430
+    },
+    {
+      "epoch": 0.4677023712183156,
+      "grad_norm": 3.4046710540384213,
+      "learning_rate": 1.1530173141113947e-05,
+      "loss": 3.0847,
+      "step": 2431
+    },
+    {
+      "epoch": 0.4678947621567024,
+      "grad_norm": 3.7565706095398803,
+      "learning_rate": 1.1524014154196063e-05,
+      "loss": 2.9757,
+      "step": 2432
+    },
+    {
+      "epoch": 0.46808715309508925,
+      "grad_norm": 3.5134879551157994,
+      "learning_rate": 1.1517854575368644e-05,
+      "loss": 3.2123,
+      "step": 2433
+    },
+    {
+      "epoch": 0.46827954403347605,
+      "grad_norm": 3.303355467311716,
+      "learning_rate": 1.1511694407023994e-05,
+      "loss": 2.9618,
+      "step": 2434
+    },
+    {
+      "epoch": 0.46847193497186285,
+      "grad_norm": 3.5278668307888847,
+      "learning_rate": 1.1505533651554654e-05,
+      "loss": 2.9984,
+      "step": 2435
+    },
+    {
+      "epoch": 0.46866432591024965,
+      "grad_norm": 3.482895661269359,
+      "learning_rate": 1.1499372311353398e-05,
+      "loss": 3.2371,
+      "step": 2436
+    },
+    {
+      "epoch": 0.46885671684863645,
+      "grad_norm": 3.4972816839975933,
+      "learning_rate": 1.149321038881321e-05,
+      "loss": 3.3607,
+      "step": 2437
+    },
+    {
+      "epoch": 0.46904910778702325,
+      "grad_norm": 3.354383681866933,
+      "learning_rate": 1.1487047886327314e-05,
+      "loss": 3.1301,
+      "step": 2438
+    },
+    {
+      "epoch": 0.46924149872541004,
+      "grad_norm": 3.5359412248720345,
+      "learning_rate": 1.1480884806289151e-05,
+      "loss": 3.224,
+      "step": 2439
+    },
+    {
+      "epoch": 0.46943388966379684,
+      "grad_norm": 3.581096262137931,
+      "learning_rate": 1.1474721151092397e-05,
+      "loss": 3.1575,
+      "step": 2440
+    },
+    {
+      "epoch": 0.46962628060218364,
+      "grad_norm": 3.7252766090804186,
+      "learning_rate": 1.1468556923130943e-05,
+      "loss": 2.9744,
+      "step": 2441
+    },
+    {
+      "epoch": 0.46981867154057044,
+      "grad_norm": 3.4079662088569287,
+      "learning_rate": 1.14623921247989e-05,
+      "loss": 3.2427,
+      "step": 2442
+    },
+    {
+      "epoch": 0.47001106247895724,
+      "grad_norm": 3.949092239105996,
+      "learning_rate": 1.1456226758490603e-05,
+      "loss": 3.1622,
+      "step": 2443
+    },
+    {
+      "epoch": 0.47020345341734404,
+      "grad_norm": 3.4012403896536525,
+      "learning_rate": 1.1450060826600618e-05,
+      "loss": 3.0066,
+      "step": 2444
+    },
+    {
+      "epoch": 0.47039584435573084,
+      "grad_norm": 3.6550121657814607,
+      "learning_rate": 1.1443894331523718e-05,
+      "loss": 3.19,
+      "step": 2445
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 3.335392465647964,
+      "learning_rate": 1.1437727275654893e-05,
+      "loss": 2.9125,
+      "step": 2446
+    },
+    {
+      "epoch": 0.47078062623250444,
+      "grad_norm": 3.634733212746068,
+      "learning_rate": 1.1431559661389362e-05,
+      "loss": 3.161,
+      "step": 2447
+    },
+    {
+      "epoch": 0.47097301717089124,
+      "grad_norm": 3.5817593806561847,
+      "learning_rate": 1.1425391491122557e-05,
+      "loss": 3.3153,
+      "step": 2448
+    },
+    {
+      "epoch": 0.47116540810927804,
+      "grad_norm": 3.276376779968075,
+      "learning_rate": 1.141922276725012e-05,
+      "loss": 3.1396,
+      "step": 2449
+    },
+    {
+      "epoch": 0.47135779904766484,
+      "grad_norm": 3.4341293679734046,
+      "learning_rate": 1.1413053492167915e-05,
+      "loss": 3.1352,
+      "step": 2450
+    },
+    {
+      "epoch": 0.47155018998605164,
+      "grad_norm": 3.33878473372695,
+      "learning_rate": 1.1406883668272015e-05,
+      "loss": 3.0425,
+      "step": 2451
+    },
+    {
+      "epoch": 0.47174258092443844,
+      "grad_norm": 3.4111784343579052,
+      "learning_rate": 1.140071329795871e-05,
+      "loss": 3.3,
+      "step": 2452
+    },
+    {
+      "epoch": 0.47193497186282524,
+      "grad_norm": 3.6108835938776402,
+      "learning_rate": 1.13945423836245e-05,
+      "loss": 3.2208,
+      "step": 2453
+    },
+    {
+      "epoch": 0.47212736280121204,
+      "grad_norm": 3.1945515636462667,
+      "learning_rate": 1.1388370927666102e-05,
+      "loss": 3.0446,
+      "step": 2454
+    },
+    {
+      "epoch": 0.4723197537395989,
+      "grad_norm": 2.467342512185269,
+      "learning_rate": 1.1382198932480429e-05,
+      "loss": 2.686,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4725121446779857,
+      "grad_norm": 3.270407111887957,
+      "learning_rate": 1.1376026400464616e-05,
+      "loss": 3.2638,
+      "step": 2456
+    },
+    {
+      "epoch": 0.4727045356163725,
+      "grad_norm": 3.5730961900964973,
+      "learning_rate": 1.136985333401601e-05,
+      "loss": 3.0349,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4728969265547593,
+      "grad_norm": 3.758651805524996,
+      "learning_rate": 1.1363679735532151e-05,
+      "loss": 3.0602,
+      "step": 2458
+    },
+    {
+      "epoch": 0.4730893174931461,
+      "grad_norm": 3.5655384191821575,
+      "learning_rate": 1.1357505607410797e-05,
+      "loss": 3.2315,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4732817084315329,
+      "grad_norm": 3.355239958297069,
+      "learning_rate": 1.1351330952049908e-05,
+      "loss": 3.1384,
+      "step": 2460
+    },
+    {
+      "epoch": 0.4734740993699197,
+      "grad_norm": 3.3942946345610294,
+      "learning_rate": 1.1345155771847646e-05,
+      "loss": 3.0342,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4736664903083065,
+      "grad_norm": 3.689530499837652,
+      "learning_rate": 1.1338980069202388e-05,
+      "loss": 3.1281,
+      "step": 2462
+    },
+    {
+      "epoch": 0.4738588812466933,
+      "grad_norm": 3.807847485981816,
+      "learning_rate": 1.1332803846512697e-05,
+      "loss": 2.9505,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4740512721850801,
+      "grad_norm": 3.3232108119557315,
+      "learning_rate": 1.1326627106177348e-05,
+      "loss": 3.0226,
+      "step": 2464
+    },
+    {
+      "epoch": 0.4742436631234669,
+      "grad_norm": 3.510041273280856,
+      "learning_rate": 1.132044985059532e-05,
+      "loss": 3.2891,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4744360540618537,
+      "grad_norm": 3.3215125396936083,
+      "learning_rate": 1.1314272082165785e-05,
+      "loss": 3.0762,
+      "step": 2466
+    },
+    {
+      "epoch": 0.4746284450002405,
+      "grad_norm": 3.299915629001998,
+      "learning_rate": 1.1308093803288119e-05,
+      "loss": 3.0762,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4748208359386273,
+      "grad_norm": 3.6067506095943687,
+      "learning_rate": 1.130191501636189e-05,
+      "loss": 3.2984,
+      "step": 2468
+    },
+    {
+      "epoch": 0.4750132268770141,
+      "grad_norm": 3.3008106920197258,
+      "learning_rate": 1.1295735723786872e-05,
+      "loss": 2.8773,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4752056178154009,
+      "grad_norm": 3.8028267281524246,
+      "learning_rate": 1.1289555927963032e-05,
+      "loss": 3.0478,
+      "step": 2470
+    },
+    {
+      "epoch": 0.4753980087537877,
+      "grad_norm": 3.1522122625450177,
+      "learning_rate": 1.1283375631290528e-05,
+      "loss": 2.8875,
+      "step": 2471
+    },
+    {
+      "epoch": 0.4755903996921745,
+      "grad_norm": 3.566409187694033,
+      "learning_rate": 1.1277194836169714e-05,
+      "loss": 3.0043,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4757827906305613,
+      "grad_norm": 3.6973282370401988,
+      "learning_rate": 1.1271013545001144e-05,
+      "loss": 3.2642,
+      "step": 2473
+    },
+    {
+      "epoch": 0.4759751815689481,
+      "grad_norm": 3.1697812811719,
+      "learning_rate": 1.1264831760185562e-05,
+      "loss": 2.5153,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4761675725073349,
+      "grad_norm": 3.154690134928266,
+      "learning_rate": 1.1258649484123895e-05,
+      "loss": 3.0606,
+      "step": 2475
+    },
+    {
+      "epoch": 0.4763599634457217,
+      "grad_norm": 3.577129376931919,
+      "learning_rate": 1.1252466719217274e-05,
+      "loss": 3.459,
+      "step": 2476
+    },
+    {
+      "epoch": 0.47655235438410853,
+      "grad_norm": 3.2391462153764214,
+      "learning_rate": 1.1246283467867012e-05,
+      "loss": 3.1735,
+      "step": 2477
+    },
+    {
+      "epoch": 0.47674474532249533,
+      "grad_norm": 3.630321681152221,
+      "learning_rate": 1.1240099732474613e-05,
+      "loss": 3.2035,
+      "step": 2478
+    },
+    {
+      "epoch": 0.47693713626088213,
+      "grad_norm": 3.6251175585169753,
+      "learning_rate": 1.1233915515441765e-05,
+      "loss": 3.1184,
+      "step": 2479
+    },
+    {
+      "epoch": 0.47712952719926893,
+      "grad_norm": 3.1232812127288923,
+      "learning_rate": 1.1227730819170349e-05,
+      "loss": 3.2196,
+      "step": 2480
+    },
+    {
+      "epoch": 0.47732191813765573,
+      "grad_norm": 3.4228164254502396,
+      "learning_rate": 1.1221545646062431e-05,
+      "loss": 3.0233,
+      "step": 2481
+    },
+    {
+      "epoch": 0.47751430907604253,
+      "grad_norm": 3.6148199985160643,
+      "learning_rate": 1.121535999852026e-05,
+      "loss": 3.3915,
+      "step": 2482
+    },
+    {
+      "epoch": 0.47770670001442933,
+      "grad_norm": 3.360082649738497,
+      "learning_rate": 1.1209173878946271e-05,
+      "loss": 3.0002,
+      "step": 2483
+    },
+    {
+      "epoch": 0.47789909095281613,
+      "grad_norm": 3.4001206819973913,
+      "learning_rate": 1.1202987289743078e-05,
+      "loss": 2.9807,
+      "step": 2484
+    },
+    {
+      "epoch": 0.47809148189120293,
+      "grad_norm": 3.541022677507985,
+      "learning_rate": 1.1196800233313488e-05,
+      "loss": 3.3298,
+      "step": 2485
+    },
+    {
+      "epoch": 0.47828387282958973,
+      "grad_norm": 3.694389816460676,
+      "learning_rate": 1.1190612712060475e-05,
+      "loss": 3.1209,
+      "step": 2486
+    },
+    {
+      "epoch": 0.47847626376797653,
+      "grad_norm": 3.584398292475285,
+      "learning_rate": 1.1184424728387204e-05,
+      "loss": 3.027,
+      "step": 2487
+    },
+    {
+      "epoch": 0.47866865470636333,
+      "grad_norm": 3.1996488719833067,
+      "learning_rate": 1.1178236284697017e-05,
+      "loss": 2.6998,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4788610456447501,
+      "grad_norm": 3.9504453231626036,
+      "learning_rate": 1.1172047383393434e-05,
+      "loss": 3.3186,
+      "step": 2489
+    },
+    {
+      "epoch": 0.4790534365831369,
+      "grad_norm": 3.2010344696213733,
+      "learning_rate": 1.1165858026880151e-05,
+      "loss": 3.1892,
+      "step": 2490
+    },
+    {
+      "epoch": 0.4792458275215237,
+      "grad_norm": 3.399238321611192,
+      "learning_rate": 1.1159668217561048e-05,
+      "loss": 3.2612,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4794382184599105,
+      "grad_norm": 3.5768890334942607,
+      "learning_rate": 1.115347795784017e-05,
+      "loss": 3.0123,
+      "step": 2492
+    },
+    {
+      "epoch": 0.4796306093982973,
+      "grad_norm": 3.4070473069431877,
+      "learning_rate": 1.1147287250121745e-05,
+      "loss": 3.2947,
+      "step": 2493
+    },
+    {
+      "epoch": 0.4798230003366841,
+      "grad_norm": 3.581670076330558,
+      "learning_rate": 1.1141096096810174e-05,
+      "loss": 3.3543,
+      "step": 2494
+    },
+    {
+      "epoch": 0.4800153912750709,
+      "grad_norm": 3.5343011947712393,
+      "learning_rate": 1.1134904500310029e-05,
+      "loss": 3.3187,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4802077822134577,
+      "grad_norm": 3.4227103981179505,
+      "learning_rate": 1.1128712463026048e-05,
+      "loss": 2.9078,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4804001731518445,
+      "grad_norm": 3.2981151004040528,
+      "learning_rate": 1.1122519987363156e-05,
+      "loss": 3.1176,
+      "step": 2497
+    },
+    {
+      "epoch": 0.4805925640902314,
+      "grad_norm": 3.5858565749400153,
+      "learning_rate": 1.1116327075726436e-05,
+      "loss": 3.2704,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4807849550286182,
+      "grad_norm": 3.40033435711184,
+      "learning_rate": 1.1110133730521142e-05,
+      "loss": 3.1889,
+      "step": 2499
+    },
+    {
+      "epoch": 0.480977345967005,
+      "grad_norm": 3.2862228609826882,
+      "learning_rate": 1.11039399541527e-05,
+      "loss": 3.1588,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4811697369053918,
+      "grad_norm": 3.653926985552167,
+      "learning_rate": 1.10977457490267e-05,
+      "loss": 3.0038,
+      "step": 2501
+    },
+    {
+      "epoch": 0.4813621278437786,
+      "grad_norm": 3.4447072572313187,
+      "learning_rate": 1.10915511175489e-05,
+      "loss": 3.053,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4815545187821654,
+      "grad_norm": 3.575826869557933,
+      "learning_rate": 1.1085356062125225e-05,
+      "loss": 3.2885,
+      "step": 2503
+    },
+    {
+      "epoch": 0.4817469097205522,
+      "grad_norm": 3.475636986017438,
+      "learning_rate": 1.1079160585161759e-05,
+      "loss": 3.233,
+      "step": 2504
+    },
+    {
+      "epoch": 0.481939300658939,
+      "grad_norm": 2.2961480251985584,
+      "learning_rate": 1.107296468906476e-05,
+      "loss": 2.5922,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4821316915973258,
+      "grad_norm": 3.250987652697073,
+      "learning_rate": 1.106676837624064e-05,
+      "loss": 3.2875,
+      "step": 2506
+    },
+    {
+      "epoch": 0.4823240825357126,
+      "grad_norm": 3.6123701888399498,
+      "learning_rate": 1.1060571649095972e-05,
+      "loss": 3.0272,
+      "step": 2507
+    },
+    {
+      "epoch": 0.48251647347409937,
+      "grad_norm": 3.2735996047484917,
+      "learning_rate": 1.10543745100375e-05,
+      "loss": 2.9927,
+      "step": 2508
+    },
+    {
+      "epoch": 0.48270886441248617,
+      "grad_norm": 3.1009118645915916,
+      "learning_rate": 1.1048176961472114e-05,
+      "loss": 3.099,
+      "step": 2509
+    },
+    {
+      "epoch": 0.48290125535087297,
+      "grad_norm": 3.587539918501533,
+      "learning_rate": 1.1041979005806876e-05,
+      "loss": 2.8766,
+      "step": 2510
+    },
+    {
+      "epoch": 0.48309364628925977,
+      "grad_norm": 3.732267437004394,
+      "learning_rate": 1.1035780645449001e-05,
+      "loss": 3.1287,
+      "step": 2511
+    },
+    {
+      "epoch": 0.48328603722764657,
+      "grad_norm": 2.3506648768053813,
+      "learning_rate": 1.1029581882805857e-05,
+      "loss": 2.6986,
+      "step": 2512
+    },
+    {
+      "epoch": 0.48347842816603337,
+      "grad_norm": 3.336670613774814,
+      "learning_rate": 1.1023382720284973e-05,
+      "loss": 3.1843,
+      "step": 2513
+    },
+    {
+      "epoch": 0.48367081910442017,
+      "grad_norm": 2.242024731090763,
+      "learning_rate": 1.1017183160294033e-05,
+      "loss": 2.6264,
+      "step": 2514
+    },
+    {
+      "epoch": 0.48386321004280697,
+      "grad_norm": 3.4219059020964067,
+      "learning_rate": 1.1010983205240878e-05,
+      "loss": 3.0028,
+      "step": 2515
+    },
+    {
+      "epoch": 0.48405560098119377,
+      "grad_norm": 3.4413084295508987,
+      "learning_rate": 1.1004782857533488e-05,
+      "loss": 3.048,
+      "step": 2516
+    },
+    {
+      "epoch": 0.48424799191958057,
+      "grad_norm": 3.643709349307844,
+      "learning_rate": 1.099858211958002e-05,
+      "loss": 3.1892,
+      "step": 2517
+    },
+    {
+      "epoch": 0.48444038285796737,
+      "grad_norm": 3.420587357458005,
+      "learning_rate": 1.0992380993788763e-05,
+      "loss": 3.1977,
+      "step": 2518
+    },
+    {
+      "epoch": 0.48463277379635417,
+      "grad_norm": 3.569748802790544,
+      "learning_rate": 1.0986179482568162e-05,
+      "loss": 3.2599,
+      "step": 2519
+    },
+    {
+      "epoch": 0.484825164734741,
+      "grad_norm": 3.6779344506200418,
+      "learning_rate": 1.0979977588326815e-05,
+      "loss": 3.228,
+      "step": 2520
+    },
+    {
+      "epoch": 0.4850175556731278,
+      "grad_norm": 3.509114030573855,
+      "learning_rate": 1.0973775313473465e-05,
+      "loss": 3.4905,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4852099466115146,
+      "grad_norm": 3.383707066371638,
+      "learning_rate": 1.0967572660417001e-05,
+      "loss": 3.2384,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4854023375499014,
+      "grad_norm": 3.5631292422699836,
+      "learning_rate": 1.0961369631566468e-05,
+      "loss": 3.2749,
+      "step": 2523
+    },
+    {
+      "epoch": 0.4855947284882882,
+      "grad_norm": 3.563015752849152,
+      "learning_rate": 1.0955166229331048e-05,
+      "loss": 3.0872,
+      "step": 2524
+    },
+    {
+      "epoch": 0.485787119426675,
+      "grad_norm": 3.723404609091628,
+      "learning_rate": 1.0948962456120068e-05,
+      "loss": 3.1215,
+      "step": 2525
+    },
+    {
+      "epoch": 0.4859795103650618,
+      "grad_norm": 3.5850424329822177,
+      "learning_rate": 1.0942758314343007e-05,
+      "loss": 3.2306,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4861719013034486,
+      "grad_norm": 3.4494566863330833,
+      "learning_rate": 1.0936553806409482e-05,
+      "loss": 3.038,
+      "step": 2527
+    },
+    {
+      "epoch": 0.4863642922418354,
+      "grad_norm": 2.2335435947467177,
+      "learning_rate": 1.0930348934729249e-05,
+      "loss": 2.5988,
+      "step": 2528
+    },
+    {
+      "epoch": 0.4865566831802222,
+      "grad_norm": 3.2923039704636583,
+      "learning_rate": 1.0924143701712211e-05,
+      "loss": 3.2955,
+      "step": 2529
+    },
+    {
+      "epoch": 0.486749074118609,
+      "grad_norm": 3.453535268851059,
+      "learning_rate": 1.0917938109768404e-05,
+      "loss": 3.2453,
+      "step": 2530
+    },
+    {
+      "epoch": 0.4869414650569958,
+      "grad_norm": 2.2862838932720853,
+      "learning_rate": 1.0911732161308014e-05,
+      "loss": 2.5178,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4871338559953826,
+      "grad_norm": 3.3794286706907157,
+      "learning_rate": 1.0905525858741364e-05,
+      "loss": 2.8518,
+      "step": 2532
+    },
+    {
+      "epoch": 0.4873262469337694,
+      "grad_norm": 3.2992879899258085,
+      "learning_rate": 1.08993192044789e-05,
+      "loss": 2.881,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4875186378721562,
+      "grad_norm": 3.39697300853983,
+      "learning_rate": 1.089311220093122e-05,
+      "loss": 3.1156,
+      "step": 2534
+    },
+    {
+      "epoch": 0.487711028810543,
+      "grad_norm": 3.4670511634932266,
+      "learning_rate": 1.0886904850509052e-05,
+      "loss": 3.2412,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4879034197489298,
+      "grad_norm": 3.4371398544709453,
+      "learning_rate": 1.0880697155623264e-05,
+      "loss": 3.1987,
+      "step": 2536
+    },
+    {
+      "epoch": 0.4880958106873166,
+      "grad_norm": 3.76710466638068,
+      "learning_rate": 1.0874489118684846e-05,
+      "loss": 3.2895,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4882882016257034,
+      "grad_norm": 3.5320782844234277,
+      "learning_rate": 1.086828074210493e-05,
+      "loss": 3.0427,
+      "step": 2538
+    },
+    {
+      "epoch": 0.4884805925640902,
+      "grad_norm": 3.127386030705917,
+      "learning_rate": 1.0862072028294777e-05,
+      "loss": 3.1326,
+      "step": 2539
+    },
+    {
+      "epoch": 0.488672983502477,
+      "grad_norm": 3.544449216861918,
+      "learning_rate": 1.0855862979665788e-05,
+      "loss": 3.0456,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4888653744408638,
+      "grad_norm": 3.282706197320416,
+      "learning_rate": 1.0849653598629477e-05,
+      "loss": 3.0015,
+      "step": 2541
+    },
+    {
+      "epoch": 0.48905776537925066,
+      "grad_norm": 3.4779783582707493,
+      "learning_rate": 1.0843443887597495e-05,
+      "loss": 3.1234,
+      "step": 2542
+    },
+    {
+      "epoch": 0.48925015631763746,
+      "grad_norm": 3.495265668313333,
+      "learning_rate": 1.0837233848981632e-05,
+      "loss": 3.011,
+      "step": 2543
+    },
+    {
+      "epoch": 0.48944254725602426,
+      "grad_norm": 2.4080096785311462,
+      "learning_rate": 1.0831023485193787e-05,
+      "loss": 2.5854,
+      "step": 2544
+    },
+    {
+      "epoch": 0.48963493819441106,
+      "grad_norm": 3.5102169893413913,
+      "learning_rate": 1.0824812798645997e-05,
+      "loss": 3.0515,
+      "step": 2545
+    },
+    {
+      "epoch": 0.48982732913279786,
+      "grad_norm": 3.0461517143636803,
+      "learning_rate": 1.0818601791750418e-05,
+      "loss": 2.9744,
+      "step": 2546
+    },
+    {
+      "epoch": 0.49001972007118466,
+      "grad_norm": 3.2419906148186812,
+      "learning_rate": 1.0812390466919337e-05,
+      "loss": 3.0964,
+      "step": 2547
+    },
+    {
+      "epoch": 0.49021211100957146,
+      "grad_norm": 3.419397061818397,
+      "learning_rate": 1.0806178826565162e-05,
+      "loss": 3.1467,
+      "step": 2548
+    },
+    {
+      "epoch": 0.49040450194795826,
+      "grad_norm": 3.6204958166470584,
+      "learning_rate": 1.0799966873100419e-05,
+      "loss": 2.9372,
+      "step": 2549
+    },
+    {
+      "epoch": 0.49059689288634506,
+      "grad_norm": 3.4261785647259826,
+      "learning_rate": 1.0793754608937758e-05,
+      "loss": 3.1768,
+      "step": 2550
+    },
+    {
+      "epoch": 0.49078928382473186,
+      "grad_norm": 3.6190211196552027,
+      "learning_rate": 1.0787542036489955e-05,
+      "loss": 3.0711,
+      "step": 2551
+    },
+    {
+      "epoch": 0.49098167476311866,
+      "grad_norm": 3.371496770283932,
+      "learning_rate": 1.0781329158169902e-05,
+      "loss": 3.3522,
+      "step": 2552
+    },
+    {
+      "epoch": 0.49117406570150546,
+      "grad_norm": 3.731235810280709,
+      "learning_rate": 1.0775115976390607e-05,
+      "loss": 3.0409,
+      "step": 2553
+    },
+    {
+      "epoch": 0.49136645663989226,
+      "grad_norm": 3.780176029210412,
+      "learning_rate": 1.0768902493565197e-05,
+      "loss": 3.1509,
+      "step": 2554
+    },
+    {
+      "epoch": 0.49155884757827906,
+      "grad_norm": 3.5863763674664844,
+      "learning_rate": 1.0762688712106918e-05,
+      "loss": 2.9596,
+      "step": 2555
+    },
+    {
+      "epoch": 0.49175123851666586,
+      "grad_norm": 3.7997034588720715,
+      "learning_rate": 1.0756474634429133e-05,
+      "loss": 3.3145,
+      "step": 2556
+    },
+    {
+      "epoch": 0.49194362945505266,
+      "grad_norm": 3.3050691026143206,
+      "learning_rate": 1.0750260262945314e-05,
+      "loss": 3.0313,
+      "step": 2557
+    },
+    {
+      "epoch": 0.49213602039343946,
+      "grad_norm": 3.4103980642200153,
+      "learning_rate": 1.0744045600069055e-05,
+      "loss": 2.9928,
+      "step": 2558
+    },
+    {
+      "epoch": 0.49232841133182625,
+      "grad_norm": 3.580330758744163,
+      "learning_rate": 1.0737830648214063e-05,
+      "loss": 3.0999,
+      "step": 2559
+    },
+    {
+      "epoch": 0.49252080227021305,
+      "grad_norm": 3.5280653215765723,
+      "learning_rate": 1.0731615409794144e-05,
+      "loss": 3.0391,
+      "step": 2560
+    },
+    {
+      "epoch": 0.49271319320859985,
+      "grad_norm": 3.473416788431021,
+      "learning_rate": 1.0725399887223234e-05,
+      "loss": 2.9906,
+      "step": 2561
+    },
+    {
+      "epoch": 0.49290558414698665,
+      "grad_norm": 3.5434405796241806,
+      "learning_rate": 1.0719184082915364e-05,
+      "loss": 3.0296,
+      "step": 2562
+    },
+    {
+      "epoch": 0.49309797508537345,
+      "grad_norm": 3.359942622029441,
+      "learning_rate": 1.0712967999284682e-05,
+      "loss": 3.1283,
+      "step": 2563
+    },
+    {
+      "epoch": 0.4932903660237603,
+      "grad_norm": 3.8259655461231232,
+      "learning_rate": 1.0706751638745448e-05,
+      "loss": 3.2278,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4934827569621471,
+      "grad_norm": 3.4816234931917553,
+      "learning_rate": 1.0700535003712023e-05,
+      "loss": 3.0868,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4936751479005339,
+      "grad_norm": 2.3803443684720382,
+      "learning_rate": 1.069431809659887e-05,
+      "loss": 2.6108,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4938675388389207,
+      "grad_norm": 3.3480521530324623,
+      "learning_rate": 1.068810091982057e-05,
+      "loss": 3.0464,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4940599297773075,
+      "grad_norm": 3.5233486807080023,
+      "learning_rate": 1.0681883475791803e-05,
+      "loss": 3.1129,
+      "step": 2568
+    },
+    {
+      "epoch": 0.4942523207156943,
+      "grad_norm": 3.465122721146094,
+      "learning_rate": 1.067566576692735e-05,
+      "loss": 3.2592,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4944447116540811,
+      "grad_norm": 3.4133027923640156,
+      "learning_rate": 1.0669447795642103e-05,
+      "loss": 3.0093,
+      "step": 2570
+    },
+    {
+      "epoch": 0.4946371025924679,
+      "grad_norm": 3.24303019439944,
+      "learning_rate": 1.066322956435104e-05,
+      "loss": 3.2187,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4948294935308547,
+      "grad_norm": 3.381642460372946,
+      "learning_rate": 1.065701107546926e-05,
+      "loss": 3.212,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4950218844692415,
+      "grad_norm": 3.5141863980187984,
+      "learning_rate": 1.065079233141195e-05,
+      "loss": 3.3454,
+      "step": 2573
+    },
+    {
+      "epoch": 0.4952142754076283,
+      "grad_norm": 3.4433746954955287,
+      "learning_rate": 1.0644573334594395e-05,
+      "loss": 3.0941,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4954066663460151,
+      "grad_norm": 3.3664233941365476,
+      "learning_rate": 1.0638354087431986e-05,
+      "loss": 3.1938,
+      "step": 2575
+    },
+    {
+      "epoch": 0.4955990572844019,
+      "grad_norm": 3.3662104985972587,
+      "learning_rate": 1.0632134592340204e-05,
+      "loss": 2.9386,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4957914482227887,
+      "grad_norm": 3.5968588178354466,
+      "learning_rate": 1.0625914851734632e-05,
+      "loss": 3.0136,
+      "step": 2577
+    },
+    {
+      "epoch": 0.4959838391611755,
+      "grad_norm": 3.4538385895234516,
+      "learning_rate": 1.0619694868030943e-05,
+      "loss": 3.0579,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4961762300995623,
+      "grad_norm": 3.07062222499145,
+      "learning_rate": 1.0613474643644907e-05,
+      "loss": 2.7443,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4963686210379491,
+      "grad_norm": 2.29007955208693,
+      "learning_rate": 1.0607254180992391e-05,
+      "loss": 2.5154,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4965610119763359,
+      "grad_norm": 3.6040389200718375,
+      "learning_rate": 1.0601033482489346e-05,
+      "loss": 3.422,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4967534029147227,
+      "grad_norm": 3.623631631267286,
+      "learning_rate": 1.0594812550551826e-05,
+      "loss": 3.1359,
+      "step": 2582
+    },
+    {
+      "epoch": 0.4969457938531095,
+      "grad_norm": 3.442630567817095,
+      "learning_rate": 1.058859138759596e-05,
+      "loss": 3.0966,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4971381847914963,
+      "grad_norm": 3.255855621085257,
+      "learning_rate": 1.0582369996037985e-05,
+      "loss": 3.0067,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4973305757298831,
+      "grad_norm": 3.2762244264761415,
+      "learning_rate": 1.0576148378294213e-05,
+      "loss": 3.0712,
+      "step": 2585
+    },
+    {
+      "epoch": 0.49752296666826995,
+      "grad_norm": 3.4215835100874714,
+      "learning_rate": 1.056992653678105e-05,
+      "loss": 3.1344,
+      "step": 2586
+    },
+    {
+      "epoch": 0.49771535760665675,
+      "grad_norm": 2.296480048436712,
+      "learning_rate": 1.0563704473914986e-05,
+      "loss": 2.6498,
+      "step": 2587
+    },
+    {
+      "epoch": 0.49790774854504355,
+      "grad_norm": 3.438773387768573,
+      "learning_rate": 1.0557482192112603e-05,
+      "loss": 3.2462,
+      "step": 2588
+    },
+    {
+      "epoch": 0.49810013948343035,
+      "grad_norm": 3.4280996658002727,
+      "learning_rate": 1.0551259693790556e-05,
+      "loss": 3.1755,
+      "step": 2589
+    },
+    {
+      "epoch": 0.49829253042181715,
+      "grad_norm": 3.5360360042070944,
+      "learning_rate": 1.0545036981365601e-05,
+      "loss": 3.2038,
+      "step": 2590
+    },
+    {
+      "epoch": 0.49848492136020395,
+      "grad_norm": 3.612698886708627,
+      "learning_rate": 1.053881405725456e-05,
+      "loss": 3.3051,
+      "step": 2591
+    },
+    {
+      "epoch": 0.49867731229859075,
+      "grad_norm": 3.737211750153202,
+      "learning_rate": 1.0532590923874349e-05,
+      "loss": 3.3243,
+      "step": 2592
+    },
+    {
+      "epoch": 0.49886970323697755,
+      "grad_norm": 3.421894304634954,
+      "learning_rate": 1.0526367583641958e-05,
+      "loss": 2.6858,
+      "step": 2593
+    },
+    {
+      "epoch": 0.49906209417536435,
+      "grad_norm": 3.620500175559969,
+      "learning_rate": 1.0520144038974468e-05,
+      "loss": 3.13,
+      "step": 2594
+    },
+    {
+      "epoch": 0.49925448511375115,
+      "grad_norm": 3.337346891643855,
+      "learning_rate": 1.0513920292289021e-05,
+      "loss": 3.3082,
+      "step": 2595
+    },
+    {
+      "epoch": 0.49944687605213794,
+      "grad_norm": 2.43946254081732,
+      "learning_rate": 1.0507696346002857e-05,
+      "loss": 2.6056,
+      "step": 2596
+    },
+    {
+      "epoch": 0.49963926699052474,
+      "grad_norm": 3.3782698509881004,
+      "learning_rate": 1.0501472202533285e-05,
+      "loss": 3.1224,
+      "step": 2597
+    },
+    {
+      "epoch": 0.49983165792891154,
+      "grad_norm": 3.85302377644686,
+      "learning_rate": 1.0495247864297684e-05,
+      "loss": 3.1984,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5000240488672983,
+      "grad_norm": 3.668502280167047,
+      "learning_rate": 1.0489023333713522e-05,
+      "loss": 3.047,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5002164398056852,
+      "grad_norm": 3.3519924534831467,
+      "learning_rate": 1.0482798613198328e-05,
+      "loss": 2.8539,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5004088307440719,
+      "grad_norm": 3.4161130469948717,
+      "learning_rate": 1.047657370516972e-05,
+      "loss": 3.3661,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5006012216824588,
+      "grad_norm": 3.5663736324561635,
+      "learning_rate": 1.0470348612045376e-05,
+      "loss": 3.052,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5007936126208455,
+      "grad_norm": 3.387204097195989,
+      "learning_rate": 1.0464123336243049e-05,
+      "loss": 2.9418,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5009860035592324,
+      "grad_norm": 3.7584152832505935,
+      "learning_rate": 1.0457897880180566e-05,
+      "loss": 2.9898,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5011783944976191,
+      "grad_norm": 3.330532539739618,
+      "learning_rate": 1.0451672246275826e-05,
+      "loss": 3.1803,
+      "step": 2605
+    },
+    {
+      "epoch": 0.501370785436006,
+      "grad_norm": 3.357133813535198,
+      "learning_rate": 1.0445446436946788e-05,
+      "loss": 3.2516,
+      "step": 2606
+    },
+    {
+      "epoch": 0.5015631763743927,
+      "grad_norm": 3.5302539692663895,
+      "learning_rate": 1.0439220454611486e-05,
+      "loss": 2.9395,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5017555673127796,
+      "grad_norm": 3.386257407970631,
+      "learning_rate": 1.0432994301688021e-05,
+      "loss": 3.0178,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5019479582511663,
+      "grad_norm": 3.331197447587106,
+      "learning_rate": 1.0426767980594559e-05,
+      "loss": 3.1554,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5021403491895532,
+      "grad_norm": 3.426009946996359,
+      "learning_rate": 1.0420541493749332e-05,
+      "loss": 2.8475,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5023327401279399,
+      "grad_norm": 3.3140645018270494,
+      "learning_rate": 1.0414314843570634e-05,
+      "loss": 2.9541,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5025251310663268,
+      "grad_norm": 3.3927748733919256,
+      "learning_rate": 1.0408088032476822e-05,
+      "loss": 3.0668,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5027175220047135,
+      "grad_norm": 3.2866454337219593,
+      "learning_rate": 1.0401861062886324e-05,
+      "loss": 3.003,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5029099129431004,
+      "grad_norm": 3.4621626102072605,
+      "learning_rate": 1.0395633937217622e-05,
+      "loss": 3.4265,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5031023038814871,
+      "grad_norm": 3.3859407643930934,
+      "learning_rate": 1.0389406657889254e-05,
+      "loss": 3.0892,
+      "step": 2615
+    },
+    {
+      "epoch": 0.503294694819874,
+      "grad_norm": 3.5582152121376875,
+      "learning_rate": 1.0383179227319826e-05,
+      "loss": 3.3374,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5034870857582607,
+      "grad_norm": 3.349095302379577,
+      "learning_rate": 1.0376951647928007e-05,
+      "loss": 3.2141,
+      "step": 2617
+    },
+    {
+      "epoch": 0.5036794766966476,
+      "grad_norm": 3.4265520168368475,
+      "learning_rate": 1.0370723922132506e-05,
+      "loss": 3.1717,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5038718676350344,
+      "grad_norm": 3.4447894458214976,
+      "learning_rate": 1.036449605235211e-05,
+      "loss": 3.0014,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5040642585734212,
+      "grad_norm": 3.6215702795502662,
+      "learning_rate": 1.0358268041005644e-05,
+      "loss": 2.9649,
+      "step": 2620
+    },
+    {
+      "epoch": 0.504256649511808,
+      "grad_norm": 3.450305562950398,
+      "learning_rate": 1.0352039890511997e-05,
+      "loss": 3.0046,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5044490404501948,
+      "grad_norm": 3.343047204078497,
+      "learning_rate": 1.034581160329012e-05,
+      "loss": 3.2642,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5046414313885816,
+      "grad_norm": 3.5874445362285523,
+      "learning_rate": 1.0339583181758997e-05,
+      "loss": 2.8002,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5048338223269684,
+      "grad_norm": 3.5248627448364913,
+      "learning_rate": 1.033335462833768e-05,
+      "loss": 3.0934,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5050262132653552,
+      "grad_norm": 3.4120790217130526,
+      "learning_rate": 1.0327125945445265e-05,
+      "loss": 3.1239,
+      "step": 2625
+    },
+    {
+      "epoch": 0.505218604203742,
+      "grad_norm": 3.221764275022081,
+      "learning_rate": 1.0320897135500904e-05,
+      "loss": 3.0474,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5054109951421288,
+      "grad_norm": 3.3510372150973864,
+      "learning_rate": 1.0314668200923791e-05,
+      "loss": 3.1879,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5056033860805156,
+      "grad_norm": 3.579919604518649,
+      "learning_rate": 1.0308439144133177e-05,
+      "loss": 3.3239,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5057957770189024,
+      "grad_norm": 3.2079938720521746,
+      "learning_rate": 1.0302209967548354e-05,
+      "loss": 3.0713,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5059881679572892,
+      "grad_norm": 3.7298195282674085,
+      "learning_rate": 1.029598067358866e-05,
+      "loss": 3.0293,
+      "step": 2630
+    },
+    {
+      "epoch": 0.506180558895676,
+      "grad_norm": 3.572280402611658,
+      "learning_rate": 1.0289751264673485e-05,
+      "loss": 3.0467,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5063729498340628,
+      "grad_norm": 3.7323562265795722,
+      "learning_rate": 1.0283521743222256e-05,
+      "loss": 3.0591,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5065653407724496,
+      "grad_norm": 3.2328940133823556,
+      "learning_rate": 1.0277292111654447e-05,
+      "loss": 3.1589,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5067577317108364,
+      "grad_norm": 3.4457221293133444,
+      "learning_rate": 1.0271062372389582e-05,
+      "loss": 3.226,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5069501226492232,
+      "grad_norm": 2.331855725779631,
+      "learning_rate": 1.0264832527847212e-05,
+      "loss": 2.6838,
+      "step": 2635
+    },
+    {
+      "epoch": 0.50714251358761,
+      "grad_norm": 3.412697657756914,
+      "learning_rate": 1.0258602580446941e-05,
+      "loss": 2.9674,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5073349045259968,
+      "grad_norm": 3.4500658506138064,
+      "learning_rate": 1.0252372532608405e-05,
+      "loss": 3.1928,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5075272954643836,
+      "grad_norm": 2.2990676781985884,
+      "learning_rate": 1.024614238675129e-05,
+      "loss": 2.6374,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5077196864027704,
+      "grad_norm": 3.239432512287971,
+      "learning_rate": 1.0239912145295303e-05,
+      "loss": 3.0717,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5079120773411573,
+      "grad_norm": 3.6197681691699803,
+      "learning_rate": 1.0233681810660207e-05,
+      "loss": 2.9228,
+      "step": 2640
+    },
+    {
+      "epoch": 0.508104468279544,
+      "grad_norm": 2.3388951079887286,
+      "learning_rate": 1.0227451385265788e-05,
+      "loss": 2.6815,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5082968592179309,
+      "grad_norm": 3.8506442817550357,
+      "learning_rate": 1.022122087153187e-05,
+      "loss": 3.2076,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5084892501563176,
+      "grad_norm": 3.553447746280496,
+      "learning_rate": 1.0214990271878319e-05,
+      "loss": 3.0072,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5086816410947045,
+      "grad_norm": 3.3644580033246476,
+      "learning_rate": 1.0208759588725016e-05,
+      "loss": 3.1174,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5088740320330912,
+      "grad_norm": 3.652383902181383,
+      "learning_rate": 1.0202528824491899e-05,
+      "loss": 3.295,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5090664229714781,
+      "grad_norm": 3.5417319478658174,
+      "learning_rate": 1.0196297981598921e-05,
+      "loss": 3.1133,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5092588139098648,
+      "grad_norm": 3.4565290810608835,
+      "learning_rate": 1.019006706246607e-05,
+      "loss": 3.2014,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5094512048482517,
+      "grad_norm": 3.5993323799167016,
+      "learning_rate": 1.018383606951336e-05,
+      "loss": 3.0913,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5096435957866384,
+      "grad_norm": 3.5152967630253893,
+      "learning_rate": 1.0177605005160837e-05,
+      "loss": 3.0699,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5098359867250253,
+      "grad_norm": 3.6072797520352693,
+      "learning_rate": 1.0171373871828578e-05,
+      "loss": 3.2379,
+      "step": 2650
+    },
+    {
+      "epoch": 0.510028377663412,
+      "grad_norm": 3.692177852829844,
+      "learning_rate": 1.0165142671936685e-05,
+      "loss": 3.1464,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5102207686017989,
+      "grad_norm": 3.55695184865762,
+      "learning_rate": 1.0158911407905279e-05,
+      "loss": 2.8528,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5104131595401856,
+      "grad_norm": 3.7501544945644656,
+      "learning_rate": 1.0152680082154514e-05,
+      "loss": 3.1443,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5106055504785725,
+      "grad_norm": 3.2837446588555568,
+      "learning_rate": 1.0146448697104561e-05,
+      "loss": 3.0577,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5107979414169592,
+      "grad_norm": 3.69581489262806,
+      "learning_rate": 1.0140217255175626e-05,
+      "loss": 3.1165,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5109903323553461,
+      "grad_norm": 3.8327951105392977,
+      "learning_rate": 1.013398575878792e-05,
+      "loss": 3.093,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5111827232937328,
+      "grad_norm": 3.456307397790415,
+      "learning_rate": 1.0127754210361694e-05,
+      "loss": 2.8424,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5113751142321197,
+      "grad_norm": 3.6188194794436126,
+      "learning_rate": 1.0121522612317204e-05,
+      "loss": 3.1335,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5115675051705064,
+      "grad_norm": 3.3521593079660916,
+      "learning_rate": 1.011529096707473e-05,
+      "loss": 3.3521,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5117598961088933,
+      "grad_norm": 2.5297125431804117,
+      "learning_rate": 1.0109059277054574e-05,
+      "loss": 2.6181,
+      "step": 2660
+    },
+    {
+      "epoch": 0.51195228704728,
+      "grad_norm": 3.3775052475519347,
+      "learning_rate": 1.010282754467705e-05,
+      "loss": 3.291,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5121446779856669,
+      "grad_norm": 3.4535983768413594,
+      "learning_rate": 1.0096595772362492e-05,
+      "loss": 3.2639,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5123370689240537,
+      "grad_norm": 3.364384291697941,
+      "learning_rate": 1.0090363962531251e-05,
+      "loss": 3.3143,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5125294598624405,
+      "grad_norm": 2.42222250775476,
+      "learning_rate": 1.0084132117603689e-05,
+      "loss": 2.6546,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5127218508008273,
+      "grad_norm": 3.1653484446493927,
+      "learning_rate": 1.0077900240000181e-05,
+      "loss": 3.252,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5129142417392141,
+      "grad_norm": 3.6212253372161465,
+      "learning_rate": 1.0071668332141115e-05,
+      "loss": 3.0896,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5131066326776009,
+      "grad_norm": 3.6901027846766445,
+      "learning_rate": 1.0065436396446899e-05,
+      "loss": 3.135,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5132990236159877,
+      "grad_norm": 3.6095414453355716,
+      "learning_rate": 1.0059204435337938e-05,
+      "loss": 3.2213,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5134914145543745,
+      "grad_norm": 3.3936757281700483,
+      "learning_rate": 1.0052972451234656e-05,
+      "loss": 3.1589,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5136838054927613,
+      "grad_norm": 3.2948712890642766,
+      "learning_rate": 1.0046740446557485e-05,
+      "loss": 3.1267,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5138761964311481,
+      "grad_norm": 3.3581185307502457,
+      "learning_rate": 1.0040508423726865e-05,
+      "loss": 3.1666,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5140685873695349,
+      "grad_norm": 3.3832036569498007,
+      "learning_rate": 1.0034276385163238e-05,
+      "loss": 3.0518,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5142609783079217,
+      "grad_norm": 3.394669113804863,
+      "learning_rate": 1.0028044333287056e-05,
+      "loss": 3.1192,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5144533692463085,
+      "grad_norm": 3.450279270317119,
+      "learning_rate": 1.002181227051878e-05,
+      "loss": 3.2413,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5146457601846953,
+      "grad_norm": 3.3590752413158356,
+      "learning_rate": 1.0015580199278873e-05,
+      "loss": 3.1076,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5148381511230821,
+      "grad_norm": 3.2510613787586813,
+      "learning_rate": 1.0009348121987795e-05,
+      "loss": 2.9988,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5150305420614689,
+      "grad_norm": 3.337132656842642,
+      "learning_rate": 1.000311604106601e-05,
+      "loss": 3.2467,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5152229329998557,
+      "grad_norm": 3.375191801629267,
+      "learning_rate": 9.996883958933993e-06,
+      "loss": 2.7839,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5154153239382425,
+      "grad_norm": 3.576841740148076,
+      "learning_rate": 9.99065187801221e-06,
+      "loss": 3.2018,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5156077148766293,
+      "grad_norm": 3.379704816884023,
+      "learning_rate": 9.984419800721132e-06,
+      "loss": 2.9875,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5158001058150161,
+      "grad_norm": 3.504461771870136,
+      "learning_rate": 9.978187729481218e-06,
+      "loss": 3.2016,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5159924967534029,
+      "grad_norm": 3.369547596007173,
+      "learning_rate": 9.971955666712945e-06,
+      "loss": 3.0531,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5161848876917897,
+      "grad_norm": 3.5549450979230026,
+      "learning_rate": 9.965723614836764e-06,
+      "loss": 3.15,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5163772786301766,
+      "grad_norm": 3.474676804665959,
+      "learning_rate": 9.959491576273139e-06,
+      "loss": 3.2728,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5165696695685633,
+      "grad_norm": 3.60164308421793,
+      "learning_rate": 9.95325955344252e-06,
+      "loss": 3.1474,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5167620605069502,
+      "grad_norm": 3.7062045332031723,
+      "learning_rate": 9.947027548765347e-06,
+      "loss": 3.4299,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5169544514453369,
+      "grad_norm": 3.495296693232079,
+      "learning_rate": 9.940795564662064e-06,
+      "loss": 3.1661,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5171468423837238,
+      "grad_norm": 3.7671135468969776,
+      "learning_rate": 9.934563603553103e-06,
+      "loss": 3.0835,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5173392333221105,
+      "grad_norm": 3.422476063889177,
+      "learning_rate": 9.928331667858886e-06,
+      "loss": 3.1792,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5175316242604974,
+      "grad_norm": 3.650028069620823,
+      "learning_rate": 9.922099759999822e-06,
+      "loss": 3.0053,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5177240151988841,
+      "grad_norm": 2.5076271151885363,
+      "learning_rate": 9.915867882396314e-06,
+      "loss": 2.61,
+      "step": 2691
+    },
+    {
+      "epoch": 0.517916406137271,
+      "grad_norm": 3.4787864907855313,
+      "learning_rate": 9.909636037468754e-06,
+      "loss": 3.1716,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5181087970756577,
+      "grad_norm": 3.6197491161304014,
+      "learning_rate": 9.90340422763751e-06,
+      "loss": 3.0821,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5183011880140446,
+      "grad_norm": 3.529962521537279,
+      "learning_rate": 9.897172455322953e-06,
+      "loss": 3.1328,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5184935789524313,
+      "grad_norm": 3.462187811694469,
+      "learning_rate": 9.890940722945429e-06,
+      "loss": 2.9558,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5186859698908182,
+      "grad_norm": 3.417751274689193,
+      "learning_rate": 9.884709032925274e-06,
+      "loss": 3.4369,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5188783608292049,
+      "grad_norm": 3.4085427664061054,
+      "learning_rate": 9.878477387682801e-06,
+      "loss": 2.9918,
+      "step": 2697
+    },
+    {
+      "epoch": 0.5190707517675918,
+      "grad_norm": 3.4409035173477283,
+      "learning_rate": 9.872245789638308e-06,
+      "loss": 2.8579,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5192631427059785,
+      "grad_norm": 3.195804538793157,
+      "learning_rate": 9.866014241212078e-06,
+      "loss": 3.0375,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5194555336443654,
+      "grad_norm": 3.641926187169992,
+      "learning_rate": 9.859782744824376e-06,
+      "loss": 3.1062,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5196479245827521,
+      "grad_norm": 3.4315128219816238,
+      "learning_rate": 9.85355130289544e-06,
+      "loss": 3.1071,
+      "step": 2701
+    },
+    {
+      "epoch": 0.519840315521139,
+      "grad_norm": 3.4541722894730227,
+      "learning_rate": 9.84731991784549e-06,
+      "loss": 3.0293,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5200327064595257,
+      "grad_norm": 3.2131525313197686,
+      "learning_rate": 9.841088592094726e-06,
+      "loss": 2.9493,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5202250973979126,
+      "grad_norm": 3.307432682922727,
+      "learning_rate": 9.834857328063316e-06,
+      "loss": 2.997,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5204174883362993,
+      "grad_norm": 3.3827652651726106,
+      "learning_rate": 9.828626128171422e-06,
+      "loss": 2.9235,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5206098792746862,
+      "grad_norm": 3.3338447297934897,
+      "learning_rate": 9.822394994839164e-06,
+      "loss": 3.0652,
+      "step": 2706
+    },
+    {
+      "epoch": 0.520802270213073,
+      "grad_norm": 3.3661392931480068,
+      "learning_rate": 9.816163930486643e-06,
+      "loss": 2.9875,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5209946611514598,
+      "grad_norm": 3.2531906638385006,
+      "learning_rate": 9.809932937533935e-06,
+      "loss": 3.2335,
+      "step": 2708
+    },
+    {
+      "epoch": 0.5211870520898466,
+      "grad_norm": 3.5886205935910582,
+      "learning_rate": 9.803702018401084e-06,
+      "loss": 3.1691,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5213794430282334,
+      "grad_norm": 3.1989498565918573,
+      "learning_rate": 9.797471175508101e-06,
+      "loss": 2.9589,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5215718339666202,
+      "grad_norm": 3.2514642414642116,
+      "learning_rate": 9.791240411274982e-06,
+      "loss": 2.8258,
+      "step": 2711
+    },
+    {
+      "epoch": 0.521764224905007,
+      "grad_norm": 3.1788316118201165,
+      "learning_rate": 9.785009728121686e-06,
+      "loss": 2.988,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5219566158433938,
+      "grad_norm": 3.369236935840319,
+      "learning_rate": 9.778779128468133e-06,
+      "loss": 2.9786,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5221490067817806,
+      "grad_norm": 3.2301381800324847,
+      "learning_rate": 9.772548614734217e-06,
+      "loss": 3.0674,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5223413977201674,
+      "grad_norm": 2.3808902948606048,
+      "learning_rate": 9.766318189339798e-06,
+      "loss": 2.6102,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5225337886585542,
+      "grad_norm": 3.2711466923973087,
+      "learning_rate": 9.760087854704697e-06,
+      "loss": 3.2669,
+      "step": 2716
+    },
+    {
+      "epoch": 0.522726179596941,
+      "grad_norm": 3.5882446753534083,
+      "learning_rate": 9.753857613248714e-06,
+      "loss": 3.2122,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5229185705353278,
+      "grad_norm": 3.3461529942836226,
+      "learning_rate": 9.747627467391596e-06,
+      "loss": 2.964,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5231109614737146,
+      "grad_norm": 3.3844653191315768,
+      "learning_rate": 9.741397419553062e-06,
+      "loss": 3.195,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5233033524121014,
+      "grad_norm": 3.1720146699804483,
+      "learning_rate": 9.735167472152793e-06,
+      "loss": 3.3264,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5234957433504882,
+      "grad_norm": 3.3603330424363325,
+      "learning_rate": 9.728937627610425e-06,
+      "loss": 3.3658,
+      "step": 2721
+    },
+    {
+      "epoch": 0.523688134288875,
+      "grad_norm": 3.686308280375822,
+      "learning_rate": 9.722707888345553e-06,
+      "loss": 3.1661,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5238805252272618,
+      "grad_norm": 3.762415206328054,
+      "learning_rate": 9.716478256777749e-06,
+      "loss": 3.0938,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5240729161656486,
+      "grad_norm": 3.5420248716235303,
+      "learning_rate": 9.710248735326519e-06,
+      "loss": 3.2665,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5242653071040354,
+      "grad_norm": 3.2110667768158265,
+      "learning_rate": 9.704019326411344e-06,
+      "loss": 3.015,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5244576980424221,
+      "grad_norm": 3.5461421455068822,
+      "learning_rate": 9.697790032451651e-06,
+      "loss": 3.2258,
+      "step": 2726
+    },
+    {
+      "epoch": 0.524650088980809,
+      "grad_norm": 3.2569757725963986,
+      "learning_rate": 9.691560855866826e-06,
+      "loss": 2.8815,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5248424799191959,
+      "grad_norm": 3.6080752590785723,
+      "learning_rate": 9.685331799076208e-06,
+      "loss": 2.8357,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5250348708575826,
+      "grad_norm": 3.2793838294972817,
+      "learning_rate": 9.6791028644991e-06,
+      "loss": 3.1663,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5252272617959695,
+      "grad_norm": 3.479829081606758,
+      "learning_rate": 9.672874054554738e-06,
+      "loss": 3.1788,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5254196527343562,
+      "grad_norm": 3.1389250334747105,
+      "learning_rate": 9.666645371662324e-06,
+      "loss": 2.9736,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5256120436727431,
+      "grad_norm": 3.2249625837144564,
+      "learning_rate": 9.660416818241007e-06,
+      "loss": 3.0028,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5258044346111298,
+      "grad_norm": 3.5303486596613136,
+      "learning_rate": 9.654188396709882e-06,
+      "loss": 3.1416,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5259968255495167,
+      "grad_norm": 3.575025880076746,
+      "learning_rate": 9.647960109488003e-06,
+      "loss": 3.3407,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5261892164879034,
+      "grad_norm": 3.305530508895833,
+      "learning_rate": 9.64173195899436e-06,
+      "loss": 3.2901,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5263816074262903,
+      "grad_norm": 3.4550345513041534,
+      "learning_rate": 9.635503947647894e-06,
+      "loss": 3.2756,
+      "step": 2736
+    },
+    {
+      "epoch": 0.526573998364677,
+      "grad_norm": 3.6035507784923455,
+      "learning_rate": 9.629276077867497e-06,
+      "loss": 3.1033,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5267663893030639,
+      "grad_norm": 2.433409320663025,
+      "learning_rate": 9.623048352071998e-06,
+      "loss": 2.7414,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5269587802414506,
+      "grad_norm": 3.2338004608603876,
+      "learning_rate": 9.616820772680174e-06,
+      "loss": 3.0111,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5271511711798375,
+      "grad_norm": 3.4222023658657705,
+      "learning_rate": 9.610593342110746e-06,
+      "loss": 2.9728,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5273435621182242,
+      "grad_norm": 3.3907012958541296,
+      "learning_rate": 9.604366062782381e-06,
+      "loss": 2.979,
+      "step": 2741
+    },
+    {
+      "epoch": 0.527535953056611,
+      "grad_norm": 3.218348975817469,
+      "learning_rate": 9.598138937113677e-06,
+      "loss": 2.8436,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5277283439949978,
+      "grad_norm": 3.4932868411894793,
+      "learning_rate": 9.59191196752318e-06,
+      "loss": 3.2519,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5279207349333846,
+      "grad_norm": 3.411134351384614,
+      "learning_rate": 9.58568515642937e-06,
+      "loss": 2.9564,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5281131258717714,
+      "grad_norm": 3.494054632889051,
+      "learning_rate": 9.579458506250668e-06,
+      "loss": 3.0162,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5283055168101582,
+      "grad_norm": 2.2153289485687693,
+      "learning_rate": 9.573232019405441e-06,
+      "loss": 2.5548,
+      "step": 2746
+    },
+    {
+      "epoch": 0.528497907748545,
+      "grad_norm": 3.780107057936906,
+      "learning_rate": 9.567005698311982e-06,
+      "loss": 2.974,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5286902986869318,
+      "grad_norm": 2.3443450763521314,
+      "learning_rate": 9.560779545388517e-06,
+      "loss": 2.6123,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5288826896253187,
+      "grad_norm": 7.628901729939195,
+      "learning_rate": 9.554553563053217e-06,
+      "loss": 3.2482,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5290750805637054,
+      "grad_norm": 3.7076221400093696,
+      "learning_rate": 9.548327753724181e-06,
+      "loss": 3.1232,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5292674715020923,
+      "grad_norm": 3.6297150334373454,
+      "learning_rate": 9.542102119819436e-06,
+      "loss": 3.2941,
+      "step": 2751
+    },
+    {
+      "epoch": 0.529459862440479,
+      "grad_norm": 3.5226382392037934,
+      "learning_rate": 9.535876663756955e-06,
+      "loss": 3.1811,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5296522533788659,
+      "grad_norm": 3.423595327441629,
+      "learning_rate": 9.529651387954628e-06,
+      "loss": 2.9438,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5298446443172526,
+      "grad_norm": 3.4422046294086264,
+      "learning_rate": 9.523426294830284e-06,
+      "loss": 3.1568,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5300370352556395,
+      "grad_norm": 3.861563379508026,
+      "learning_rate": 9.517201386801675e-06,
+      "loss": 3.176,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5302294261940262,
+      "grad_norm": 3.316483637788951,
+      "learning_rate": 9.510976666286484e-06,
+      "loss": 2.997,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5304218171324131,
+      "grad_norm": 3.3530269536919115,
+      "learning_rate": 9.504752135702318e-06,
+      "loss": 3.0981,
+      "step": 2757
+    },
+    {
+      "epoch": 0.5306142080707998,
+      "grad_norm": 3.325201671642127,
+      "learning_rate": 9.498527797466718e-06,
+      "loss": 3.0071,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5308065990091867,
+      "grad_norm": 3.3649455476967414,
+      "learning_rate": 9.492303653997146e-06,
+      "loss": 2.8951,
+      "step": 2759
+    },
+    {
+      "epoch": 0.5309989899475734,
+      "grad_norm": 3.3628988248773757,
+      "learning_rate": 9.48607970771098e-06,
+      "loss": 3.455,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5311913808859603,
+      "grad_norm": 3.6748155900364923,
+      "learning_rate": 9.479855961025538e-06,
+      "loss": 3.148,
+      "step": 2761
+    },
+    {
+      "epoch": 0.531383771824347,
+      "grad_norm": 3.5153971246929148,
+      "learning_rate": 9.473632416358045e-06,
+      "loss": 3.3449,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5315761627627339,
+      "grad_norm": 3.3564992513492786,
+      "learning_rate": 9.467409076125653e-06,
+      "loss": 3.0251,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5317685537011206,
+      "grad_norm": 3.445385375375759,
+      "learning_rate": 9.461185942745443e-06,
+      "loss": 3.0498,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5319609446395075,
+      "grad_norm": 3.7678845637533707,
+      "learning_rate": 9.454963018634402e-06,
+      "loss": 3.1997,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5321533355778942,
+      "grad_norm": 3.8761629017328594,
+      "learning_rate": 9.448740306209447e-06,
+      "loss": 3.2466,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5323457265162811,
+      "grad_norm": 3.3992637093288893,
+      "learning_rate": 9.442517807887402e-06,
+      "loss": 3.1311,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5325381174546678,
+      "grad_norm": 3.823347608027158,
+      "learning_rate": 9.436295526085016e-06,
+      "loss": 3.0168,
+      "step": 2768
+    },
+    {
+      "epoch": 0.5327305083930547,
+      "grad_norm": 2.3805767768679122,
+      "learning_rate": 9.430073463218952e-06,
+      "loss": 2.6208,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5329228993314414,
+      "grad_norm": 3.597178173606514,
+      "learning_rate": 9.423851621705789e-06,
+      "loss": 3.1017,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5331152902698283,
+      "grad_norm": 3.5488776565373557,
+      "learning_rate": 9.41763000396202e-06,
+      "loss": 2.9626,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5333076812082151,
+      "grad_norm": 3.530730493850447,
+      "learning_rate": 9.411408612404043e-06,
+      "loss": 2.8927,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5335000721466019,
+      "grad_norm": 3.289133591435827,
+      "learning_rate": 9.40518744944818e-06,
+      "loss": 2.6994,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5336924630849887,
+      "grad_norm": 3.458817545461412,
+      "learning_rate": 9.398966517510654e-06,
+      "loss": 3.0296,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5338848540233755,
+      "grad_norm": 3.425110926142774,
+      "learning_rate": 9.39274581900761e-06,
+      "loss": 3.0634,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5340772449617623,
+      "grad_norm": 3.3899384392761753,
+      "learning_rate": 9.386525356355095e-06,
+      "loss": 3.1716,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5342696359001491,
+      "grad_norm": 2.582558914597202,
+      "learning_rate": 9.380305131969059e-06,
+      "loss": 2.8512,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5344620268385359,
+      "grad_norm": 2.3984921557250822,
+      "learning_rate": 9.374085148265372e-06,
+      "loss": 2.6818,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5346544177769227,
+      "grad_norm": 3.403827205604075,
+      "learning_rate": 9.3678654076598e-06,
+      "loss": 2.9551,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5348468087153095,
+      "grad_norm": 3.3714764064344243,
+      "learning_rate": 9.361645912568015e-06,
+      "loss": 3.0611,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5350391996536963,
+      "grad_norm": 3.4349451487104252,
+      "learning_rate": 9.355426665405607e-06,
+      "loss": 3.1356,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5352315905920831,
+      "grad_norm": 3.4265861042821273,
+      "learning_rate": 9.349207668588053e-06,
+      "loss": 3.1253,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5354239815304699,
+      "grad_norm": 3.2272422925231474,
+      "learning_rate": 9.342988924530742e-06,
+      "loss": 2.9466,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5356163724688567,
+      "grad_norm": 3.4660788398726488,
+      "learning_rate": 9.336770435648963e-06,
+      "loss": 3.0811,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5358087634072435,
+      "grad_norm": 3.545352262594877,
+      "learning_rate": 9.330552204357904e-06,
+      "loss": 3.1244,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5360011543456303,
+      "grad_norm": 3.503127390476538,
+      "learning_rate": 9.32433423307265e-06,
+      "loss": 3.103,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5361935452840171,
+      "grad_norm": 2.333765248676852,
+      "learning_rate": 9.318116524208198e-06,
+      "loss": 2.6636,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5363859362224039,
+      "grad_norm": 3.201131262750119,
+      "learning_rate": 9.311899080179433e-06,
+      "loss": 2.8028,
+      "step": 2788
+    },
+    {
+      "epoch": 0.5365783271607907,
+      "grad_norm": 3.405706150803864,
+      "learning_rate": 9.305681903401133e-06,
+      "loss": 3.0834,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5367707180991775,
+      "grad_norm": 3.526428755833103,
+      "learning_rate": 9.299464996287984e-06,
+      "loss": 3.2256,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5369631090375643,
+      "grad_norm": 3.418079890450734,
+      "learning_rate": 9.293248361254557e-06,
+      "loss": 2.8575,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5371554999759511,
+      "grad_norm": 3.3051606855545055,
+      "learning_rate": 9.287032000715318e-06,
+      "loss": 3.1109,
+      "step": 2792
+    },
+    {
+      "epoch": 0.537347890914338,
+      "grad_norm": 3.334078743957499,
+      "learning_rate": 9.28081591708464e-06,
+      "loss": 3.1105,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5375402818527247,
+      "grad_norm": 3.368038536051019,
+      "learning_rate": 9.27460011277677e-06,
+      "loss": 2.8369,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5377326727911116,
+      "grad_norm": 3.513301382133477,
+      "learning_rate": 9.268384590205858e-06,
+      "loss": 2.9155,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5379250637294983,
+      "grad_norm": 3.8101126370372,
+      "learning_rate": 9.262169351785944e-06,
+      "loss": 3.1673,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5381174546678852,
+      "grad_norm": 3.2075562557242727,
+      "learning_rate": 9.255954399930948e-06,
+      "loss": 2.9232,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5383098456062719,
+      "grad_norm": 3.4503619780803993,
+      "learning_rate": 9.249739737054686e-06,
+      "loss": 3.0034,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5385022365446588,
+      "grad_norm": 3.104942761109755,
+      "learning_rate": 9.24352536557087e-06,
+      "loss": 2.8764,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5386946274830455,
+      "grad_norm": 3.543184080906699,
+      "learning_rate": 9.237311287893086e-06,
+      "loss": 3.0701,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5388870184214324,
+      "grad_norm": 4.0052125578186155,
+      "learning_rate": 9.231097506434808e-06,
+      "loss": 3.1409,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5390794093598191,
+      "grad_norm": 3.3857102379111788,
+      "learning_rate": 9.224884023609398e-06,
+      "loss": 3.1515,
+      "step": 2802
+    },
+    {
+      "epoch": 0.539271800298206,
+      "grad_norm": 3.5590806493851415,
+      "learning_rate": 9.218670841830098e-06,
+      "loss": 3.0063,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5394641912365927,
+      "grad_norm": 3.717502087928466,
+      "learning_rate": 9.212457963510045e-06,
+      "loss": 3.2599,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5396565821749796,
+      "grad_norm": 3.1817007100541304,
+      "learning_rate": 9.206245391062243e-06,
+      "loss": 2.98,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5398489731133663,
+      "grad_norm": 3.245787240240229,
+      "learning_rate": 9.200033126899585e-06,
+      "loss": 2.9581,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5400413640517532,
+      "grad_norm": 3.638814104831748,
+      "learning_rate": 9.193821173434843e-06,
+      "loss": 2.951,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5402337549901399,
+      "grad_norm": 3.4062503032372686,
+      "learning_rate": 9.187609533080668e-06,
+      "loss": 3.029,
+      "step": 2808
+    },
+    {
+      "epoch": 0.5404261459285268,
+      "grad_norm": 3.2289459258547906,
+      "learning_rate": 9.181398208249583e-06,
+      "loss": 3.0293,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5406185368669135,
+      "grad_norm": 2.321885461846329,
+      "learning_rate": 9.175187201354005e-06,
+      "loss": 2.6387,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5408109278053004,
+      "grad_norm": 3.374376411477994,
+      "learning_rate": 9.168976514806216e-06,
+      "loss": 3.0567,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5410033187436871,
+      "grad_norm": 2.4230967196509177,
+      "learning_rate": 9.162766151018372e-06,
+      "loss": 2.7444,
+      "step": 2812
+    },
+    {
+      "epoch": 0.541195709682074,
+      "grad_norm": 3.6815350067591144,
+      "learning_rate": 9.156556112402508e-06,
+      "loss": 3.299,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5413881006204607,
+      "grad_norm": 3.372463404472258,
+      "learning_rate": 9.150346401370528e-06,
+      "loss": 3.019,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5415804915588476,
+      "grad_norm": 3.4194133627100505,
+      "learning_rate": 9.144137020334214e-06,
+      "loss": 3.1156,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5417728824972344,
+      "grad_norm": 3.9306849620218625,
+      "learning_rate": 9.137927971705223e-06,
+      "loss": 3.1014,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5419652734356212,
+      "grad_norm": 3.398935497471376,
+      "learning_rate": 9.131719257895074e-06,
+      "loss": 3.2364,
+      "step": 2817
+    },
+    {
+      "epoch": 0.542157664374008,
+      "grad_norm": 3.359341222116963,
+      "learning_rate": 9.125510881315159e-06,
+      "loss": 3.1233,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5423500553123948,
+      "grad_norm": 3.4292392618764618,
+      "learning_rate": 9.119302844376741e-06,
+      "loss": 2.9753,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5425424462507816,
+      "grad_norm": 3.3755583679009926,
+      "learning_rate": 9.113095149490951e-06,
+      "loss": 2.8927,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5427348371891684,
+      "grad_norm": 3.2953527041681068,
+      "learning_rate": 9.106887799068782e-06,
+      "loss": 3.2179,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5429272281275552,
+      "grad_norm": 3.3531230437161486,
+      "learning_rate": 9.100680795521104e-06,
+      "loss": 3.1452,
+      "step": 2822
+    },
+    {
+      "epoch": 0.543119619065942,
+      "grad_norm": 3.636724756298331,
+      "learning_rate": 9.09447414125864e-06,
+      "loss": 3.1816,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5433120100043288,
+      "grad_norm": 3.233568938023358,
+      "learning_rate": 9.088267838691987e-06,
+      "loss": 3.1551,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5435044009427156,
+      "grad_norm": 3.443925574712492,
+      "learning_rate": 9.0820618902316e-06,
+      "loss": 3.0378,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5436967918811024,
+      "grad_norm": 3.610290365705572,
+      "learning_rate": 9.075856298287796e-06,
+      "loss": 2.9466,
+      "step": 2826
+    },
+    {
+      "epoch": 0.5438891828194892,
+      "grad_norm": 3.5246916960170522,
+      "learning_rate": 9.069651065270753e-06,
+      "loss": 3.36,
+      "step": 2827
+    },
+    {
+      "epoch": 0.544081573757876,
+      "grad_norm": 3.5203646455204796,
+      "learning_rate": 9.06344619359052e-06,
+      "loss": 2.6875,
+      "step": 2828
+    },
+    {
+      "epoch": 0.5442739646962628,
+      "grad_norm": 3.3469843320801176,
+      "learning_rate": 9.057241685656995e-06,
+      "loss": 3.183,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5444663556346496,
+      "grad_norm": 3.4933436795440755,
+      "learning_rate": 9.051037543879933e-06,
+      "loss": 3.2399,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5446587465730364,
+      "grad_norm": 3.817756469984606,
+      "learning_rate": 9.044833770668957e-06,
+      "loss": 3.1688,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5448511375114232,
+      "grad_norm": 3.4978453277628616,
+      "learning_rate": 9.038630368433537e-06,
+      "loss": 3.0597,
+      "step": 2832
+    },
+    {
+      "epoch": 0.54504352844981,
+      "grad_norm": 3.456690003493995,
+      "learning_rate": 9.032427339583e-06,
+      "loss": 2.9245,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5452359193881968,
+      "grad_norm": 2.4359432435692194,
+      "learning_rate": 9.026224686526539e-06,
+      "loss": 2.6185,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5454283103265836,
+      "grad_norm": 3.4286411171122535,
+      "learning_rate": 9.020022411673186e-06,
+      "loss": 3.1192,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5456207012649704,
+      "grad_norm": 2.1734044546883804,
+      "learning_rate": 9.013820517431841e-06,
+      "loss": 2.5723,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5458130922033573,
+      "grad_norm": 3.3824334077406277,
+      "learning_rate": 9.00761900621124e-06,
+      "loss": 3.0612,
+      "step": 2837
+    },
+    {
+      "epoch": 0.546005483141744,
+      "grad_norm": 2.412131563171212,
+      "learning_rate": 9.00141788041998e-06,
+      "loss": 2.605,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5461978740801309,
+      "grad_norm": 3.394465399903514,
+      "learning_rate": 8.99521714246651e-06,
+      "loss": 2.9238,
+      "step": 2839
+    },
+    {
+      "epoch": 0.5463902650185176,
+      "grad_norm": 3.4274359490756536,
+      "learning_rate": 8.989016794759127e-06,
+      "loss": 2.9438,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5465826559569045,
+      "grad_norm": 3.2495396751783154,
+      "learning_rate": 8.98281683970597e-06,
+      "loss": 3.3459,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5467750468952912,
+      "grad_norm": 3.287088285849496,
+      "learning_rate": 8.97661727971503e-06,
+      "loss": 2.8948,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5469674378336781,
+      "grad_norm": 3.5118259214681964,
+      "learning_rate": 8.970418117194146e-06,
+      "loss": 3.2677,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5471598287720648,
+      "grad_norm": 3.369540408135911,
+      "learning_rate": 8.964219354550999e-06,
+      "loss": 2.9579,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5473522197104517,
+      "grad_norm": 3.522405400187918,
+      "learning_rate": 8.958020994193124e-06,
+      "loss": 3.2657,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5475446106488384,
+      "grad_norm": 3.4589662928563176,
+      "learning_rate": 8.951823038527887e-06,
+      "loss": 3.2518,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5477370015872253,
+      "grad_norm": 3.4554304528724034,
+      "learning_rate": 8.945625489962503e-06,
+      "loss": 3.2881,
+      "step": 2847
+    },
+    {
+      "epoch": 0.547929392525612,
+      "grad_norm": 3.3881874436619572,
+      "learning_rate": 8.93942835090403e-06,
+      "loss": 2.9587,
+      "step": 2848
+    },
+    {
+      "epoch": 0.5481217834639989,
+      "grad_norm": 3.806267341113042,
+      "learning_rate": 8.933231623759365e-06,
+      "loss": 2.9005,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5483141744023856,
+      "grad_norm": 3.2400626483160653,
+      "learning_rate": 8.927035310935241e-06,
+      "loss": 2.9342,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5485065653407725,
+      "grad_norm": 3.7712829559564116,
+      "learning_rate": 8.920839414838243e-06,
+      "loss": 3.1059,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5486989562791592,
+      "grad_norm": 3.2383461360403105,
+      "learning_rate": 8.914643937874778e-06,
+      "loss": 3.0215,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5488913472175461,
+      "grad_norm": 3.180619921044246,
+      "learning_rate": 8.908448882451104e-06,
+      "loss": 3.0143,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5490837381559328,
+      "grad_norm": 3.7065205321745647,
+      "learning_rate": 8.902254250973306e-06,
+      "loss": 3.1338,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5492761290943197,
+      "grad_norm": 2.795509583368623,
+      "learning_rate": 8.896060045847305e-06,
+      "loss": 2.6473,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5494685200327064,
+      "grad_norm": 3.46216472779845,
+      "learning_rate": 8.88986626947886e-06,
+      "loss": 2.9409,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5496609109710933,
+      "grad_norm": 3.378013490554524,
+      "learning_rate": 8.883672924273566e-06,
+      "loss": 2.9818,
+      "step": 2857
+    },
+    {
+      "epoch": 0.5498533019094801,
+      "grad_norm": 3.525164306782934,
+      "learning_rate": 8.877480012636847e-06,
+      "loss": 3.0118,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5500456928478669,
+      "grad_norm": 3.3351677036765457,
+      "learning_rate": 8.871287536973954e-06,
+      "loss": 2.9026,
+      "step": 2859
+    },
+    {
+      "epoch": 0.5502380837862537,
+      "grad_norm": 3.3505982376700496,
+      "learning_rate": 8.865095499689978e-06,
+      "loss": 2.7997,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5504304747246405,
+      "grad_norm": 3.2853019630091747,
+      "learning_rate": 8.85890390318983e-06,
+      "loss": 2.9008,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5506228656630273,
+      "grad_norm": 3.3907805654342233,
+      "learning_rate": 8.852712749878255e-06,
+      "loss": 3.3629,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5508152566014141,
+      "grad_norm": 3.4455492209486582,
+      "learning_rate": 8.846522042159833e-06,
+      "loss": 2.9368,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5510076475398009,
+      "grad_norm": 3.2026881881415155,
+      "learning_rate": 8.840331782438954e-06,
+      "loss": 2.7091,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5512000384781877,
+      "grad_norm": 3.2108878453141174,
+      "learning_rate": 8.83414197311985e-06,
+      "loss": 2.9465,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5513924294165745,
+      "grad_norm": 3.3387980512469966,
+      "learning_rate": 8.82795261660657e-06,
+      "loss": 2.9704,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5515848203549613,
+      "grad_norm": 3.5190928076257695,
+      "learning_rate": 8.821763715302986e-06,
+      "loss": 3.0223,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5517772112933481,
+      "grad_norm": 3.7389603393839934,
+      "learning_rate": 8.815575271612798e-06,
+      "loss": 3.1279,
+      "step": 2868
+    },
+    {
+      "epoch": 0.5519696022317349,
+      "grad_norm": 3.219233598113543,
+      "learning_rate": 8.809387287939528e-06,
+      "loss": 2.9454,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5521619931701217,
+      "grad_norm": 3.335783466964186,
+      "learning_rate": 8.803199766686517e-06,
+      "loss": 3.2237,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5523543841085085,
+      "grad_norm": 3.37151251698765,
+      "learning_rate": 8.797012710256923e-06,
+      "loss": 3.0469,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5525467750468953,
+      "grad_norm": 2.7190179756868638,
+      "learning_rate": 8.790826121053732e-06,
+      "loss": 2.6898,
+      "step": 2872
+    },
+    {
+      "epoch": 0.552739165985282,
+      "grad_norm": 3.1965723402749218,
+      "learning_rate": 8.784640001479741e-06,
+      "loss": 2.859,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5529315569236689,
+      "grad_norm": 3.57217832343962,
+      "learning_rate": 8.77845435393757e-06,
+      "loss": 3.0699,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5531239478620557,
+      "grad_norm": 3.222771507323968,
+      "learning_rate": 8.772269180829653e-06,
+      "loss": 2.9663,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5533163388004425,
+      "grad_norm": 3.422804591312974,
+      "learning_rate": 8.766084484558237e-06,
+      "loss": 3.0034,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5535087297388293,
+      "grad_norm": 3.30633461798492,
+      "learning_rate": 8.759900267525393e-06,
+      "loss": 3.0158,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5537011206772161,
+      "grad_norm": 3.4092142544532447,
+      "learning_rate": 8.753716532132992e-06,
+      "loss": 2.9705,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5538935116156029,
+      "grad_norm": 3.4708882247935287,
+      "learning_rate": 8.747533280782725e-06,
+      "loss": 3.0486,
+      "step": 2879
+    },
+    {
+      "epoch": 0.5540859025539897,
+      "grad_norm": 3.514430143643134,
+      "learning_rate": 8.741350515876103e-06,
+      "loss": 3.1395,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5542782934923766,
+      "grad_norm": 3.2469986324395705,
+      "learning_rate": 8.73516823981444e-06,
+      "loss": 2.9557,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5544706844307633,
+      "grad_norm": 3.435738220365889,
+      "learning_rate": 8.728986454998858e-06,
+      "loss": 3.0621,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5546630753691502,
+      "grad_norm": 3.281550450024847,
+      "learning_rate": 8.72280516383029e-06,
+      "loss": 3.0759,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5548554663075369,
+      "grad_norm": 3.2436589360606574,
+      "learning_rate": 8.716624368709477e-06,
+      "loss": 2.7888,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5550478572459238,
+      "grad_norm": 3.573131221647325,
+      "learning_rate": 8.71044407203697e-06,
+      "loss": 3.0718,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5552402481843105,
+      "grad_norm": 3.471004175880475,
+      "learning_rate": 8.70426427621313e-06,
+      "loss": 2.954,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5554326391226974,
+      "grad_norm": 3.7092758171874443,
+      "learning_rate": 8.698084983638111e-06,
+      "loss": 2.8961,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5556250300610841,
+      "grad_norm": 3.605951443464281,
+      "learning_rate": 8.691906196711884e-06,
+      "loss": 2.9463,
+      "step": 2888
+    },
+    {
+      "epoch": 0.555817420999471,
+      "grad_norm": 2.2550830861038738,
+      "learning_rate": 8.685727917834218e-06,
+      "loss": 2.5856,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5560098119378577,
+      "grad_norm": 3.428997949768597,
+      "learning_rate": 8.679550149404685e-06,
+      "loss": 3.0831,
+      "step": 2890
+    },
+    {
+      "epoch": 0.5562022028762446,
+      "grad_norm": 3.2838203834440693,
+      "learning_rate": 8.673372893822653e-06,
+      "loss": 3.0361,
+      "step": 2891
+    },
+    {
+      "epoch": 0.5563945938146313,
+      "grad_norm": 3.4658799645650973,
+      "learning_rate": 8.667196153487308e-06,
+      "loss": 3.1635,
+      "step": 2892
+    },
+    {
+      "epoch": 0.5565869847530182,
+      "grad_norm": 3.1657872188116096,
+      "learning_rate": 8.661019930797615e-06,
+      "loss": 2.9473,
+      "step": 2893
+    },
+    {
+      "epoch": 0.5567793756914049,
+      "grad_norm": 3.4105728282848204,
+      "learning_rate": 8.654844228152355e-06,
+      "loss": 3.0604,
+      "step": 2894
+    },
+    {
+      "epoch": 0.5569717666297918,
+      "grad_norm": 3.608751453263586,
+      "learning_rate": 8.648669047950097e-06,
+      "loss": 3.026,
+      "step": 2895
+    },
+    {
+      "epoch": 0.5571641575681785,
+      "grad_norm": 3.4237334199511307,
+      "learning_rate": 8.642494392589206e-06,
+      "loss": 3.054,
+      "step": 2896
+    },
+    {
+      "epoch": 0.5573565485065654,
+      "grad_norm": 3.5138800681937017,
+      "learning_rate": 8.63632026446785e-06,
+      "loss": 3.0863,
+      "step": 2897
+    },
+    {
+      "epoch": 0.5575489394449521,
+      "grad_norm": 3.7130924152724165,
+      "learning_rate": 8.630146665983993e-06,
+      "loss": 3.0258,
+      "step": 2898
+    },
+    {
+      "epoch": 0.557741330383339,
+      "grad_norm": 3.3868920871211947,
+      "learning_rate": 8.623973599535385e-06,
+      "loss": 2.9442,
+      "step": 2899
+    },
+    {
+      "epoch": 0.5579337213217257,
+      "grad_norm": 3.593043781405917,
+      "learning_rate": 8.617801067519575e-06,
+      "loss": 2.8862,
+      "step": 2900
+    },
+    {
+      "epoch": 0.5581261122601126,
+      "grad_norm": 3.6313810504542374,
+      "learning_rate": 8.611629072333905e-06,
+      "loss": 3.0054,
+      "step": 2901
+    },
+    {
+      "epoch": 0.5583185031984994,
+      "grad_norm": 3.6112710518540765,
+      "learning_rate": 8.605457616375503e-06,
+      "loss": 3.4025,
+      "step": 2902
+    },
+    {
+      "epoch": 0.5585108941368861,
+      "grad_norm": 3.450232124994151,
+      "learning_rate": 8.599286702041292e-06,
+      "loss": 3.2714,
+      "step": 2903
+    },
+    {
+      "epoch": 0.558703285075273,
+      "grad_norm": 3.4682271682299963,
+      "learning_rate": 8.593116331727987e-06,
+      "loss": 3.1321,
+      "step": 2904
+    },
+    {
+      "epoch": 0.5588956760136597,
+      "grad_norm": 3.4480821256262164,
+      "learning_rate": 8.586946507832088e-06,
+      "loss": 3.2508,
+      "step": 2905
+    },
+    {
+      "epoch": 0.5590880669520466,
+      "grad_norm": 3.2650104419141903,
+      "learning_rate": 8.580777232749883e-06,
+      "loss": 3.0223,
+      "step": 2906
+    },
+    {
+      "epoch": 0.5592804578904333,
+      "grad_norm": 3.284409904985459,
+      "learning_rate": 8.574608508877448e-06,
+      "loss": 3.0131,
+      "step": 2907
+    },
+    {
+      "epoch": 0.5594728488288202,
+      "grad_norm": 3.282694333199841,
+      "learning_rate": 8.568440338610638e-06,
+      "loss": 3.2158,
+      "step": 2908
+    },
+    {
+      "epoch": 0.559665239767207,
+      "grad_norm": 3.72312420125456,
+      "learning_rate": 8.562272724345108e-06,
+      "loss": 2.9547,
+      "step": 2909
+    },
+    {
+      "epoch": 0.5598576307055938,
+      "grad_norm": 3.5184521585464514,
+      "learning_rate": 8.556105668476287e-06,
+      "loss": 2.8993,
+      "step": 2910
+    },
+    {
+      "epoch": 0.5600500216439805,
+      "grad_norm": 3.315271489569505,
+      "learning_rate": 8.549939173399385e-06,
+      "loss": 2.8808,
+      "step": 2911
+    },
+    {
+      "epoch": 0.5602424125823674,
+      "grad_norm": 3.2905275849223843,
+      "learning_rate": 8.5437732415094e-06,
+      "loss": 3.1053,
+      "step": 2912
+    },
+    {
+      "epoch": 0.5604348035207541,
+      "grad_norm": 3.2853303380135825,
+      "learning_rate": 8.537607875201106e-06,
+      "loss": 3.0566,
+      "step": 2913
+    },
+    {
+      "epoch": 0.560627194459141,
+      "grad_norm": 3.368118940144854,
+      "learning_rate": 8.531443076869058e-06,
+      "loss": 2.9923,
+      "step": 2914
+    },
+    {
+      "epoch": 0.5608195853975277,
+      "grad_norm": 3.4832990800229746,
+      "learning_rate": 8.525278848907603e-06,
+      "loss": 3.2017,
+      "step": 2915
+    },
+    {
+      "epoch": 0.5610119763359146,
+      "grad_norm": 3.774280597551869,
+      "learning_rate": 8.51911519371085e-06,
+      "loss": 3.4012,
+      "step": 2916
+    },
+    {
+      "epoch": 0.5612043672743013,
+      "grad_norm": 3.528024211911425,
+      "learning_rate": 8.512952113672689e-06,
+      "loss": 3.0059,
+      "step": 2917
+    },
+    {
+      "epoch": 0.5613967582126882,
+      "grad_norm": 3.489036400779609,
+      "learning_rate": 8.506789611186794e-06,
+      "loss": 3.2132,
+      "step": 2918
+    },
+    {
+      "epoch": 0.5615891491510749,
+      "grad_norm": 3.366079245930363,
+      "learning_rate": 8.500627688646607e-06,
+      "loss": 3.2744,
+      "step": 2919
+    },
+    {
+      "epoch": 0.5617815400894618,
+      "grad_norm": 3.5902287274165645,
+      "learning_rate": 8.494466348445345e-06,
+      "loss": 2.9654,
+      "step": 2920
+    },
+    {
+      "epoch": 0.5619739310278485,
+      "grad_norm": 3.379724298915237,
+      "learning_rate": 8.48830559297601e-06,
+      "loss": 3.2528,
+      "step": 2921
+    },
+    {
+      "epoch": 0.5621663219662354,
+      "grad_norm": 3.5137924586837768,
+      "learning_rate": 8.48214542463136e-06,
+      "loss": 3.1431,
+      "step": 2922
+    },
+    {
+      "epoch": 0.5623587129046221,
+      "grad_norm": 3.499585661583127,
+      "learning_rate": 8.475985845803938e-06,
+      "loss": 3.131,
+      "step": 2923
+    },
+    {
+      "epoch": 0.562551103843009,
+      "grad_norm": 3.533082082305384,
+      "learning_rate": 8.469826858886054e-06,
+      "loss": 3.0347,
+      "step": 2924
+    },
+    {
+      "epoch": 0.5627434947813958,
+      "grad_norm": 3.533706379743962,
+      "learning_rate": 8.463668466269785e-06,
+      "loss": 3.3207,
+      "step": 2925
+    },
+    {
+      "epoch": 0.5629358857197826,
+      "grad_norm": 3.2593512484904688,
+      "learning_rate": 8.457510670346976e-06,
+      "loss": 2.8732,
+      "step": 2926
+    },
+    {
+      "epoch": 0.5631282766581694,
+      "grad_norm": 3.6129641324124044,
+      "learning_rate": 8.451353473509254e-06,
+      "loss": 3.1069,
+      "step": 2927
+    },
+    {
+      "epoch": 0.5633206675965562,
+      "grad_norm": 3.4265139738235,
+      "learning_rate": 8.445196878147997e-06,
+      "loss": 3.1692,
+      "step": 2928
+    },
+    {
+      "epoch": 0.563513058534943,
+      "grad_norm": 3.571167852286114,
+      "learning_rate": 8.439040886654354e-06,
+      "loss": 2.919,
+      "step": 2929
+    },
+    {
+      "epoch": 0.5637054494733298,
+      "grad_norm": 3.7805958967176436,
+      "learning_rate": 8.432885501419248e-06,
+      "loss": 3.1263,
+      "step": 2930
+    },
+    {
+      "epoch": 0.5638978404117166,
+      "grad_norm": 3.2835405407466283,
+      "learning_rate": 8.426730724833354e-06,
+      "loss": 3.0693,
+      "step": 2931
+    },
+    {
+      "epoch": 0.5640902313501034,
+      "grad_norm": 3.5069777416253967,
+      "learning_rate": 8.420576559287112e-06,
+      "loss": 3.2257,
+      "step": 2932
+    },
+    {
+      "epoch": 0.5642826222884902,
+      "grad_norm": 3.357672754512838,
+      "learning_rate": 8.414423007170742e-06,
+      "loss": 3.1644,
+      "step": 2933
+    },
+    {
+      "epoch": 0.564475013226877,
+      "grad_norm": 3.1743229254256295,
+      "learning_rate": 8.408270070874201e-06,
+      "loss": 2.8605,
+      "step": 2934
+    },
+    {
+      "epoch": 0.5646674041652638,
+      "grad_norm": 3.571499318980105,
+      "learning_rate": 8.402117752787225e-06,
+      "loss": 2.9937,
+      "step": 2935
+    },
+    {
+      "epoch": 0.5648597951036506,
+      "grad_norm": 3.359085060579703,
+      "learning_rate": 8.395966055299302e-06,
+      "loss": 3.035,
+      "step": 2936
+    },
+    {
+      "epoch": 0.5650521860420374,
+      "grad_norm": 3.171470956508546,
+      "learning_rate": 8.389814980799679e-06,
+      "loss": 3.0577,
+      "step": 2937
+    },
+    {
+      "epoch": 0.5652445769804242,
+      "grad_norm": 3.5071385888492275,
+      "learning_rate": 8.38366453167736e-06,
+      "loss": 2.81,
+      "step": 2938
+    },
+    {
+      "epoch": 0.565436967918811,
+      "grad_norm": 3.4248099886593497,
+      "learning_rate": 8.377514710321117e-06,
+      "loss": 2.8068,
+      "step": 2939
+    },
+    {
+      "epoch": 0.5656293588571978,
+      "grad_norm": 3.6673842512616073,
+      "learning_rate": 8.371365519119463e-06,
+      "loss": 3.0,
+      "step": 2940
+    },
+    {
+      "epoch": 0.5658217497955846,
+      "grad_norm": 3.3817917433056293,
+      "learning_rate": 8.365216960460675e-06,
+      "loss": 3.014,
+      "step": 2941
+    },
+    {
+      "epoch": 0.5660141407339714,
+      "grad_norm": 3.5198732080077972,
+      "learning_rate": 8.359069036732781e-06,
+      "loss": 3.0686,
+      "step": 2942
+    },
+    {
+      "epoch": 0.5662065316723582,
+      "grad_norm": 3.3188344438236563,
+      "learning_rate": 8.352921750323562e-06,
+      "loss": 3.1,
+      "step": 2943
+    },
+    {
+      "epoch": 0.566398922610745,
+      "grad_norm": 3.7009262019549944,
+      "learning_rate": 8.346775103620559e-06,
+      "loss": 3.1501,
+      "step": 2944
+    },
+    {
+      "epoch": 0.5665913135491318,
+      "grad_norm": 3.3920254776815977,
+      "learning_rate": 8.340629099011057e-06,
+      "loss": 2.8532,
+      "step": 2945
+    },
+    {
+      "epoch": 0.5667837044875187,
+      "grad_norm": 3.8324378900225873,
+      "learning_rate": 8.33448373888209e-06,
+      "loss": 2.9175,
+      "step": 2946
+    },
+    {
+      "epoch": 0.5669760954259054,
+      "grad_norm": 3.6189802555309645,
+      "learning_rate": 8.328339025620449e-06,
+      "loss": 2.9208,
+      "step": 2947
+    },
+    {
+      "epoch": 0.5671684863642923,
+      "grad_norm": 3.236661969014474,
+      "learning_rate": 8.322194961612668e-06,
+      "loss": 3.0404,
+      "step": 2948
+    },
+    {
+      "epoch": 0.567360877302679,
+      "grad_norm": 3.3844621852197267,
+      "learning_rate": 8.316051549245026e-06,
+      "loss": 2.8865,
+      "step": 2949
+    },
+    {
+      "epoch": 0.5675532682410659,
+      "grad_norm": 3.034052835295303,
+      "learning_rate": 8.309908790903562e-06,
+      "loss": 2.8786,
+      "step": 2950
+    },
+    {
+      "epoch": 0.5677456591794526,
+      "grad_norm": 3.4390675028227102,
+      "learning_rate": 8.303766688974047e-06,
+      "loss": 3.0431,
+      "step": 2951
+    },
+    {
+      "epoch": 0.5679380501178395,
+      "grad_norm": 3.334790532333487,
+      "learning_rate": 8.297625245842006e-06,
+      "loss": 3.1969,
+      "step": 2952
+    },
+    {
+      "epoch": 0.5681304410562262,
+      "grad_norm": 3.281969183630088,
+      "learning_rate": 8.291484463892703e-06,
+      "loss": 3.1446,
+      "step": 2953
+    },
+    {
+      "epoch": 0.5683228319946131,
+      "grad_norm": 3.377469829073395,
+      "learning_rate": 8.285344345511147e-06,
+      "loss": 3.1934,
+      "step": 2954
+    },
+    {
+      "epoch": 0.5685152229329998,
+      "grad_norm": 3.352351013569528,
+      "learning_rate": 8.279204893082083e-06,
+      "loss": 3.0105,
+      "step": 2955
+    },
+    {
+      "epoch": 0.5687076138713867,
+      "grad_norm": 3.836892860617646,
+      "learning_rate": 8.273066108990017e-06,
+      "loss": 3.0693,
+      "step": 2956
+    },
+    {
+      "epoch": 0.5689000048097734,
+      "grad_norm": 3.4209389433201083,
+      "learning_rate": 8.266927995619175e-06,
+      "loss": 3.2102,
+      "step": 2957
+    },
+    {
+      "epoch": 0.5690923957481603,
+      "grad_norm": 3.4577948977812496,
+      "learning_rate": 8.260790555353526e-06,
+      "loss": 3.0298,
+      "step": 2958
+    },
+    {
+      "epoch": 0.569284786686547,
+      "grad_norm": 3.5063414950462013,
+      "learning_rate": 8.254653790576787e-06,
+      "loss": 3.2014,
+      "step": 2959
+    },
+    {
+      "epoch": 0.5694771776249339,
+      "grad_norm": 3.356685894106561,
+      "learning_rate": 8.248517703672405e-06,
+      "loss": 2.915,
+      "step": 2960
+    },
+    {
+      "epoch": 0.5696695685633206,
+      "grad_norm": 3.267464620821728,
+      "learning_rate": 8.242382297023558e-06,
+      "loss": 3.0016,
+      "step": 2961
+    },
+    {
+      "epoch": 0.5698619595017075,
+      "grad_norm": 3.6409775462758844,
+      "learning_rate": 8.23624757301318e-06,
+      "loss": 3.2165,
+      "step": 2962
+    },
+    {
+      "epoch": 0.5700543504400942,
+      "grad_norm": 3.23152499626095,
+      "learning_rate": 8.230113534023917e-06,
+      "loss": 2.9727,
+      "step": 2963
+    },
+    {
+      "epoch": 0.5702467413784811,
+      "grad_norm": 3.2308803934698798,
+      "learning_rate": 8.223980182438167e-06,
+      "loss": 3.0227,
+      "step": 2964
+    },
+    {
+      "epoch": 0.5704391323168678,
+      "grad_norm": 3.5913508901849243,
+      "learning_rate": 8.217847520638049e-06,
+      "loss": 3.2513,
+      "step": 2965
+    },
+    {
+      "epoch": 0.5706315232552547,
+      "grad_norm": 3.292211241142333,
+      "learning_rate": 8.211715551005414e-06,
+      "loss": 2.926,
+      "step": 2966
+    },
+    {
+      "epoch": 0.5708239141936414,
+      "grad_norm": 3.6866407464864643,
+      "learning_rate": 8.205584275921854e-06,
+      "loss": 3.042,
+      "step": 2967
+    },
+    {
+      "epoch": 0.5710163051320283,
+      "grad_norm": 3.5006873347789753,
+      "learning_rate": 8.199453697768686e-06,
+      "loss": 3.2883,
+      "step": 2968
+    },
+    {
+      "epoch": 0.5712086960704151,
+      "grad_norm": 2.493337615167186,
+      "learning_rate": 8.193323818926955e-06,
+      "loss": 2.6831,
+      "step": 2969
+    },
+    {
+      "epoch": 0.5714010870088019,
+      "grad_norm": 3.534753950993612,
+      "learning_rate": 8.187194641777432e-06,
+      "loss": 2.9424,
+      "step": 2970
+    },
+    {
+      "epoch": 0.5715934779471887,
+      "grad_norm": 3.3360582870291857,
+      "learning_rate": 8.181066168700622e-06,
+      "loss": 3.1199,
+      "step": 2971
+    },
+    {
+      "epoch": 0.5717858688855755,
+      "grad_norm": 3.1487239506407283,
+      "learning_rate": 8.174938402076754e-06,
+      "loss": 2.9166,
+      "step": 2972
+    },
+    {
+      "epoch": 0.5719782598239623,
+      "grad_norm": 3.3054529390584166,
+      "learning_rate": 8.168811344285776e-06,
+      "loss": 2.9802,
+      "step": 2973
+    },
+    {
+      "epoch": 0.5721706507623491,
+      "grad_norm": 3.4881789199158897,
+      "learning_rate": 8.162684997707374e-06,
+      "loss": 3.0119,
+      "step": 2974
+    },
+    {
+      "epoch": 0.5723630417007359,
+      "grad_norm": 3.7962507674609838,
+      "learning_rate": 8.156559364720947e-06,
+      "loss": 3.0753,
+      "step": 2975
+    },
+    {
+      "epoch": 0.5725554326391227,
+      "grad_norm": 3.53318457718163,
+      "learning_rate": 8.150434447705623e-06,
+      "loss": 2.9484,
+      "step": 2976
+    },
+    {
+      "epoch": 0.5727478235775095,
+      "grad_norm": 3.362794753954557,
+      "learning_rate": 8.144310249040246e-06,
+      "loss": 3.0618,
+      "step": 2977
+    },
+    {
+      "epoch": 0.5729402145158963,
+      "grad_norm": 3.4409597699859256,
+      "learning_rate": 8.138186771103382e-06,
+      "loss": 2.9986,
+      "step": 2978
+    },
+    {
+      "epoch": 0.5731326054542831,
+      "grad_norm": 3.572797650431118,
+      "learning_rate": 8.132064016273325e-06,
+      "loss": 3.068,
+      "step": 2979
+    },
+    {
+      "epoch": 0.5733249963926699,
+      "grad_norm": 3.2613810521341406,
+      "learning_rate": 8.12594198692808e-06,
+      "loss": 2.9289,
+      "step": 2980
+    },
+    {
+      "epoch": 0.5735173873310567,
+      "grad_norm": 3.2328309763254888,
+      "learning_rate": 8.119820685445373e-06,
+      "loss": 2.9935,
+      "step": 2981
+    },
+    {
+      "epoch": 0.5737097782694435,
+      "grad_norm": 3.4489256052180863,
+      "learning_rate": 8.113700114202647e-06,
+      "loss": 2.8898,
+      "step": 2982
+    },
+    {
+      "epoch": 0.5739021692078303,
+      "grad_norm": 3.4832246106334472,
+      "learning_rate": 8.107580275577059e-06,
+      "loss": 2.7139,
+      "step": 2983
+    },
+    {
+      "epoch": 0.5740945601462171,
+      "grad_norm": 3.1713938037988063,
+      "learning_rate": 8.101461171945483e-06,
+      "loss": 3.0166,
+      "step": 2984
+    },
+    {
+      "epoch": 0.5742869510846039,
+      "grad_norm": 3.498939719927801,
+      "learning_rate": 8.095342805684516e-06,
+      "loss": 3.1148,
+      "step": 2985
+    },
+    {
+      "epoch": 0.5744793420229907,
+      "grad_norm": 2.4539559462622975,
+      "learning_rate": 8.089225179170454e-06,
+      "loss": 2.6118,
+      "step": 2986
+    },
+    {
+      "epoch": 0.5746717329613775,
+      "grad_norm": 3.361144017731216,
+      "learning_rate": 8.083108294779313e-06,
+      "loss": 3.0305,
+      "step": 2987
+    },
+    {
+      "epoch": 0.5748641238997643,
+      "grad_norm": 3.415729351186661,
+      "learning_rate": 8.076992154886826e-06,
+      "loss": 3.0725,
+      "step": 2988
+    },
+    {
+      "epoch": 0.5750565148381511,
+      "grad_norm": 3.2086155848015494,
+      "learning_rate": 8.070876761868426e-06,
+      "loss": 2.9596,
+      "step": 2989
+    },
+    {
+      "epoch": 0.575248905776538,
+      "grad_norm": 2.4580469966618343,
+      "learning_rate": 8.064762118099258e-06,
+      "loss": 2.7535,
+      "step": 2990
+    },
+    {
+      "epoch": 0.5754412967149247,
+      "grad_norm": 2.682182354666669,
+      "learning_rate": 8.058648225954188e-06,
+      "loss": 2.67,
+      "step": 2991
+    },
+    {
+      "epoch": 0.5756336876533116,
+      "grad_norm": 3.230951236021042,
+      "learning_rate": 8.052535087807774e-06,
+      "loss": 2.7785,
+      "step": 2992
+    },
+    {
+      "epoch": 0.5758260785916983,
+      "grad_norm": 3.528568997773176,
+      "learning_rate": 8.046422706034294e-06,
+      "loss": 2.9096,
+      "step": 2993
+    },
+    {
+      "epoch": 0.5760184695300852,
+      "grad_norm": 3.4702968962647796,
+      "learning_rate": 8.040311083007725e-06,
+      "loss": 3.0215,
+      "step": 2994
+    },
+    {
+      "epoch": 0.5762108604684719,
+      "grad_norm": 3.564456650362154,
+      "learning_rate": 8.034200221101746e-06,
+      "loss": 3.2232,
+      "step": 2995
+    },
+    {
+      "epoch": 0.5764032514068588,
+      "grad_norm": 3.119021265399688,
+      "learning_rate": 8.028090122689747e-06,
+      "loss": 3.0406,
+      "step": 2996
+    },
+    {
+      "epoch": 0.5765956423452455,
+      "grad_norm": 3.4190560920196598,
+      "learning_rate": 8.021980790144828e-06,
+      "loss": 2.9084,
+      "step": 2997
+    },
+    {
+      "epoch": 0.5767880332836324,
+      "grad_norm": 3.361544982886083,
+      "learning_rate": 8.015872225839776e-06,
+      "loss": 3.0581,
+      "step": 2998
+    },
+    {
+      "epoch": 0.5769804242220191,
+      "grad_norm": 2.321924135335796,
+      "learning_rate": 8.009764432147086e-06,
+      "loss": 2.5864,
+      "step": 2999
+    },
+    {
+      "epoch": 0.577172815160406,
+      "grad_norm": 3.634083048260218,
+      "learning_rate": 8.003657411438961e-06,
+      "loss": 2.795,
+      "step": 3000
+    },
+    {
+      "epoch": 0.5773652060987927,
+      "grad_norm": 3.3736392316896806,
+      "learning_rate": 7.997551166087293e-06,
+      "loss": 3.0451,
+      "step": 3001
+    },
+    {
+      "epoch": 0.5775575970371796,
+      "grad_norm": 3.1000796905600025,
+      "learning_rate": 7.991445698463672e-06,
+      "loss": 2.7592,
+      "step": 3002
+    },
+    {
+      "epoch": 0.5777499879755663,
+      "grad_norm": 2.4820539490258797,
+      "learning_rate": 7.985341010939402e-06,
+      "loss": 2.6492,
+      "step": 3003
+    },
+    {
+      "epoch": 0.5779423789139532,
+      "grad_norm": 3.866423748614394,
+      "learning_rate": 7.979237105885467e-06,
+      "loss": 3.284,
+      "step": 3004
+    },
+    {
+      "epoch": 0.5781347698523399,
+      "grad_norm": 3.3964141924976246,
+      "learning_rate": 7.973133985672558e-06,
+      "loss": 2.9805,
+      "step": 3005
+    },
+    {
+      "epoch": 0.5783271607907268,
+      "grad_norm": 3.6217934829275955,
+      "learning_rate": 7.967031652671051e-06,
+      "loss": 3.0306,
+      "step": 3006
+    },
+    {
+      "epoch": 0.5785195517291135,
+      "grad_norm": 3.716737469431339,
+      "learning_rate": 7.960930109251023e-06,
+      "loss": 3.019,
+      "step": 3007
+    },
+    {
+      "epoch": 0.5787119426675004,
+      "grad_norm": 3.3784990820574117,
+      "learning_rate": 7.954829357782243e-06,
+      "loss": 3.1523,
+      "step": 3008
+    },
+    {
+      "epoch": 0.5789043336058871,
+      "grad_norm": 3.2358529771901092,
+      "learning_rate": 7.948729400634178e-06,
+      "loss": 2.9981,
+      "step": 3009
+    },
+    {
+      "epoch": 0.579096724544274,
+      "grad_norm": 3.3612609838331378,
+      "learning_rate": 7.942630240175977e-06,
+      "loss": 3.0723,
+      "step": 3010
+    },
+    {
+      "epoch": 0.5792891154826608,
+      "grad_norm": 3.5610555292739323,
+      "learning_rate": 7.936531878776484e-06,
+      "loss": 3.1077,
+      "step": 3011
+    },
+    {
+      "epoch": 0.5794815064210476,
+      "grad_norm": 3.18072211223373,
+      "learning_rate": 7.930434318804229e-06,
+      "loss": 3.1039,
+      "step": 3012
+    },
+    {
+      "epoch": 0.5796738973594344,
+      "grad_norm": 3.673652439396747,
+      "learning_rate": 7.924337562627435e-06,
+      "loss": 3.0375,
+      "step": 3013
+    },
+    {
+      "epoch": 0.5798662882978212,
+      "grad_norm": 3.21736897115576,
+      "learning_rate": 7.918241612614016e-06,
+      "loss": 3.0955,
+      "step": 3014
+    },
+    {
+      "epoch": 0.580058679236208,
+      "grad_norm": 3.272841523790372,
+      "learning_rate": 7.91214647113157e-06,
+      "loss": 2.9327,
+      "step": 3015
+    },
+    {
+      "epoch": 0.5802510701745948,
+      "grad_norm": 3.285028357253793,
+      "learning_rate": 7.906052140547373e-06,
+      "loss": 3.0564,
+      "step": 3016
+    },
+    {
+      "epoch": 0.5804434611129816,
+      "grad_norm": 3.3219420176248575,
+      "learning_rate": 7.899958623228398e-06,
+      "loss": 3.1639,
+      "step": 3017
+    },
+    {
+      "epoch": 0.5806358520513684,
+      "grad_norm": 3.53574944471683,
+      "learning_rate": 7.893865921541294e-06,
+      "loss": 3.1602,
+      "step": 3018
+    },
+    {
+      "epoch": 0.5808282429897552,
+      "grad_norm": 3.3983572459584424,
+      "learning_rate": 7.887774037852395e-06,
+      "loss": 3.3447,
+      "step": 3019
+    },
+    {
+      "epoch": 0.581020633928142,
+      "grad_norm": 3.5188430483502944,
+      "learning_rate": 7.881682974527723e-06,
+      "loss": 3.1068,
+      "step": 3020
+    },
+    {
+      "epoch": 0.5812130248665288,
+      "grad_norm": 3.4456964400229992,
+      "learning_rate": 7.875592733932972e-06,
+      "loss": 3.1081,
+      "step": 3021
+    },
+    {
+      "epoch": 0.5814054158049156,
+      "grad_norm": 3.7215520692296415,
+      "learning_rate": 7.869503318433529e-06,
+      "loss": 3.2989,
+      "step": 3022
+    },
+    {
+      "epoch": 0.5815978067433024,
+      "grad_norm": 3.336765029150241,
+      "learning_rate": 7.863414730394444e-06,
+      "loss": 2.8011,
+      "step": 3023
+    },
+    {
+      "epoch": 0.5817901976816892,
+      "grad_norm": 3.1010216584647265,
+      "learning_rate": 7.857326972180455e-06,
+      "loss": 2.8232,
+      "step": 3024
+    },
+    {
+      "epoch": 0.581982588620076,
+      "grad_norm": 3.149234327618063,
+      "learning_rate": 7.85124004615598e-06,
+      "loss": 2.8477,
+      "step": 3025
+    },
+    {
+      "epoch": 0.5821749795584628,
+      "grad_norm": 3.5025210209629782,
+      "learning_rate": 7.845153954685114e-06,
+      "loss": 3.0759,
+      "step": 3026
+    },
+    {
+      "epoch": 0.5823673704968496,
+      "grad_norm": 3.281400615717622,
+      "learning_rate": 7.839068700131623e-06,
+      "loss": 2.9386,
+      "step": 3027
+    },
+    {
+      "epoch": 0.5825597614352364,
+      "grad_norm": 3.268352713937481,
+      "learning_rate": 7.832984284858946e-06,
+      "loss": 2.8786,
+      "step": 3028
+    },
+    {
+      "epoch": 0.5827521523736232,
+      "grad_norm": 3.3856245520717345,
+      "learning_rate": 7.826900711230204e-06,
+      "loss": 3.1974,
+      "step": 3029
+    },
+    {
+      "epoch": 0.58294454331201,
+      "grad_norm": 3.575174394106374,
+      "learning_rate": 7.820817981608185e-06,
+      "loss": 2.989,
+      "step": 3030
+    },
+    {
+      "epoch": 0.5831369342503968,
+      "grad_norm": 2.2391338948717125,
+      "learning_rate": 7.814736098355348e-06,
+      "loss": 2.4947,
+      "step": 3031
+    },
+    {
+      "epoch": 0.5833293251887836,
+      "grad_norm": 3.3883478228514567,
+      "learning_rate": 7.808655063833832e-06,
+      "loss": 3.0011,
+      "step": 3032
+    },
+    {
+      "epoch": 0.5835217161271704,
+      "grad_norm": 3.2480899177871483,
+      "learning_rate": 7.802574880405438e-06,
+      "loss": 3.1019,
+      "step": 3033
+    },
+    {
+      "epoch": 0.5837141070655573,
+      "grad_norm": 3.581042182560137,
+      "learning_rate": 7.79649555043164e-06,
+      "loss": 3.2023,
+      "step": 3034
+    },
+    {
+      "epoch": 0.583906498003944,
+      "grad_norm": 2.3022853949104545,
+      "learning_rate": 7.790417076273581e-06,
+      "loss": 2.6907,
+      "step": 3035
+    },
+    {
+      "epoch": 0.5840988889423309,
+      "grad_norm": 3.404090162113957,
+      "learning_rate": 7.784339460292065e-06,
+      "loss": 3.0088,
+      "step": 3036
+    },
+    {
+      "epoch": 0.5842912798807176,
+      "grad_norm": 3.281219410672763,
+      "learning_rate": 7.77826270484757e-06,
+      "loss": 3.1413,
+      "step": 3037
+    },
+    {
+      "epoch": 0.5844836708191045,
+      "grad_norm": 3.755794623557134,
+      "learning_rate": 7.772186812300244e-06,
+      "loss": 3.2524,
+      "step": 3038
+    },
+    {
+      "epoch": 0.5846760617574912,
+      "grad_norm": 3.4452647495957938,
+      "learning_rate": 7.766111785009888e-06,
+      "loss": 2.7283,
+      "step": 3039
+    },
+    {
+      "epoch": 0.5848684526958781,
+      "grad_norm": 3.340809305386598,
+      "learning_rate": 7.760037625335973e-06,
+      "loss": 3.2897,
+      "step": 3040
+    },
+    {
+      "epoch": 0.5850608436342648,
+      "grad_norm": 3.6715150286990923,
+      "learning_rate": 7.753964335637634e-06,
+      "loss": 3.0428,
+      "step": 3041
+    },
+    {
+      "epoch": 0.5852532345726517,
+      "grad_norm": 3.366073340606852,
+      "learning_rate": 7.747891918273668e-06,
+      "loss": 3.1568,
+      "step": 3042
+    },
+    {
+      "epoch": 0.5854456255110384,
+      "grad_norm": 3.4573304111228658,
+      "learning_rate": 7.741820375602524e-06,
+      "loss": 2.8395,
+      "step": 3043
+    },
+    {
+      "epoch": 0.5856380164494253,
+      "grad_norm": 3.351855419771378,
+      "learning_rate": 7.735749709982329e-06,
+      "loss": 3.0336,
+      "step": 3044
+    },
+    {
+      "epoch": 0.585830407387812,
+      "grad_norm": 3.616320885738499,
+      "learning_rate": 7.729679923770855e-06,
+      "loss": 3.0559,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5860227983261989,
+      "grad_norm": 3.3038454053123396,
+      "learning_rate": 7.723611019325538e-06,
+      "loss": 3.0554,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5862151892645856,
+      "grad_norm": 3.5322948960877607,
+      "learning_rate": 7.71754299900347e-06,
+      "loss": 2.8059,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5864075802029725,
+      "grad_norm": 3.287047847950497,
+      "learning_rate": 7.7114758651614e-06,
+      "loss": 3.1064,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5865999711413592,
+      "grad_norm": 3.7989242824931075,
+      "learning_rate": 7.705409620155733e-06,
+      "loss": 3.0008,
+      "step": 3049
+    },
+    {
+      "epoch": 0.586792362079746,
+      "grad_norm": 3.4338193329894815,
+      "learning_rate": 7.699344266342529e-06,
+      "loss": 2.9649,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5869847530181328,
+      "grad_norm": 3.5017226588922705,
+      "learning_rate": 7.693279806077504e-06,
+      "loss": 2.8668,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5871771439565197,
+      "grad_norm": 3.7045620616179304,
+      "learning_rate": 7.68721624171602e-06,
+      "loss": 3.2483,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5873695348949064,
+      "grad_norm": 3.6848374010899403,
+      "learning_rate": 7.681153575613098e-06,
+      "loss": 3.138,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5875619258332933,
+      "grad_norm": 3.55830901050331,
+      "learning_rate": 7.675091810123404e-06,
+      "loss": 2.9664,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5877543167716801,
+      "grad_norm": 3.566189675020707,
+      "learning_rate": 7.669030947601265e-06,
+      "loss": 3.0675,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5879467077100669,
+      "grad_norm": 3.3370122218053906,
+      "learning_rate": 7.662970990400647e-06,
+      "loss": 3.071,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5881390986484537,
+      "grad_norm": 2.213911922471267,
+      "learning_rate": 7.656911940875163e-06,
+      "loss": 2.6345,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5883314895868405,
+      "grad_norm": 3.701494469769157,
+      "learning_rate": 7.650853801378084e-06,
+      "loss": 3.0193,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5885238805252273,
+      "grad_norm": 3.3641510450430254,
+      "learning_rate": 7.644796574262322e-06,
+      "loss": 3.0529,
+      "step": 3059
+    },
+    {
+      "epoch": 0.588716271463614,
+      "grad_norm": 3.445078075274408,
+      "learning_rate": 7.638740261880423e-06,
+      "loss": 3.0011,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5889086624020009,
+      "grad_norm": 3.2657803285365694,
+      "learning_rate": 7.632684866584606e-06,
+      "loss": 3.0141,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5891010533403876,
+      "grad_norm": 3.0642967210277825,
+      "learning_rate": 7.626630390726704e-06,
+      "loss": 2.8915,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5892934442787745,
+      "grad_norm": 3.265320765738573,
+      "learning_rate": 7.620576836658212e-06,
+      "loss": 3.0631,
+      "step": 3063
+    },
+    {
+      "epoch": 0.5894858352171612,
+      "grad_norm": 3.6852411318457445,
+      "learning_rate": 7.61452420673026e-06,
+      "loss": 3.2117,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5896782261555481,
+      "grad_norm": 3.391883275288293,
+      "learning_rate": 7.608472503293615e-06,
+      "loss": 3.0619,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5898706170939348,
+      "grad_norm": 3.6346285225787027,
+      "learning_rate": 7.60242172869869e-06,
+      "loss": 2.9027,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5900630080323217,
+      "grad_norm": 3.5537763513004603,
+      "learning_rate": 7.596371885295542e-06,
+      "loss": 3.1737,
+      "step": 3067
+    },
+    {
+      "epoch": 0.5902553989707084,
+      "grad_norm": 3.8541933900367678,
+      "learning_rate": 7.590322975433857e-06,
+      "loss": 3.2539,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5904477899090953,
+      "grad_norm": 3.483419947787561,
+      "learning_rate": 7.584275001462961e-06,
+      "loss": 3.1693,
+      "step": 3069
+    },
+    {
+      "epoch": 0.590640180847482,
+      "grad_norm": 3.5939217220791337,
+      "learning_rate": 7.578227965731819e-06,
+      "loss": 3.1168,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5908325717858689,
+      "grad_norm": 3.405423320585252,
+      "learning_rate": 7.572181870589028e-06,
+      "loss": 2.9785,
+      "step": 3071
+    },
+    {
+      "epoch": 0.5910249627242556,
+      "grad_norm": 3.288333388009658,
+      "learning_rate": 7.566136718382821e-06,
+      "loss": 2.988,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5912173536626425,
+      "grad_norm": 2.3422149828831733,
+      "learning_rate": 7.560092511461069e-06,
+      "loss": 2.5153,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5914097446010292,
+      "grad_norm": 3.3007577584327747,
+      "learning_rate": 7.55404925217127e-06,
+      "loss": 2.9639,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5916021355394161,
+      "grad_norm": 3.2596658213584933,
+      "learning_rate": 7.548006942860557e-06,
+      "loss": 3.2312,
+      "step": 3075
+    },
+    {
+      "epoch": 0.5917945264778028,
+      "grad_norm": 3.588245735459122,
+      "learning_rate": 7.541965585875695e-06,
+      "loss": 3.0178,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5919869174161897,
+      "grad_norm": 4.018567159837494,
+      "learning_rate": 7.535925183563073e-06,
+      "loss": 2.7857,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5921793083545766,
+      "grad_norm": 3.3348116625443263,
+      "learning_rate": 7.529885738268714e-06,
+      "loss": 2.905,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5923716992929633,
+      "grad_norm": 3.5360645702900015,
+      "learning_rate": 7.523847252338274e-06,
+      "loss": 2.8994,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5925640902313501,
+      "grad_norm": 3.5383593109007965,
+      "learning_rate": 7.51780972811703e-06,
+      "loss": 3.0906,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5927564811697369,
+      "grad_norm": 3.445028607879897,
+      "learning_rate": 7.511773167949885e-06,
+      "loss": 3.024,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5929488721081237,
+      "grad_norm": 3.582399385431015,
+      "learning_rate": 7.5057375741813685e-06,
+      "loss": 2.8427,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5931412630465105,
+      "grad_norm": 3.544001734748186,
+      "learning_rate": 7.499702949155634e-06,
+      "loss": 3.0147,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5933336539848973,
+      "grad_norm": 3.3853262144368848,
+      "learning_rate": 7.493669295216467e-06,
+      "loss": 3.0378,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5935260449232841,
+      "grad_norm": 3.1979273298496906,
+      "learning_rate": 7.487636614707265e-06,
+      "loss": 3.0738,
+      "step": 3085
+    },
+    {
+      "epoch": 0.593718435861671,
+      "grad_norm": 3.5554023002037507,
+      "learning_rate": 7.48160490997105e-06,
+      "loss": 2.9468,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5939108268000577,
+      "grad_norm": 4.070002639283372,
+      "learning_rate": 7.475574183350471e-06,
+      "loss": 3.4172,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5941032177384445,
+      "grad_norm": 3.8309352366969587,
+      "learning_rate": 7.46954443718779e-06,
+      "loss": 2.9597,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5942956086768313,
+      "grad_norm": 3.673524081739812,
+      "learning_rate": 7.463515673824888e-06,
+      "loss": 3.0698,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5944879996152181,
+      "grad_norm": 3.4546609999210607,
+      "learning_rate": 7.457487895603273e-06,
+      "loss": 2.9959,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5946803905536049,
+      "grad_norm": 3.1153238412569597,
+      "learning_rate": 7.451461104864061e-06,
+      "loss": 2.8207,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5948727814919917,
+      "grad_norm": 3.4919737096791352,
+      "learning_rate": 7.44543530394799e-06,
+      "loss": 3.0623,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5950651724303785,
+      "grad_norm": 3.520909706905806,
+      "learning_rate": 7.439410495195411e-06,
+      "loss": 2.99,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5952575633687653,
+      "grad_norm": 3.355800071269815,
+      "learning_rate": 7.433386680946288e-06,
+      "loss": 2.9033,
+      "step": 3094
+    },
+    {
+      "epoch": 0.5954499543071521,
+      "grad_norm": 3.4449105687490515,
+      "learning_rate": 7.427363863540202e-06,
+      "loss": 3.0093,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5956423452455389,
+      "grad_norm": 3.6087143934134884,
+      "learning_rate": 7.421342045316351e-06,
+      "loss": 3.3827,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5958347361839257,
+      "grad_norm": 3.595066227353702,
+      "learning_rate": 7.415321228613534e-06,
+      "loss": 2.9165,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5960271271223125,
+      "grad_norm": 3.306305121217399,
+      "learning_rate": 7.409301415770168e-06,
+      "loss": 2.9028,
+      "step": 3098
+    },
+    {
+      "epoch": 0.5962195180606994,
+      "grad_norm": 2.3511255640124755,
+      "learning_rate": 7.403282609124281e-06,
+      "loss": 2.5814,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5964119089990861,
+      "grad_norm": 3.487749294241079,
+      "learning_rate": 7.397264811013507e-06,
+      "loss": 3.0101,
+      "step": 3100
+    },
+    {
+      "epoch": 0.596604299937473,
+      "grad_norm": 3.6328830518172244,
+      "learning_rate": 7.391248023775084e-06,
+      "loss": 2.9919,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5967966908758597,
+      "grad_norm": 3.419236705707502,
+      "learning_rate": 7.385232249745873e-06,
+      "loss": 2.7885,
+      "step": 3102
+    },
+    {
+      "epoch": 0.5969890818142466,
+      "grad_norm": 3.422289125378827,
+      "learning_rate": 7.379217491262325e-06,
+      "loss": 3.1263,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5971814727526333,
+      "grad_norm": 3.2809091597821656,
+      "learning_rate": 7.373203750660505e-06,
+      "loss": 2.7861,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5973738636910202,
+      "grad_norm": 3.3936918863648367,
+      "learning_rate": 7.36719103027608e-06,
+      "loss": 2.949,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5975662546294069,
+      "grad_norm": 3.3005139979212266,
+      "learning_rate": 7.361179332444318e-06,
+      "loss": 3.0251,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5977586455677938,
+      "grad_norm": 3.368078931616573,
+      "learning_rate": 7.355168659500094e-06,
+      "loss": 3.0145,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5979510365061805,
+      "grad_norm": 3.6440102733372837,
+      "learning_rate": 7.3491590137778915e-06,
+      "loss": 2.9011,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5981434274445674,
+      "grad_norm": 3.436971678504415,
+      "learning_rate": 7.343150397611782e-06,
+      "loss": 2.9608,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5983358183829541,
+      "grad_norm": 3.400366696445,
+      "learning_rate": 7.3371428133354435e-06,
+      "loss": 2.7891,
+      "step": 3110
+    },
+    {
+      "epoch": 0.598528209321341,
+      "grad_norm": 3.5433515174088064,
+      "learning_rate": 7.33113626328215e-06,
+      "loss": 2.8539,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5987206002597277,
+      "grad_norm": 3.243742678266504,
+      "learning_rate": 7.325130749784781e-06,
+      "loss": 3.0169,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5989129911981146,
+      "grad_norm": 3.873058722936152,
+      "learning_rate": 7.3191262751758005e-06,
+      "loss": 2.8933,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5991053821365013,
+      "grad_norm": 3.2155897821583865,
+      "learning_rate": 7.3131228417872905e-06,
+      "loss": 2.6388,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5992977730748882,
+      "grad_norm": 3.823301913395358,
+      "learning_rate": 7.307120451950902e-06,
+      "loss": 3.2579,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5994901640132749,
+      "grad_norm": 3.3479290275431244,
+      "learning_rate": 7.301119107997905e-06,
+      "loss": 2.9136,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5996825549516618,
+      "grad_norm": 3.2460318513466393,
+      "learning_rate": 7.295118812259145e-06,
+      "loss": 3.1201,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5998749458900485,
+      "grad_norm": 3.4050582431173555,
+      "learning_rate": 7.289119567065068e-06,
+      "loss": 2.8828,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6000673368284354,
+      "grad_norm": 3.5859635926103364,
+      "learning_rate": 7.2831213747457155e-06,
+      "loss": 2.87,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6002597277668222,
+      "grad_norm": 3.748480659596424,
+      "learning_rate": 7.2771242376307125e-06,
+      "loss": 3.4091,
+      "step": 3120
+    },
+    {
+      "epoch": 0.600452118705209,
+      "grad_norm": 3.37931143420728,
+      "learning_rate": 7.271128158049283e-06,
+      "loss": 3.2656,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6006445096435958,
+      "grad_norm": 3.6348663086833697,
+      "learning_rate": 7.2651331383302326e-06,
+      "loss": 2.8314,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6008369005819826,
+      "grad_norm": 3.4210486689205015,
+      "learning_rate": 7.2591391808019555e-06,
+      "loss": 3.2925,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6010292915203694,
+      "grad_norm": 3.098564180265442,
+      "learning_rate": 7.253146287792434e-06,
+      "loss": 2.7346,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6012216824587562,
+      "grad_norm": 3.3278854813639214,
+      "learning_rate": 7.247154461629248e-06,
+      "loss": 3.0224,
+      "step": 3125
+    },
+    {
+      "epoch": 0.601414073397143,
+      "grad_norm": 3.4300508435051036,
+      "learning_rate": 7.241163704639547e-06,
+      "loss": 3.0149,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6016064643355298,
+      "grad_norm": 3.2958140631070774,
+      "learning_rate": 7.235174019150071e-06,
+      "loss": 3.006,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6017988552739166,
+      "grad_norm": 3.1176269160560017,
+      "learning_rate": 7.229185407487149e-06,
+      "loss": 2.9163,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6019912462123034,
+      "grad_norm": 3.526348539427453,
+      "learning_rate": 7.2231978719766884e-06,
+      "loss": 3.1718,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6021836371506902,
+      "grad_norm": 3.757462358291769,
+      "learning_rate": 7.217211414944171e-06,
+      "loss": 2.863,
+      "step": 3130
+    },
+    {
+      "epoch": 0.602376028089077,
+      "grad_norm": 3.3578111799259722,
+      "learning_rate": 7.2112260387146784e-06,
+      "loss": 2.9993,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6025684190274638,
+      "grad_norm": 3.3582485432899283,
+      "learning_rate": 7.2052417456128565e-06,
+      "loss": 2.684,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6027608099658506,
+      "grad_norm": 3.5806988018143593,
+      "learning_rate": 7.199258537962936e-06,
+      "loss": 3.0472,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6029532009042374,
+      "grad_norm": 2.1653922348311667,
+      "learning_rate": 7.193276418088729e-06,
+      "loss": 2.6214,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6031455918426242,
+      "grad_norm": 3.3542423761179068,
+      "learning_rate": 7.187295388313618e-06,
+      "loss": 2.9758,
+      "step": 3135
+    },
+    {
+      "epoch": 0.603337982781011,
+      "grad_norm": 3.4048342260610727,
+      "learning_rate": 7.181315450960562e-06,
+      "loss": 2.8211,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6035303737193978,
+      "grad_norm": 3.4850598074327808,
+      "learning_rate": 7.175336608352113e-06,
+      "loss": 3.0592,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6037227646577846,
+      "grad_norm": 3.2364875758750182,
+      "learning_rate": 7.169358862810374e-06,
+      "loss": 3.1616,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6039151555961714,
+      "grad_norm": 3.742702483214844,
+      "learning_rate": 7.163382216657033e-06,
+      "loss": 2.9775,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6041075465345582,
+      "grad_norm": 3.3865835242315336,
+      "learning_rate": 7.1574066722133565e-06,
+      "loss": 3.3371,
+      "step": 3140
+    },
+    {
+      "epoch": 0.604299937472945,
+      "grad_norm": 3.5314264911982707,
+      "learning_rate": 7.151432231800173e-06,
+      "loss": 3.233,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6044923284113318,
+      "grad_norm": 3.6092800830013556,
+      "learning_rate": 7.145458897737882e-06,
+      "loss": 2.978,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6046847193497187,
+      "grad_norm": 3.4987756663439127,
+      "learning_rate": 7.139486672346466e-06,
+      "loss": 2.9574,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6048771102881054,
+      "grad_norm": 3.4921694626753297,
+      "learning_rate": 7.133515557945463e-06,
+      "loss": 3.0013,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6050695012264923,
+      "grad_norm": 3.607330363309266,
+      "learning_rate": 7.12754555685399e-06,
+      "loss": 3.1353,
+      "step": 3145
+    },
+    {
+      "epoch": 0.605261892164879,
+      "grad_norm": 3.174229488423223,
+      "learning_rate": 7.121576671390722e-06,
+      "loss": 2.8778,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6054542831032659,
+      "grad_norm": 3.306266375667237,
+      "learning_rate": 7.115608903873905e-06,
+      "loss": 3.1216,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6056466740416526,
+      "grad_norm": 2.8733769023282902,
+      "learning_rate": 7.109642256621353e-06,
+      "loss": 2.8598,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6058390649800395,
+      "grad_norm": 2.203623228894518,
+      "learning_rate": 7.103676731950443e-06,
+      "loss": 2.5191,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6060314559184262,
+      "grad_norm": 3.3611368481923467,
+      "learning_rate": 7.0977123321781176e-06,
+      "loss": 3.1719,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6062238468568131,
+      "grad_norm": 3.312278309115096,
+      "learning_rate": 7.091749059620881e-06,
+      "loss": 2.8939,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6064162377951998,
+      "grad_norm": 3.3931110475570194,
+      "learning_rate": 7.0857869165947945e-06,
+      "loss": 3.2059,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6066086287335867,
+      "grad_norm": 3.5159985394072013,
+      "learning_rate": 7.079825905415491e-06,
+      "loss": 3.1346,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6068010196719734,
+      "grad_norm": 3.2752811634243884,
+      "learning_rate": 7.073866028398153e-06,
+      "loss": 2.7267,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6069934106103603,
+      "grad_norm": 3.4660737804268456,
+      "learning_rate": 7.067907287857535e-06,
+      "loss": 3.2446,
+      "step": 3155
+    },
+    {
+      "epoch": 0.607185801548747,
+      "grad_norm": 3.6512329590358186,
+      "learning_rate": 7.061949686107938e-06,
+      "loss": 3.0949,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6073781924871339,
+      "grad_norm": 3.639876240639287,
+      "learning_rate": 7.0559932254632315e-06,
+      "loss": 3.0725,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6075705834255206,
+      "grad_norm": 3.39698862965262,
+      "learning_rate": 7.0500379082368305e-06,
+      "loss": 2.9313,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6077629743639075,
+      "grad_norm": 3.719398650966615,
+      "learning_rate": 7.044083736741711e-06,
+      "loss": 3.3202,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6079553653022942,
+      "grad_norm": 2.2680747618689714,
+      "learning_rate": 7.03813071329041e-06,
+      "loss": 2.6959,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6081477562406811,
+      "grad_norm": 2.3971776460524077,
+      "learning_rate": 7.032178840195009e-06,
+      "loss": 2.6066,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6083401471790678,
+      "grad_norm": 3.3183890430596947,
+      "learning_rate": 7.026228119767149e-06,
+      "loss": 3.0563,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6085325381174547,
+      "grad_norm": 3.3044897879941675,
+      "learning_rate": 7.020278554318023e-06,
+      "loss": 3.0149,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6087249290558415,
+      "grad_norm": 3.6789172892446422,
+      "learning_rate": 7.014330146158367e-06,
+      "loss": 2.9888,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6089173199942283,
+      "grad_norm": 3.6729119833667316,
+      "learning_rate": 7.008382897598477e-06,
+      "loss": 3.2416,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6091097109326151,
+      "grad_norm": 3.5378914435640225,
+      "learning_rate": 7.002436810948201e-06,
+      "loss": 3.0836,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6093021018710019,
+      "grad_norm": 3.586562302623624,
+      "learning_rate": 6.996491888516927e-06,
+      "loss": 3.0799,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6094944928093887,
+      "grad_norm": 3.31250536530379,
+      "learning_rate": 6.990548132613592e-06,
+      "loss": 2.8737,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6096868837477755,
+      "grad_norm": 3.544617644928949,
+      "learning_rate": 6.984605545546686e-06,
+      "loss": 3.176,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6098792746861623,
+      "grad_norm": 3.219858357264792,
+      "learning_rate": 6.978664129624241e-06,
+      "loss": 3.0272,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6100716656245491,
+      "grad_norm": 3.818163839892229,
+      "learning_rate": 6.972723887153828e-06,
+      "loss": 2.8507,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6102640565629359,
+      "grad_norm": 3.620259227955746,
+      "learning_rate": 6.9667848204425785e-06,
+      "loss": 2.9165,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6104564475013227,
+      "grad_norm": 3.7304517826572203,
+      "learning_rate": 6.960846931797152e-06,
+      "loss": 3.0253,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6106488384397095,
+      "grad_norm": 3.560039270196854,
+      "learning_rate": 6.9549102235237565e-06,
+      "loss": 3.0118,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6108412293780963,
+      "grad_norm": 3.543213725082684,
+      "learning_rate": 6.948974697928144e-06,
+      "loss": 3.0159,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6110336203164831,
+      "grad_norm": 2.4226556660068383,
+      "learning_rate": 6.943040357315598e-06,
+      "loss": 2.5968,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6112260112548699,
+      "grad_norm": 3.400882883399907,
+      "learning_rate": 6.9371072039909515e-06,
+      "loss": 2.8642,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6114184021932567,
+      "grad_norm": 2.236368591069986,
+      "learning_rate": 6.931175240258576e-06,
+      "loss": 2.5316,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6116107931316435,
+      "grad_norm": 3.211345712004628,
+      "learning_rate": 6.9252444684223765e-06,
+      "loss": 2.98,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6118031840700303,
+      "grad_norm": 3.1421801685662083,
+      "learning_rate": 6.919314890785793e-06,
+      "loss": 3.0485,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6119955750084171,
+      "grad_norm": 3.4535292484864675,
+      "learning_rate": 6.913386509651807e-06,
+      "loss": 3.133,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6121879659468039,
+      "grad_norm": 3.6690787239284512,
+      "learning_rate": 6.907459327322934e-06,
+      "loss": 3.1773,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6123803568851907,
+      "grad_norm": 3.6914510501220725,
+      "learning_rate": 6.90153334610122e-06,
+      "loss": 3.0082,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6125727478235775,
+      "grad_norm": 3.4876964352934925,
+      "learning_rate": 6.895608568288255e-06,
+      "loss": 2.9954,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6127651387619643,
+      "grad_norm": 3.7908789021393234,
+      "learning_rate": 6.889684996185148e-06,
+      "loss": 2.9853,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6129575297003511,
+      "grad_norm": 3.498901342936075,
+      "learning_rate": 6.88376263209255e-06,
+      "loss": 3.2433,
+      "step": 3186
+    },
+    {
+      "epoch": 0.613149920638738,
+      "grad_norm": 3.5051542130329114,
+      "learning_rate": 6.877841478310639e-06,
+      "loss": 3.1169,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6133423115771247,
+      "grad_norm": 3.448614903740431,
+      "learning_rate": 6.871921537139117e-06,
+      "loss": 2.9843,
+      "step": 3188
+    },
+    {
+      "epoch": 0.6135347025155116,
+      "grad_norm": 3.156814020159379,
+      "learning_rate": 6.866002810877224e-06,
+      "loss": 2.9418,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6137270934538983,
+      "grad_norm": 3.8101238370114654,
+      "learning_rate": 6.860085301823729e-06,
+      "loss": 3.1439,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6139194843922852,
+      "grad_norm": 3.614880249135818,
+      "learning_rate": 6.854169012276923e-06,
+      "loss": 2.7752,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6141118753306719,
+      "grad_norm": 3.367434479421538,
+      "learning_rate": 6.848253944534622e-06,
+      "loss": 3.0766,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6143042662690588,
+      "grad_norm": 3.3225868473902267,
+      "learning_rate": 6.84234010089417e-06,
+      "loss": 2.8818,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6144966572074455,
+      "grad_norm": 3.258495841579265,
+      "learning_rate": 6.836427483652436e-06,
+      "loss": 3.0047,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6146890481458324,
+      "grad_norm": 3.4070506415282544,
+      "learning_rate": 6.830516095105817e-06,
+      "loss": 2.8596,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6148814390842191,
+      "grad_norm": 3.432980732938711,
+      "learning_rate": 6.824605937550224e-06,
+      "loss": 3.0833,
+      "step": 3196
+    },
+    {
+      "epoch": 0.615073830022606,
+      "grad_norm": 3.512127216853676,
+      "learning_rate": 6.818697013281093e-06,
+      "loss": 3.1045,
+      "step": 3197
+    },
+    {
+      "epoch": 0.6152662209609927,
+      "grad_norm": 3.532987291616427,
+      "learning_rate": 6.8127893245933864e-06,
+      "loss": 2.8308,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6154586118993796,
+      "grad_norm": 3.5785220943316545,
+      "learning_rate": 6.806882873781579e-06,
+      "loss": 3.1378,
+      "step": 3199
+    },
+    {
+      "epoch": 0.6156510028377663,
+      "grad_norm": 3.3337878725072327,
+      "learning_rate": 6.800977663139666e-06,
+      "loss": 3.05,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6158433937761532,
+      "grad_norm": 3.5505018694922286,
+      "learning_rate": 6.795073694961171e-06,
+      "loss": 3.1008,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6160357847145399,
+      "grad_norm": 3.1316921294529725,
+      "learning_rate": 6.789170971539119e-06,
+      "loss": 3.0092,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6162281756529268,
+      "grad_norm": 3.7400682358043804,
+      "learning_rate": 6.783269495166066e-06,
+      "loss": 2.6934,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6164205665913135,
+      "grad_norm": 3.3353675464808754,
+      "learning_rate": 6.777369268134076e-06,
+      "loss": 2.9647,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6166129575297004,
+      "grad_norm": 3.3758436834436436,
+      "learning_rate": 6.771470292734723e-06,
+      "loss": 2.9453,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6168053484680871,
+      "grad_norm": 3.331236152868043,
+      "learning_rate": 6.7655725712591055e-06,
+      "loss": 3.0155,
+      "step": 3206
+    },
+    {
+      "epoch": 0.616997739406474,
+      "grad_norm": 3.345955672796357,
+      "learning_rate": 6.759676105997834e-06,
+      "loss": 3.111,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6171901303448608,
+      "grad_norm": 3.3180561835627156,
+      "learning_rate": 6.753780899241027e-06,
+      "loss": 2.9063,
+      "step": 3208
+    },
+    {
+      "epoch": 0.6173825212832476,
+      "grad_norm": 3.312941282890529,
+      "learning_rate": 6.747886953278311e-06,
+      "loss": 3.2274,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6175749122216344,
+      "grad_norm": 3.277875219729665,
+      "learning_rate": 6.741994270398826e-06,
+      "loss": 2.79,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6177673031600212,
+      "grad_norm": 3.309477682844906,
+      "learning_rate": 6.736102852891227e-06,
+      "loss": 2.8622,
+      "step": 3211
+    },
+    {
+      "epoch": 0.617959694098408,
+      "grad_norm": 3.358900090354382,
+      "learning_rate": 6.730212703043666e-06,
+      "loss": 2.793,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6181520850367948,
+      "grad_norm": 2.4729098500631883,
+      "learning_rate": 6.7243238231438176e-06,
+      "loss": 2.745,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6183444759751816,
+      "grad_norm": 3.3793230328720436,
+      "learning_rate": 6.718436215478849e-06,
+      "loss": 2.8696,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6185368669135684,
+      "grad_norm": 3.403431080440354,
+      "learning_rate": 6.712549882335442e-06,
+      "loss": 3.1922,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6187292578519552,
+      "grad_norm": 3.24454604384241,
+      "learning_rate": 6.70666482599978e-06,
+      "loss": 3.0658,
+      "step": 3216
+    },
+    {
+      "epoch": 0.618921648790342,
+      "grad_norm": 3.5435074803393154,
+      "learning_rate": 6.7007810487575475e-06,
+      "loss": 3.2842,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6191140397287288,
+      "grad_norm": 3.669195909654402,
+      "learning_rate": 6.694898552893941e-06,
+      "loss": 2.9529,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6193064306671155,
+      "grad_norm": 3.5244662338142856,
+      "learning_rate": 6.6890173406936485e-06,
+      "loss": 3.1623,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6194988216055024,
+      "grad_norm": 3.7479402101266444,
+      "learning_rate": 6.683137414440872e-06,
+      "loss": 3.1102,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6196912125438891,
+      "grad_norm": 3.48198674726512,
+      "learning_rate": 6.677258776419304e-06,
+      "loss": 3.3272,
+      "step": 3221
+    },
+    {
+      "epoch": 0.619883603482276,
+      "grad_norm": 3.3614636951139345,
+      "learning_rate": 6.671381428912138e-06,
+      "loss": 3.151,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6200759944206627,
+      "grad_norm": 3.488042605501595,
+      "learning_rate": 6.66550537420207e-06,
+      "loss": 2.9867,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6202683853590496,
+      "grad_norm": 3.2482353750153057,
+      "learning_rate": 6.659630614571287e-06,
+      "loss": 3.0344,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6204607762974363,
+      "grad_norm": 3.635587198279006,
+      "learning_rate": 6.653757152301488e-06,
+      "loss": 3.2666,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6206531672358232,
+      "grad_norm": 3.422597070444857,
+      "learning_rate": 6.647884989673849e-06,
+      "loss": 3.1003,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6208455581742099,
+      "grad_norm": 3.2886320786625522,
+      "learning_rate": 6.642014128969055e-06,
+      "loss": 3.185,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6210379491125968,
+      "grad_norm": 3.3111279925033177,
+      "learning_rate": 6.63614457246728e-06,
+      "loss": 2.956,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6212303400509837,
+      "grad_norm": 3.4114206246724263,
+      "learning_rate": 6.630276322448188e-06,
+      "loss": 2.9542,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6214227309893704,
+      "grad_norm": 3.287293374426287,
+      "learning_rate": 6.624409381190946e-06,
+      "loss": 2.9783,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6216151219277573,
+      "grad_norm": 3.122961866306714,
+      "learning_rate": 6.618543750974202e-06,
+      "loss": 2.9208,
+      "step": 3231
+    },
+    {
+      "epoch": 0.621807512866144,
+      "grad_norm": 3.620165939305507,
+      "learning_rate": 6.6126794340761025e-06,
+      "loss": 3.1486,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6219999038045309,
+      "grad_norm": 3.443138880368688,
+      "learning_rate": 6.606816432774279e-06,
+      "loss": 3.3667,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6221922947429176,
+      "grad_norm": 3.5364417154317027,
+      "learning_rate": 6.600954749345851e-06,
+      "loss": 3.0742,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6223846856813045,
+      "grad_norm": 3.906403258065009,
+      "learning_rate": 6.595094386067428e-06,
+      "loss": 2.9133,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6225770766196912,
+      "grad_norm": 3.3596341463525285,
+      "learning_rate": 6.589235345215117e-06,
+      "loss": 3.1142,
+      "step": 3236
+    },
+    {
+      "epoch": 0.622769467558078,
+      "grad_norm": 3.4848824561424276,
+      "learning_rate": 6.583377629064494e-06,
+      "loss": 2.7936,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6229618584964648,
+      "grad_norm": 3.304709872308674,
+      "learning_rate": 6.5775212398906295e-06,
+      "loss": 3.1847,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6231542494348516,
+      "grad_norm": 3.6151575110176206,
+      "learning_rate": 6.571666179968079e-06,
+      "loss": 3.1213,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6233466403732384,
+      "grad_norm": 3.512235338634056,
+      "learning_rate": 6.565812451570881e-06,
+      "loss": 3.0209,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6235390313116252,
+      "grad_norm": 3.9962359109895735,
+      "learning_rate": 6.5599600569725495e-06,
+      "loss": 3.2633,
+      "step": 3241
+    },
+    {
+      "epoch": 0.623731422250012,
+      "grad_norm": 3.658877133352506,
+      "learning_rate": 6.554108998446096e-06,
+      "loss": 2.9449,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6239238131883988,
+      "grad_norm": 3.311862530234695,
+      "learning_rate": 6.548259278263999e-06,
+      "loss": 2.9681,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6241162041267856,
+      "grad_norm": 3.421959532256755,
+      "learning_rate": 6.542410898698226e-06,
+      "loss": 2.9131,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6243085950651724,
+      "grad_norm": 3.4007388788069575,
+      "learning_rate": 6.536563862020218e-06,
+      "loss": 2.852,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6245009860035592,
+      "grad_norm": 3.3924909365183553,
+      "learning_rate": 6.530718170500896e-06,
+      "loss": 2.9192,
+      "step": 3246
+    },
+    {
+      "epoch": 0.624693376941946,
+      "grad_norm": 3.7972341645657623,
+      "learning_rate": 6.524873826410658e-06,
+      "loss": 3.2341,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6248857678803328,
+      "grad_norm": 3.5132342374391565,
+      "learning_rate": 6.519030832019383e-06,
+      "loss": 3.1084,
+      "step": 3248
+    },
+    {
+      "epoch": 0.6250781588187196,
+      "grad_norm": 3.1786800091233713,
+      "learning_rate": 6.513189189596422e-06,
+      "loss": 3.0301,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6252705497571064,
+      "grad_norm": 3.6056471731398227,
+      "learning_rate": 6.507348901410604e-06,
+      "loss": 3.0628,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6254629406954932,
+      "grad_norm": 3.6048715883742486,
+      "learning_rate": 6.501509969730224e-06,
+      "loss": 3.1435,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6256553316338801,
+      "grad_norm": 3.422058420739333,
+      "learning_rate": 6.495672396823061e-06,
+      "loss": 3.1608,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6258477225722668,
+      "grad_norm": 3.6459049450862655,
+      "learning_rate": 6.489836184956353e-06,
+      "loss": 3.159,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6260401135106537,
+      "grad_norm": 3.8349486883759583,
+      "learning_rate": 6.484001336396828e-06,
+      "loss": 3.0791,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6262325044490404,
+      "grad_norm": 3.338462416778856,
+      "learning_rate": 6.478167853410668e-06,
+      "loss": 2.8739,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6264248953874273,
+      "grad_norm": 3.6168175685907267,
+      "learning_rate": 6.472335738263534e-06,
+      "loss": 3.262,
+      "step": 3256
+    },
+    {
+      "epoch": 0.626617286325814,
+      "grad_norm": 3.302902004863746,
+      "learning_rate": 6.466504993220548e-06,
+      "loss": 3.0352,
+      "step": 3257
+    },
+    {
+      "epoch": 0.6268096772642009,
+      "grad_norm": 3.521123585992802,
+      "learning_rate": 6.460675620546305e-06,
+      "loss": 3.0292,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6270020682025876,
+      "grad_norm": 3.464972309265321,
+      "learning_rate": 6.454847622504867e-06,
+      "loss": 2.9111,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6271944591409745,
+      "grad_norm": 3.1872322281091816,
+      "learning_rate": 6.4490210013597635e-06,
+      "loss": 3.2731,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6273868500793612,
+      "grad_norm": 3.3426572409937454,
+      "learning_rate": 6.4431957593739845e-06,
+      "loss": 2.9673,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6275792410177481,
+      "grad_norm": 3.274957893202695,
+      "learning_rate": 6.4373718988099896e-06,
+      "loss": 2.8579,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6277716319561348,
+      "grad_norm": 3.3349367676697304,
+      "learning_rate": 6.431549421929694e-06,
+      "loss": 2.9395,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6279640228945217,
+      "grad_norm": 3.2997148197495756,
+      "learning_rate": 6.4257283309944804e-06,
+      "loss": 2.9885,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6281564138329084,
+      "grad_norm": 3.228222227580711,
+      "learning_rate": 6.419908628265203e-06,
+      "loss": 3.0531,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6283488047712953,
+      "grad_norm": 3.446157460306644,
+      "learning_rate": 6.414090316002161e-06,
+      "loss": 2.7967,
+      "step": 3266
+    },
+    {
+      "epoch": 0.628541195709682,
+      "grad_norm": 3.346536607444672,
+      "learning_rate": 6.4082733964651166e-06,
+      "loss": 3.0634,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6287335866480689,
+      "grad_norm": 3.3057578755619232,
+      "learning_rate": 6.4024578719133e-06,
+      "loss": 2.9587,
+      "step": 3268
+    },
+    {
+      "epoch": 0.6289259775864556,
+      "grad_norm": 3.3441336500739625,
+      "learning_rate": 6.396643744605391e-06,
+      "loss": 2.9484,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6291183685248425,
+      "grad_norm": 3.422486816731877,
+      "learning_rate": 6.390831016799527e-06,
+      "loss": 2.7449,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6293107594632292,
+      "grad_norm": 3.8265741157884374,
+      "learning_rate": 6.385019690753311e-06,
+      "loss": 3.2865,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6295031504016161,
+      "grad_norm": 3.365697821931397,
+      "learning_rate": 6.379209768723791e-06,
+      "loss": 3.0227,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6296955413400029,
+      "grad_norm": 3.704697265229041,
+      "learning_rate": 6.373401252967475e-06,
+      "loss": 3.0749,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6298879322783897,
+      "grad_norm": 3.4170175187932963,
+      "learning_rate": 6.367594145740324e-06,
+      "loss": 3.1846,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6300803232167765,
+      "grad_norm": 3.473596669658972,
+      "learning_rate": 6.361788449297748e-06,
+      "loss": 2.8329,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6302727141551633,
+      "grad_norm": 3.3170989112568754,
+      "learning_rate": 6.355984165894613e-06,
+      "loss": 3.1242,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6304651050935501,
+      "grad_norm": 3.3791940859385305,
+      "learning_rate": 6.350181297785242e-06,
+      "loss": 2.9973,
+      "step": 3277
+    },
+    {
+      "epoch": 0.6306574960319369,
+      "grad_norm": 2.447819267442939,
+      "learning_rate": 6.344379847223398e-06,
+      "loss": 2.6004,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6308498869703237,
+      "grad_norm": 3.2726908335822014,
+      "learning_rate": 6.338579816462298e-06,
+      "loss": 2.8506,
+      "step": 3279
+    },
+    {
+      "epoch": 0.6310422779087105,
+      "grad_norm": 3.3625888586605064,
+      "learning_rate": 6.332781207754605e-06,
+      "loss": 3.2284,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6312346688470973,
+      "grad_norm": 3.769085479578533,
+      "learning_rate": 6.326984023352435e-06,
+      "loss": 3.0307,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6314270597854841,
+      "grad_norm": 3.639066321187574,
+      "learning_rate": 6.321188265507342e-06,
+      "loss": 3.3117,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6316194507238709,
+      "grad_norm": 2.3609954864503657,
+      "learning_rate": 6.31539393647034e-06,
+      "loss": 2.6845,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6318118416622577,
+      "grad_norm": 3.301589766722046,
+      "learning_rate": 6.309601038491874e-06,
+      "loss": 3.0102,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6320042326006445,
+      "grad_norm": 3.4952251651549826,
+      "learning_rate": 6.303809573821842e-06,
+      "loss": 2.959,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6321966235390313,
+      "grad_norm": 3.752767908604963,
+      "learning_rate": 6.298019544709579e-06,
+      "loss": 2.9472,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6323890144774181,
+      "grad_norm": 3.38067040194525,
+      "learning_rate": 6.292230953403866e-06,
+      "loss": 3.2113,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6325814054158049,
+      "grad_norm": 3.3091884263126743,
+      "learning_rate": 6.286443802152926e-06,
+      "loss": 3.2757,
+      "step": 3288
+    },
+    {
+      "epoch": 0.6327737963541917,
+      "grad_norm": 3.333508999154406,
+      "learning_rate": 6.280658093204422e-06,
+      "loss": 2.9535,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6329661872925785,
+      "grad_norm": 4.008037274777768,
+      "learning_rate": 6.274873828805459e-06,
+      "loss": 3.094,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6331585782309653,
+      "grad_norm": 3.30987584318709,
+      "learning_rate": 6.269091011202576e-06,
+      "loss": 3.0434,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6333509691693521,
+      "grad_norm": 3.5145777751907965,
+      "learning_rate": 6.263309642641751e-06,
+      "loss": 3.048,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6335433601077389,
+      "grad_norm": 3.1162465544003206,
+      "learning_rate": 6.257529725368405e-06,
+      "loss": 2.9885,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6337357510461257,
+      "grad_norm": 3.210799887490705,
+      "learning_rate": 6.251751261627386e-06,
+      "loss": 3.037,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6339281419845125,
+      "grad_norm": 3.6823013031179164,
+      "learning_rate": 6.245974253662988e-06,
+      "loss": 2.866,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6341205329228994,
+      "grad_norm": 3.864388812647437,
+      "learning_rate": 6.240198703718932e-06,
+      "loss": 3.2479,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6343129238612861,
+      "grad_norm": 3.5094157296404243,
+      "learning_rate": 6.234424614038375e-06,
+      "loss": 3.0269,
+      "step": 3297
+    },
+    {
+      "epoch": 0.634505314799673,
+      "grad_norm": 3.587076144296471,
+      "learning_rate": 6.2286519868639095e-06,
+      "loss": 3.0513,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6346977057380597,
+      "grad_norm": 3.446108474919643,
+      "learning_rate": 6.222880824437549e-06,
+      "loss": 2.9804,
+      "step": 3299
+    },
+    {
+      "epoch": 0.6348900966764466,
+      "grad_norm": 3.534923957109299,
+      "learning_rate": 6.217111129000759e-06,
+      "loss": 2.8775,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6350824876148333,
+      "grad_norm": 3.529528972879504,
+      "learning_rate": 6.211342902794413e-06,
+      "loss": 3.1834,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6352748785532202,
+      "grad_norm": 3.497028323556413,
+      "learning_rate": 6.205576148058828e-06,
+      "loss": 3.0679,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6354672694916069,
+      "grad_norm": 3.5556496400829665,
+      "learning_rate": 6.199810867033745e-06,
+      "loss": 3.2325,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6356596604299938,
+      "grad_norm": 3.341559408994591,
+      "learning_rate": 6.19404706195833e-06,
+      "loss": 3.1902,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6358520513683805,
+      "grad_norm": 3.415387832817926,
+      "learning_rate": 6.188284735071177e-06,
+      "loss": 2.8899,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6360444423067674,
+      "grad_norm": 3.389553205548809,
+      "learning_rate": 6.182523888610316e-06,
+      "loss": 3.2365,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6362368332451541,
+      "grad_norm": 3.6545893244411864,
+      "learning_rate": 6.176764524813187e-06,
+      "loss": 3.0898,
+      "step": 3307
+    },
+    {
+      "epoch": 0.636429224183541,
+      "grad_norm": 3.48689121400775,
+      "learning_rate": 6.171006645916662e-06,
+      "loss": 2.9151,
+      "step": 3308
+    },
+    {
+      "epoch": 0.6366216151219277,
+      "grad_norm": 3.327132120820244,
+      "learning_rate": 6.165250254157032e-06,
+      "loss": 2.9478,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6368140060603146,
+      "grad_norm": 3.5656367623635274,
+      "learning_rate": 6.159495351770017e-06,
+      "loss": 2.9919,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6370063969987013,
+      "grad_norm": 3.46040254337264,
+      "learning_rate": 6.153741940990749e-06,
+      "loss": 3.0908,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6371987879370882,
+      "grad_norm": 3.6613348367701106,
+      "learning_rate": 6.1479900240537956e-06,
+      "loss": 2.9459,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6373911788754749,
+      "grad_norm": 3.3347060956854304,
+      "learning_rate": 6.142239603193128e-06,
+      "loss": 2.7269,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6375835698138618,
+      "grad_norm": 3.4386915952252917,
+      "learning_rate": 6.136490680642146e-06,
+      "loss": 3.2211,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6377759607522485,
+      "grad_norm": 3.6092606606035895,
+      "learning_rate": 6.130743258633667e-06,
+      "loss": 3.0417,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6379683516906354,
+      "grad_norm": 3.5113984236577753,
+      "learning_rate": 6.124997339399916e-06,
+      "loss": 3.0419,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6381607426290222,
+      "grad_norm": 3.5211636066135883,
+      "learning_rate": 6.119252925172549e-06,
+      "loss": 2.8251,
+      "step": 3317
+    },
+    {
+      "epoch": 0.638353133567409,
+      "grad_norm": 3.6457443416444866,
+      "learning_rate": 6.113510018182628e-06,
+      "loss": 2.8272,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6385455245057958,
+      "grad_norm": 3.4518225362223434,
+      "learning_rate": 6.107768620660633e-06,
+      "loss": 2.814,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6387379154441826,
+      "grad_norm": 3.0404730698195905,
+      "learning_rate": 6.102028734836456e-06,
+      "loss": 2.6028,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6389303063825694,
+      "grad_norm": 2.4182495098141885,
+      "learning_rate": 6.0962903629394e-06,
+      "loss": 2.5578,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6391226973209562,
+      "grad_norm": 3.3348317717463134,
+      "learning_rate": 6.090553507198187e-06,
+      "loss": 2.7613,
+      "step": 3322
+    },
+    {
+      "epoch": 0.639315088259343,
+      "grad_norm": 3.151632862532061,
+      "learning_rate": 6.0848181698409384e-06,
+      "loss": 3.047,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6395074791977298,
+      "grad_norm": 3.4320177455379905,
+      "learning_rate": 6.079084353095202e-06,
+      "loss": 3.2892,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6396998701361166,
+      "grad_norm": 3.4968597854173584,
+      "learning_rate": 6.07335205918792e-06,
+      "loss": 3.0532,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6398922610745034,
+      "grad_norm": 3.7005239596325294,
+      "learning_rate": 6.067621290345455e-06,
+      "loss": 3.1065,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6400846520128902,
+      "grad_norm": 3.497518202388355,
+      "learning_rate": 6.061892048793568e-06,
+      "loss": 3.0881,
+      "step": 3327
+    },
+    {
+      "epoch": 0.640277042951277,
+      "grad_norm": 3.2208494396261584,
+      "learning_rate": 6.056164336757426e-06,
+      "loss": 2.787,
+      "step": 3328
+    },
+    {
+      "epoch": 0.6404694338896638,
+      "grad_norm": 3.6345453182123233,
+      "learning_rate": 6.050438156461613e-06,
+      "loss": 3.13,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6406618248280506,
+      "grad_norm": 3.480103481683372,
+      "learning_rate": 6.044713510130108e-06,
+      "loss": 2.9192,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6408542157664374,
+      "grad_norm": 3.325394237315888,
+      "learning_rate": 6.038990399986302e-06,
+      "loss": 3.0629,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6410466067048242,
+      "grad_norm": 3.4906026358337603,
+      "learning_rate": 6.03326882825298e-06,
+      "loss": 2.9803,
+      "step": 3332
+    },
+    {
+      "epoch": 0.641238997643211,
+      "grad_norm": 3.616353126309492,
+      "learning_rate": 6.027548797152336e-06,
+      "loss": 3.0838,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6414313885815978,
+      "grad_norm": 3.511665461137589,
+      "learning_rate": 6.021830308905963e-06,
+      "loss": 3.0852,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6416237795199846,
+      "grad_norm": 3.868358052246518,
+      "learning_rate": 6.016113365734861e-06,
+      "loss": 3.1601,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6418161704583714,
+      "grad_norm": 3.376843992367136,
+      "learning_rate": 6.0103979698594215e-06,
+      "loss": 2.7089,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6420085613967582,
+      "grad_norm": 3.532235539046074,
+      "learning_rate": 6.004684123499436e-06,
+      "loss": 3.2784,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6422009523351451,
+      "grad_norm": 3.5930587257370776,
+      "learning_rate": 5.998971828874102e-06,
+      "loss": 3.1909,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6423933432735318,
+      "grad_norm": 2.511536589848054,
+      "learning_rate": 5.993261088202005e-06,
+      "loss": 2.7442,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6425857342119187,
+      "grad_norm": 3.2804460579583057,
+      "learning_rate": 5.987551903701128e-06,
+      "loss": 3.2745,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6427781251503054,
+      "grad_norm": 3.308436429305467,
+      "learning_rate": 5.9818442775888595e-06,
+      "loss": 3.0469,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6429705160886923,
+      "grad_norm": 3.69978239950363,
+      "learning_rate": 5.97613821208197e-06,
+      "loss": 3.0228,
+      "step": 3342
+    },
+    {
+      "epoch": 0.643162907027079,
+      "grad_norm": 3.531545723855125,
+      "learning_rate": 5.970433709396635e-06,
+      "loss": 3.0231,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6433552979654659,
+      "grad_norm": 3.4388978714660525,
+      "learning_rate": 5.964730771748415e-06,
+      "loss": 3.2148,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6435476889038526,
+      "grad_norm": 3.6615188680857367,
+      "learning_rate": 5.959029401352262e-06,
+      "loss": 2.9799,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6437400798422395,
+      "grad_norm": 3.7882955177384234,
+      "learning_rate": 5.953329600422524e-06,
+      "loss": 3.0809,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6439324707806262,
+      "grad_norm": 3.5295530513640485,
+      "learning_rate": 5.947631371172943e-06,
+      "loss": 3.0862,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6441248617190131,
+      "grad_norm": 3.6985309980408267,
+      "learning_rate": 5.941934715816642e-06,
+      "loss": 3.3246,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6443172526573998,
+      "grad_norm": 3.346617181084521,
+      "learning_rate": 5.936239636566137e-06,
+      "loss": 3.1234,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6445096435957867,
+      "grad_norm": 3.266996999661194,
+      "learning_rate": 5.930546135633327e-06,
+      "loss": 2.8529,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6447020345341734,
+      "grad_norm": 3.3333812248461623,
+      "learning_rate": 5.924854215229509e-06,
+      "loss": 3.073,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6448944254725603,
+      "grad_norm": 3.221539690622712,
+      "learning_rate": 5.919163877565351e-06,
+      "loss": 2.8851,
+      "step": 3352
+    },
+    {
+      "epoch": 0.645086816410947,
+      "grad_norm": 3.3745101654627097,
+      "learning_rate": 5.9134751248509236e-06,
+      "loss": 2.8808,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6452792073493339,
+      "grad_norm": 3.424933615827668,
+      "learning_rate": 5.9077879592956675e-06,
+      "loss": 3.2329,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6454715982877206,
+      "grad_norm": 3.27784870489329,
+      "learning_rate": 5.902102383108415e-06,
+      "loss": 2.9825,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6456639892261075,
+      "grad_norm": 3.507428615851996,
+      "learning_rate": 5.896418398497377e-06,
+      "loss": 3.0109,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6458563801644942,
+      "grad_norm": 3.2397270021061337,
+      "learning_rate": 5.890736007670144e-06,
+      "loss": 2.9444,
+      "step": 3357
+    },
+    {
+      "epoch": 0.6460487711028811,
+      "grad_norm": 3.3150669748214443,
+      "learning_rate": 5.885055212833696e-06,
+      "loss": 3.0372,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6462411620412678,
+      "grad_norm": 3.3832594721673797,
+      "learning_rate": 5.879376016194387e-06,
+      "loss": 2.9068,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6464335529796547,
+      "grad_norm": 3.614203076941966,
+      "learning_rate": 5.873698419957952e-06,
+      "loss": 3.1943,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6466259439180415,
+      "grad_norm": 3.3329719477469646,
+      "learning_rate": 5.8680224263295045e-06,
+      "loss": 2.9726,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6468183348564283,
+      "grad_norm": 3.6270602481916954,
+      "learning_rate": 5.862348037513533e-06,
+      "loss": 2.8449,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6470107257948151,
+      "grad_norm": 3.6749397683272904,
+      "learning_rate": 5.856675255713905e-06,
+      "loss": 2.9674,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6472031167332019,
+      "grad_norm": 3.58717014467732,
+      "learning_rate": 5.851004083133862e-06,
+      "loss": 2.9624,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6473955076715887,
+      "grad_norm": 3.6367948994176467,
+      "learning_rate": 5.8453345219760275e-06,
+      "loss": 3.0499,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6475878986099755,
+      "grad_norm": 3.947387140920793,
+      "learning_rate": 5.839666574442389e-06,
+      "loss": 3.0218,
+      "step": 3366
+    },
+    {
+      "epoch": 0.6477802895483623,
+      "grad_norm": 3.800402146068601,
+      "learning_rate": 5.834000242734317e-06,
+      "loss": 2.9218,
+      "step": 3367
+    },
+    {
+      "epoch": 0.647972680486749,
+      "grad_norm": 3.2587011154902212,
+      "learning_rate": 5.828335529052541e-06,
+      "loss": 2.9943,
+      "step": 3368
+    },
+    {
+      "epoch": 0.6481650714251359,
+      "grad_norm": 3.303844762181033,
+      "learning_rate": 5.822672435597172e-06,
+      "loss": 3.1923,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6483574623635227,
+      "grad_norm": 3.715896053564414,
+      "learning_rate": 5.817010964567702e-06,
+      "loss": 3.1679,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6485498533019095,
+      "grad_norm": 3.301743535771389,
+      "learning_rate": 5.811351118162969e-06,
+      "loss": 2.8502,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6487422442402963,
+      "grad_norm": 3.885873028453468,
+      "learning_rate": 5.805692898581196e-06,
+      "loss": 2.9801,
+      "step": 3372
+    },
+    {
+      "epoch": 0.6489346351786831,
+      "grad_norm": 3.198160287131093,
+      "learning_rate": 5.800036308019974e-06,
+      "loss": 3.1674,
+      "step": 3373
+    },
+    {
+      "epoch": 0.6491270261170698,
+      "grad_norm": 3.528398581262105,
+      "learning_rate": 5.79438134867625e-06,
+      "loss": 3.2577,
+      "step": 3374
+    },
+    {
+      "epoch": 0.6493194170554567,
+      "grad_norm": 3.3596960757297842,
+      "learning_rate": 5.788728022746348e-06,
+      "loss": 2.7552,
+      "step": 3375
+    },
+    {
+      "epoch": 0.6495118079938434,
+      "grad_norm": 3.4914925496672136,
+      "learning_rate": 5.783076332425957e-06,
+      "loss": 2.997,
+      "step": 3376
+    },
+    {
+      "epoch": 0.6497041989322303,
+      "grad_norm": 3.3827392554891316,
+      "learning_rate": 5.777426279910125e-06,
+      "loss": 2.9945,
+      "step": 3377
+    },
+    {
+      "epoch": 0.649896589870617,
+      "grad_norm": 3.4176398194691844,
+      "learning_rate": 5.771777867393275e-06,
+      "loss": 3.0152,
+      "step": 3378
+    },
+    {
+      "epoch": 0.6500889808090039,
+      "grad_norm": 3.622047547290269,
+      "learning_rate": 5.766131097069174e-06,
+      "loss": 3.0316,
+      "step": 3379
+    },
+    {
+      "epoch": 0.6502813717473906,
+      "grad_norm": 3.8546238447242445,
+      "learning_rate": 5.760485971130969e-06,
+      "loss": 2.9004,
+      "step": 3380
+    },
+    {
+      "epoch": 0.6504737626857775,
+      "grad_norm": 3.394176862089281,
+      "learning_rate": 5.7548424917711596e-06,
+      "loss": 2.9942,
+      "step": 3381
+    },
+    {
+      "epoch": 0.6506661536241644,
+      "grad_norm": 3.6910249861766737,
+      "learning_rate": 5.749200661181611e-06,
+      "loss": 2.7658,
+      "step": 3382
+    },
+    {
+      "epoch": 0.6508585445625511,
+      "grad_norm": 3.249695389644881,
+      "learning_rate": 5.7435604815535475e-06,
+      "loss": 3.1391,
+      "step": 3383
+    },
+    {
+      "epoch": 0.651050935500938,
+      "grad_norm": 3.3365820243222712,
+      "learning_rate": 5.7379219550775415e-06,
+      "loss": 2.8668,
+      "step": 3384
+    },
+    {
+      "epoch": 0.6512433264393247,
+      "grad_norm": 3.599026601674678,
+      "learning_rate": 5.732285083943537e-06,
+      "loss": 2.8203,
+      "step": 3385
+    },
+    {
+      "epoch": 0.6514357173777116,
+      "grad_norm": 3.3528513533834516,
+      "learning_rate": 5.726649870340833e-06,
+      "loss": 3.0397,
+      "step": 3386
+    },
+    {
+      "epoch": 0.6516281083160983,
+      "grad_norm": 3.6862655297473292,
+      "learning_rate": 5.721016316458068e-06,
+      "loss": 3.1602,
+      "step": 3387
+    },
+    {
+      "epoch": 0.6518204992544852,
+      "grad_norm": 3.6270350731118226,
+      "learning_rate": 5.715384424483268e-06,
+      "loss": 3.0354,
+      "step": 3388
+    },
+    {
+      "epoch": 0.6520128901928719,
+      "grad_norm": 3.3502963124284286,
+      "learning_rate": 5.709754196603781e-06,
+      "loss": 2.916,
+      "step": 3389
+    },
+    {
+      "epoch": 0.6522052811312588,
+      "grad_norm": 3.2931999532290845,
+      "learning_rate": 5.704125635006329e-06,
+      "loss": 2.9761,
+      "step": 3390
+    },
+    {
+      "epoch": 0.6523976720696455,
+      "grad_norm": 3.5414677742777636,
+      "learning_rate": 5.6984987418769825e-06,
+      "loss": 3.2096,
+      "step": 3391
+    },
+    {
+      "epoch": 0.6525900630080324,
+      "grad_norm": 3.5341905195123475,
+      "learning_rate": 5.692873519401154e-06,
+      "loss": 2.6726,
+      "step": 3392
+    },
+    {
+      "epoch": 0.6527824539464191,
+      "grad_norm": 3.240768523133126,
+      "learning_rate": 5.6872499697636195e-06,
+      "loss": 2.9749,
+      "step": 3393
+    },
+    {
+      "epoch": 0.652974844884806,
+      "grad_norm": 3.4948641841794816,
+      "learning_rate": 5.681628095148502e-06,
+      "loss": 2.8089,
+      "step": 3394
+    },
+    {
+      "epoch": 0.6531672358231927,
+      "grad_norm": 3.288937878764572,
+      "learning_rate": 5.6760078977392706e-06,
+      "loss": 3.0068,
+      "step": 3395
+    },
+    {
+      "epoch": 0.6533596267615795,
+      "grad_norm": 3.8888568299538857,
+      "learning_rate": 5.67038937971875e-06,
+      "loss": 3.0517,
+      "step": 3396
+    },
+    {
+      "epoch": 0.6535520176999663,
+      "grad_norm": 2.4149395660166837,
+      "learning_rate": 5.664772543269101e-06,
+      "loss": 2.6849,
+      "step": 3397
+    },
+    {
+      "epoch": 0.6537444086383531,
+      "grad_norm": 3.4396799454903295,
+      "learning_rate": 5.659157390571842e-06,
+      "loss": 2.9334,
+      "step": 3398
+    },
+    {
+      "epoch": 0.6539367995767399,
+      "grad_norm": 3.607697731523285,
+      "learning_rate": 5.653543923807833e-06,
+      "loss": 3.0309,
+      "step": 3399
+    },
+    {
+      "epoch": 0.6541291905151267,
+      "grad_norm": 3.5058081621823476,
+      "learning_rate": 5.6479321451572785e-06,
+      "loss": 3.0273,
+      "step": 3400
+    },
+    {
+      "epoch": 0.6543215814535135,
+      "grad_norm": 3.7160568240123406,
+      "learning_rate": 5.642322056799732e-06,
+      "loss": 2.786,
+      "step": 3401
+    },
+    {
+      "epoch": 0.6545139723919003,
+      "grad_norm": 3.482818034566111,
+      "learning_rate": 5.636713660914087e-06,
+      "loss": 3.3139,
+      "step": 3402
+    },
+    {
+      "epoch": 0.6547063633302871,
+      "grad_norm": 3.5330678536233866,
+      "learning_rate": 5.631106959678575e-06,
+      "loss": 2.8959,
+      "step": 3403
+    },
+    {
+      "epoch": 0.6548987542686739,
+      "grad_norm": 3.578490492953155,
+      "learning_rate": 5.625501955270777e-06,
+      "loss": 2.8305,
+      "step": 3404
+    },
+    {
+      "epoch": 0.6550911452070608,
+      "grad_norm": 3.4738751131722245,
+      "learning_rate": 5.619898649867612e-06,
+      "loss": 3.1104,
+      "step": 3405
+    },
+    {
+      "epoch": 0.6552835361454475,
+      "grad_norm": 3.5869505750930184,
+      "learning_rate": 5.614297045645339e-06,
+      "loss": 3.1083,
+      "step": 3406
+    },
+    {
+      "epoch": 0.6554759270838344,
+      "grad_norm": 3.6446019714196156,
+      "learning_rate": 5.6086971447795625e-06,
+      "loss": 3.0295,
+      "step": 3407
+    },
+    {
+      "epoch": 0.6556683180222211,
+      "grad_norm": 3.452666156982138,
+      "learning_rate": 5.603098949445209e-06,
+      "loss": 3.1747,
+      "step": 3408
+    },
+    {
+      "epoch": 0.655860708960608,
+      "grad_norm": 3.3606872466385975,
+      "learning_rate": 5.597502461816557e-06,
+      "loss": 2.7856,
+      "step": 3409
+    },
+    {
+      "epoch": 0.6560530998989947,
+      "grad_norm": 3.629344029306742,
+      "learning_rate": 5.5919076840672215e-06,
+      "loss": 2.9998,
+      "step": 3410
+    },
+    {
+      "epoch": 0.6562454908373816,
+      "grad_norm": 3.358827603417466,
+      "learning_rate": 5.5863146183701454e-06,
+      "loss": 2.7358,
+      "step": 3411
+    },
+    {
+      "epoch": 0.6564378817757683,
+      "grad_norm": 3.557564224393113,
+      "learning_rate": 5.580723266897616e-06,
+      "loss": 2.951,
+      "step": 3412
+    },
+    {
+      "epoch": 0.6566302727141552,
+      "grad_norm": 3.574489218033161,
+      "learning_rate": 5.575133631821243e-06,
+      "loss": 3.0981,
+      "step": 3413
+    },
+    {
+      "epoch": 0.6568226636525419,
+      "grad_norm": 3.420915012800928,
+      "learning_rate": 5.5695457153119806e-06,
+      "loss": 3.1744,
+      "step": 3414
+    },
+    {
+      "epoch": 0.6570150545909288,
+      "grad_norm": 3.4978254284714003,
+      "learning_rate": 5.563959519540114e-06,
+      "loss": 3.1758,
+      "step": 3415
+    },
+    {
+      "epoch": 0.6572074455293155,
+      "grad_norm": 3.2029900772177298,
+      "learning_rate": 5.558375046675244e-06,
+      "loss": 2.934,
+      "step": 3416
+    },
+    {
+      "epoch": 0.6573998364677024,
+      "grad_norm": 3.5118902325321453,
+      "learning_rate": 5.552792298886335e-06,
+      "loss": 3.2516,
+      "step": 3417
+    },
+    {
+      "epoch": 0.6575922274060891,
+      "grad_norm": 3.6177336039944974,
+      "learning_rate": 5.547211278341646e-06,
+      "loss": 3.0205,
+      "step": 3418
+    },
+    {
+      "epoch": 0.657784618344476,
+      "grad_norm": 3.3438654359107782,
+      "learning_rate": 5.541631987208789e-06,
+      "loss": 2.8662,
+      "step": 3419
+    },
+    {
+      "epoch": 0.6579770092828627,
+      "grad_norm": 3.5538179735820274,
+      "learning_rate": 5.536054427654698e-06,
+      "loss": 2.8043,
+      "step": 3420
+    },
+    {
+      "epoch": 0.6581694002212496,
+      "grad_norm": 3.460568309925604,
+      "learning_rate": 5.530478601845624e-06,
+      "loss": 3.0179,
+      "step": 3421
+    },
+    {
+      "epoch": 0.6583617911596363,
+      "grad_norm": 3.515638956445024,
+      "learning_rate": 5.52490451194716e-06,
+      "loss": 3.1757,
+      "step": 3422
+    },
+    {
+      "epoch": 0.6585541820980232,
+      "grad_norm": 3.292934869327124,
+      "learning_rate": 5.519332160124215e-06,
+      "loss": 2.8188,
+      "step": 3423
+    },
+    {
+      "epoch": 0.6587465730364099,
+      "grad_norm": 3.6027860920973795,
+      "learning_rate": 5.513761548541032e-06,
+      "loss": 2.7926,
+      "step": 3424
+    },
+    {
+      "epoch": 0.6589389639747968,
+      "grad_norm": 3.5269026068590885,
+      "learning_rate": 5.508192679361169e-06,
+      "loss": 2.8311,
+      "step": 3425
+    },
+    {
+      "epoch": 0.6591313549131836,
+      "grad_norm": 3.5535083183340435,
+      "learning_rate": 5.502625554747508e-06,
+      "loss": 3.3569,
+      "step": 3426
+    },
+    {
+      "epoch": 0.6593237458515704,
+      "grad_norm": 3.532428728017737,
+      "learning_rate": 5.497060176862259e-06,
+      "loss": 3.0067,
+      "step": 3427
+    },
+    {
+      "epoch": 0.6595161367899572,
+      "grad_norm": 3.366561992365177,
+      "learning_rate": 5.491496547866948e-06,
+      "loss": 2.9507,
+      "step": 3428
+    },
+    {
+      "epoch": 0.659708527728344,
+      "grad_norm": 3.424960392702373,
+      "learning_rate": 5.485934669922428e-06,
+      "loss": 3.0183,
+      "step": 3429
+    },
+    {
+      "epoch": 0.6599009186667308,
+      "grad_norm": 3.818757324731149,
+      "learning_rate": 5.480374545188866e-06,
+      "loss": 3.2019,
+      "step": 3430
+    },
+    {
+      "epoch": 0.6600933096051176,
+      "grad_norm": 3.5098728402587254,
+      "learning_rate": 5.474816175825754e-06,
+      "loss": 3.0021,
+      "step": 3431
+    },
+    {
+      "epoch": 0.6602857005435044,
+      "grad_norm": 3.477021759542291,
+      "learning_rate": 5.469259563991894e-06,
+      "loss": 3.0251,
+      "step": 3432
+    },
+    {
+      "epoch": 0.6604780914818912,
+      "grad_norm": 3.666881513103784,
+      "learning_rate": 5.46370471184541e-06,
+      "loss": 2.8683,
+      "step": 3433
+    },
+    {
+      "epoch": 0.660670482420278,
+      "grad_norm": 3.2752167299919814,
+      "learning_rate": 5.458151621543744e-06,
+      "loss": 3.1364,
+      "step": 3434
+    },
+    {
+      "epoch": 0.6608628733586648,
+      "grad_norm": 3.4068019923422606,
+      "learning_rate": 5.452600295243653e-06,
+      "loss": 3.0055,
+      "step": 3435
+    },
+    {
+      "epoch": 0.6610552642970516,
+      "grad_norm": 3.3452487559166815,
+      "learning_rate": 5.4470507351012116e-06,
+      "loss": 2.6157,
+      "step": 3436
+    },
+    {
+      "epoch": 0.6612476552354384,
+      "grad_norm": 3.3668848873829944,
+      "learning_rate": 5.441502943271797e-06,
+      "loss": 2.9028,
+      "step": 3437
+    },
+    {
+      "epoch": 0.6614400461738252,
+      "grad_norm": 3.448794818995321,
+      "learning_rate": 5.4359569219101115e-06,
+      "loss": 3.1519,
+      "step": 3438
+    },
+    {
+      "epoch": 0.661632437112212,
+      "grad_norm": 3.412011058564034,
+      "learning_rate": 5.430412673170167e-06,
+      "loss": 2.7159,
+      "step": 3439
+    },
+    {
+      "epoch": 0.6618248280505988,
+      "grad_norm": 3.3241013012600007,
+      "learning_rate": 5.424870199205283e-06,
+      "loss": 3.1405,
+      "step": 3440
+    },
+    {
+      "epoch": 0.6620172189889856,
+      "grad_norm": 3.4812555525800764,
+      "learning_rate": 5.4193295021681e-06,
+      "loss": 3.1788,
+      "step": 3441
+    },
+    {
+      "epoch": 0.6622096099273724,
+      "grad_norm": 2.465576943540714,
+      "learning_rate": 5.413790584210551e-06,
+      "loss": 2.7129,
+      "step": 3442
+    },
+    {
+      "epoch": 0.6624020008657592,
+      "grad_norm": 3.23505964876115,
+      "learning_rate": 5.408253447483892e-06,
+      "loss": 3.0157,
+      "step": 3443
+    },
+    {
+      "epoch": 0.662594391804146,
+      "grad_norm": 3.624815091508012,
+      "learning_rate": 5.402718094138688e-06,
+      "loss": 3.0138,
+      "step": 3444
+    },
+    {
+      "epoch": 0.6627867827425328,
+      "grad_norm": 3.6201444899031605,
+      "learning_rate": 5.397184526324792e-06,
+      "loss": 3.1351,
+      "step": 3445
+    },
+    {
+      "epoch": 0.6629791736809196,
+      "grad_norm": 3.6139041091055244,
+      "learning_rate": 5.391652746191398e-06,
+      "loss": 3.0711,
+      "step": 3446
+    },
+    {
+      "epoch": 0.6631715646193065,
+      "grad_norm": 3.3287466861570083,
+      "learning_rate": 5.38612275588697e-06,
+      "loss": 3.0847,
+      "step": 3447
+    },
+    {
+      "epoch": 0.6633639555576932,
+      "grad_norm": 3.5973263711800363,
+      "learning_rate": 5.380594557559298e-06,
+      "loss": 3.0782,
+      "step": 3448
+    },
+    {
+      "epoch": 0.6635563464960801,
+      "grad_norm": 3.468745499895028,
+      "learning_rate": 5.375068153355474e-06,
+      "loss": 2.9715,
+      "step": 3449
+    },
+    {
+      "epoch": 0.6637487374344668,
+      "grad_norm": 3.526840541140222,
+      "learning_rate": 5.369543545421883e-06,
+      "loss": 2.9409,
+      "step": 3450
+    },
+    {
+      "epoch": 0.6639411283728537,
+      "grad_norm": 3.6427605884610865,
+      "learning_rate": 5.364020735904223e-06,
+      "loss": 2.9755,
+      "step": 3451
+    },
+    {
+      "epoch": 0.6641335193112404,
+      "grad_norm": 3.402384936203152,
+      "learning_rate": 5.358499726947488e-06,
+      "loss": 2.9788,
+      "step": 3452
+    },
+    {
+      "epoch": 0.6643259102496273,
+      "grad_norm": 3.457817379986368,
+      "learning_rate": 5.352980520695974e-06,
+      "loss": 3.0773,
+      "step": 3453
+    },
+    {
+      "epoch": 0.664518301188014,
+      "grad_norm": 3.4463501985839864,
+      "learning_rate": 5.347463119293283e-06,
+      "loss": 3.2325,
+      "step": 3454
+    },
+    {
+      "epoch": 0.6647106921264009,
+      "grad_norm": 3.556212985767232,
+      "learning_rate": 5.341947524882301e-06,
+      "loss": 3.188,
+      "step": 3455
+    },
+    {
+      "epoch": 0.6649030830647876,
+      "grad_norm": 3.7660380352813307,
+      "learning_rate": 5.336433739605227e-06,
+      "loss": 3.2678,
+      "step": 3456
+    },
+    {
+      "epoch": 0.6650954740031745,
+      "grad_norm": 3.5334879078898203,
+      "learning_rate": 5.330921765603549e-06,
+      "loss": 2.8187,
+      "step": 3457
+    },
+    {
+      "epoch": 0.6652878649415612,
+      "grad_norm": 3.101955866440567,
+      "learning_rate": 5.325411605018056e-06,
+      "loss": 2.7819,
+      "step": 3458
+    },
+    {
+      "epoch": 0.6654802558799481,
+      "grad_norm": 3.2624462691872393,
+      "learning_rate": 5.31990325998883e-06,
+      "loss": 3.1369,
+      "step": 3459
+    },
+    {
+      "epoch": 0.6656726468183348,
+      "grad_norm": 3.375873354682192,
+      "learning_rate": 5.314396732655253e-06,
+      "loss": 3.01,
+      "step": 3460
+    },
+    {
+      "epoch": 0.6658650377567217,
+      "grad_norm": 3.4257481623420882,
+      "learning_rate": 5.308892025155989e-06,
+      "loss": 3.1841,
+      "step": 3461
+    },
+    {
+      "epoch": 0.6660574286951084,
+      "grad_norm": 3.2415641020235992,
+      "learning_rate": 5.303389139629007e-06,
+      "loss": 2.5853,
+      "step": 3462
+    },
+    {
+      "epoch": 0.6662498196334953,
+      "grad_norm": 2.309700636608047,
+      "learning_rate": 5.297888078211564e-06,
+      "loss": 2.7381,
+      "step": 3463
+    },
+    {
+      "epoch": 0.666442210571882,
+      "grad_norm": 3.3156666449652246,
+      "learning_rate": 5.2923888430402085e-06,
+      "loss": 3.1466,
+      "step": 3464
+    },
+    {
+      "epoch": 0.6666346015102689,
+      "grad_norm": 3.7984049238815563,
+      "learning_rate": 5.286891436250785e-06,
+      "loss": 3.2748,
+      "step": 3465
+    },
+    {
+      "epoch": 0.6668269924486556,
+      "grad_norm": 3.4098200299831056,
+      "learning_rate": 5.281395859978414e-06,
+      "loss": 3.0326,
+      "step": 3466
+    },
+    {
+      "epoch": 0.6670193833870425,
+      "grad_norm": 3.4704954546702367,
+      "learning_rate": 5.2759021163575184e-06,
+      "loss": 3.0519,
+      "step": 3467
+    },
+    {
+      "epoch": 0.6672117743254292,
+      "grad_norm": 3.3331449603711376,
+      "learning_rate": 5.27041020752181e-06,
+      "loss": 3.1028,
+      "step": 3468
+    },
+    {
+      "epoch": 0.6674041652638161,
+      "grad_norm": 3.3269489728426698,
+      "learning_rate": 5.26492013560427e-06,
+      "loss": 2.9086,
+      "step": 3469
+    },
+    {
+      "epoch": 0.6675965562022029,
+      "grad_norm": 2.2245256234317465,
+      "learning_rate": 5.259431902737195e-06,
+      "loss": 2.56,
+      "step": 3470
+    },
+    {
+      "epoch": 0.6677889471405897,
+      "grad_norm": 3.503274655980237,
+      "learning_rate": 5.2539455110521385e-06,
+      "loss": 3.0295,
+      "step": 3471
+    },
+    {
+      "epoch": 0.6679813380789765,
+      "grad_norm": 3.5388330697903205,
+      "learning_rate": 5.248460962679958e-06,
+      "loss": 3.1675,
+      "step": 3472
+    },
+    {
+      "epoch": 0.6681737290173633,
+      "grad_norm": 3.0963454766245,
+      "learning_rate": 5.24297825975079e-06,
+      "loss": 2.9921,
+      "step": 3473
+    },
+    {
+      "epoch": 0.6683661199557501,
+      "grad_norm": 3.224316147146273,
+      "learning_rate": 5.237497404394044e-06,
+      "loss": 3.0515,
+      "step": 3474
+    },
+    {
+      "epoch": 0.6685585108941369,
+      "grad_norm": 3.299079382070122,
+      "learning_rate": 5.232018398738436e-06,
+      "loss": 3.1143,
+      "step": 3475
+    },
+    {
+      "epoch": 0.6687509018325237,
+      "grad_norm": 3.4967087229600797,
+      "learning_rate": 5.226541244911936e-06,
+      "loss": 3.1873,
+      "step": 3476
+    },
+    {
+      "epoch": 0.6689432927709105,
+      "grad_norm": 3.5927264295964685,
+      "learning_rate": 5.221065945041811e-06,
+      "loss": 2.831,
+      "step": 3477
+    },
+    {
+      "epoch": 0.6691356837092973,
+      "grad_norm": 3.4934447075504274,
+      "learning_rate": 5.215592501254609e-06,
+      "loss": 3.0509,
+      "step": 3478
+    },
+    {
+      "epoch": 0.6693280746476841,
+      "grad_norm": 3.2055845987930724,
+      "learning_rate": 5.210120915676147e-06,
+      "loss": 2.8451,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6695204655860709,
+      "grad_norm": 3.483304316565873,
+      "learning_rate": 5.2046511904315265e-06,
+      "loss": 3.0261,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6697128565244577,
+      "grad_norm": 3.115559802848036,
+      "learning_rate": 5.199183327645128e-06,
+      "loss": 2.9559,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6699052474628445,
+      "grad_norm": 3.297756033101178,
+      "learning_rate": 5.193717329440604e-06,
+      "loss": 2.9837,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6700976384012313,
+      "grad_norm": 3.4794624531151346,
+      "learning_rate": 5.188253197940889e-06,
+      "loss": 3.2285,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6702900293396181,
+      "grad_norm": 3.27611733292452,
+      "learning_rate": 5.182790935268185e-06,
+      "loss": 2.9544,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6704824202780049,
+      "grad_norm": 3.2781773164009773,
+      "learning_rate": 5.177330543543971e-06,
+      "loss": 3.0626,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6706748112163917,
+      "grad_norm": 3.5281823110715558,
+      "learning_rate": 5.171872024889004e-06,
+      "loss": 3.081,
+      "step": 3486
+    },
+    {
+      "epoch": 0.6708672021547785,
+      "grad_norm": 3.306206835153916,
+      "learning_rate": 5.166415381423306e-06,
+      "loss": 2.8434,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6710595930931653,
+      "grad_norm": 3.3710451947783224,
+      "learning_rate": 5.160960615266179e-06,
+      "loss": 3.1572,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6712519840315521,
+      "grad_norm": 3.1777158842244075,
+      "learning_rate": 5.155507728536191e-06,
+      "loss": 3.0383,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6714443749699389,
+      "grad_norm": 3.4515201137641083,
+      "learning_rate": 5.150056723351173e-06,
+      "loss": 3.0606,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6716367659083258,
+      "grad_norm": 3.729044786550854,
+      "learning_rate": 5.14460760182824e-06,
+      "loss": 3.0719,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6718291568467125,
+      "grad_norm": 3.6707572502781027,
+      "learning_rate": 5.139160366083765e-06,
+      "loss": 3.0386,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6720215477850994,
+      "grad_norm": 3.840358750760315,
+      "learning_rate": 5.133715018233393e-06,
+      "loss": 2.9821,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6722139387234861,
+      "grad_norm": 3.339018857051246,
+      "learning_rate": 5.128271560392037e-06,
+      "loss": 3.1358,
+      "step": 3494
+    },
+    {
+      "epoch": 0.672406329661873,
+      "grad_norm": 3.374594318973455,
+      "learning_rate": 5.122829994673866e-06,
+      "loss": 2.9448,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6725987206002597,
+      "grad_norm": 3.4815157367310787,
+      "learning_rate": 5.117390323192326e-06,
+      "loss": 3.0471,
+      "step": 3496
+    },
+    {
+      "epoch": 0.6727911115386466,
+      "grad_norm": 3.47058662649197,
+      "learning_rate": 5.111952548060126e-06,
+      "loss": 2.9339,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6729835024770333,
+      "grad_norm": 3.2880919078305664,
+      "learning_rate": 5.106516671389224e-06,
+      "loss": 2.8369,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6731758934154202,
+      "grad_norm": 3.621850648755694,
+      "learning_rate": 5.101082695290866e-06,
+      "loss": 3.0638,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6733682843538069,
+      "grad_norm": 3.256888709860925,
+      "learning_rate": 5.0956506218755344e-06,
+      "loss": 3.1121,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6735606752921938,
+      "grad_norm": 3.660577578563313,
+      "learning_rate": 5.09022045325299e-06,
+      "loss": 2.8057,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6737530662305805,
+      "grad_norm": 3.368716417277223,
+      "learning_rate": 5.0847921915322486e-06,
+      "loss": 2.9088,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6739454571689674,
+      "grad_norm": 3.5457373822678644,
+      "learning_rate": 5.07936583882158e-06,
+      "loss": 2.9406,
+      "step": 3503
+    },
+    {
+      "epoch": 0.6741378481073541,
+      "grad_norm": 3.5146291184187364,
+      "learning_rate": 5.073941397228518e-06,
+      "loss": 2.9362,
+      "step": 3504
+    },
+    {
+      "epoch": 0.674330239045741,
+      "grad_norm": 3.452581322598457,
+      "learning_rate": 5.068518868859854e-06,
+      "loss": 3.2815,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6745226299841277,
+      "grad_norm": 3.644346541719036,
+      "learning_rate": 5.063098255821637e-06,
+      "loss": 2.9056,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6747150209225146,
+      "grad_norm": 3.5133965610849893,
+      "learning_rate": 5.0576795602191734e-06,
+      "loss": 3.0816,
+      "step": 3507
+    },
+    {
+      "epoch": 0.6749074118609013,
+      "grad_norm": 3.597367536881128,
+      "learning_rate": 5.052262784157014e-06,
+      "loss": 3.0279,
+      "step": 3508
+    },
+    {
+      "epoch": 0.6750998027992882,
+      "grad_norm": 3.4295509506552673,
+      "learning_rate": 5.046847929738971e-06,
+      "loss": 2.9094,
+      "step": 3509
+    },
+    {
+      "epoch": 0.6752921937376749,
+      "grad_norm": 2.2615055466967764,
+      "learning_rate": 5.041434999068127e-06,
+      "loss": 2.6231,
+      "step": 3510
+    },
+    {
+      "epoch": 0.6754845846760618,
+      "grad_norm": 3.6195236098637587,
+      "learning_rate": 5.036023994246787e-06,
+      "loss": 3.0431,
+      "step": 3511
+    },
+    {
+      "epoch": 0.6756769756144485,
+      "grad_norm": 3.762368115498139,
+      "learning_rate": 5.030614917376532e-06,
+      "loss": 2.9815,
+      "step": 3512
+    },
+    {
+      "epoch": 0.6758693665528354,
+      "grad_norm": 3.462805396304912,
+      "learning_rate": 5.025207770558176e-06,
+      "loss": 2.8288,
+      "step": 3513
+    },
+    {
+      "epoch": 0.6760617574912222,
+      "grad_norm": 3.6346378672432467,
+      "learning_rate": 5.0198025558917985e-06,
+      "loss": 3.134,
+      "step": 3514
+    },
+    {
+      "epoch": 0.676254148429609,
+      "grad_norm": 3.4148306182176653,
+      "learning_rate": 5.014399275476721e-06,
+      "loss": 2.9646,
+      "step": 3515
+    },
+    {
+      "epoch": 0.6764465393679958,
+      "grad_norm": 3.4247543224055557,
+      "learning_rate": 5.008997931411517e-06,
+      "loss": 2.9789,
+      "step": 3516
+    },
+    {
+      "epoch": 0.6766389303063826,
+      "grad_norm": 3.356696858981768,
+      "learning_rate": 5.003598525794002e-06,
+      "loss": 2.8492,
+      "step": 3517
+    },
+    {
+      "epoch": 0.6768313212447694,
+      "grad_norm": 3.7564008382494802,
+      "learning_rate": 4.998201060721253e-06,
+      "loss": 3.064,
+      "step": 3518
+    },
+    {
+      "epoch": 0.6770237121831562,
+      "grad_norm": 3.65216466777436,
+      "learning_rate": 4.992805538289571e-06,
+      "loss": 3.0264,
+      "step": 3519
+    },
+    {
+      "epoch": 0.677216103121543,
+      "grad_norm": 3.864257796947702,
+      "learning_rate": 4.987411960594521e-06,
+      "loss": 3.226,
+      "step": 3520
+    },
+    {
+      "epoch": 0.6774084940599298,
+      "grad_norm": 3.4957484510953267,
+      "learning_rate": 4.982020329730904e-06,
+      "loss": 2.8062,
+      "step": 3521
+    },
+    {
+      "epoch": 0.6776008849983166,
+      "grad_norm": 3.477318804798542,
+      "learning_rate": 4.976630647792771e-06,
+      "loss": 2.951,
+      "step": 3522
+    },
+    {
+      "epoch": 0.6777932759367034,
+      "grad_norm": 3.6294678518174663,
+      "learning_rate": 4.971242916873412e-06,
+      "loss": 2.9788,
+      "step": 3523
+    },
+    {
+      "epoch": 0.6779856668750902,
+      "grad_norm": 3.267878909203834,
+      "learning_rate": 4.965857139065354e-06,
+      "loss": 2.7558,
+      "step": 3524
+    },
+    {
+      "epoch": 0.678178057813477,
+      "grad_norm": 3.6213825386512344,
+      "learning_rate": 4.9604733164603755e-06,
+      "loss": 3.1389,
+      "step": 3525
+    },
+    {
+      "epoch": 0.6783704487518638,
+      "grad_norm": 3.3201119071660954,
+      "learning_rate": 4.955091451149495e-06,
+      "loss": 2.7608,
+      "step": 3526
+    },
+    {
+      "epoch": 0.6785628396902506,
+      "grad_norm": 3.3305295916811035,
+      "learning_rate": 4.9497115452229535e-06,
+      "loss": 2.8253,
+      "step": 3527
+    },
+    {
+      "epoch": 0.6787552306286374,
+      "grad_norm": 3.9015117394583383,
+      "learning_rate": 4.9443336007702614e-06,
+      "loss": 3.2,
+      "step": 3528
+    },
+    {
+      "epoch": 0.6789476215670242,
+      "grad_norm": 2.2850975969952283,
+      "learning_rate": 4.938957619880138e-06,
+      "loss": 2.55,
+      "step": 3529
+    },
+    {
+      "epoch": 0.679140012505411,
+      "grad_norm": 3.4679875295737386,
+      "learning_rate": 4.9335836046405575e-06,
+      "loss": 3.0666,
+      "step": 3530
+    },
+    {
+      "epoch": 0.6793324034437977,
+      "grad_norm": 3.782806367875205,
+      "learning_rate": 4.928211557138728e-06,
+      "loss": 2.9054,
+      "step": 3531
+    },
+    {
+      "epoch": 0.6795247943821846,
+      "grad_norm": 2.3218720116394107,
+      "learning_rate": 4.922841479461083e-06,
+      "loss": 2.537,
+      "step": 3532
+    },
+    {
+      "epoch": 0.6797171853205713,
+      "grad_norm": 3.771977834722463,
+      "learning_rate": 4.917473373693305e-06,
+      "loss": 3.0048,
+      "step": 3533
+    },
+    {
+      "epoch": 0.6799095762589582,
+      "grad_norm": 3.5440074250363534,
+      "learning_rate": 4.9121072419203016e-06,
+      "loss": 3.0269,
+      "step": 3534
+    },
+    {
+      "epoch": 0.6801019671973451,
+      "grad_norm": 3.566982502117534,
+      "learning_rate": 4.906743086226218e-06,
+      "loss": 3.2178,
+      "step": 3535
+    },
+    {
+      "epoch": 0.6802943581357318,
+      "grad_norm": 3.6156504561035385,
+      "learning_rate": 4.901380908694434e-06,
+      "loss": 2.8648,
+      "step": 3536
+    },
+    {
+      "epoch": 0.6804867490741187,
+      "grad_norm": 3.5130886932347978,
+      "learning_rate": 4.8960207114075495e-06,
+      "loss": 3.2068,
+      "step": 3537
+    },
+    {
+      "epoch": 0.6806791400125054,
+      "grad_norm": 3.584013660975283,
+      "learning_rate": 4.890662496447407e-06,
+      "loss": 2.9038,
+      "step": 3538
+    },
+    {
+      "epoch": 0.6808715309508923,
+      "grad_norm": 3.2566690033653773,
+      "learning_rate": 4.8853062658950765e-06,
+      "loss": 3.2058,
+      "step": 3539
+    },
+    {
+      "epoch": 0.681063921889279,
+      "grad_norm": 3.697356688078167,
+      "learning_rate": 4.879952021830856e-06,
+      "loss": 2.9705,
+      "step": 3540
+    },
+    {
+      "epoch": 0.6812563128276659,
+      "grad_norm": 3.278465622001884,
+      "learning_rate": 4.874599766334276e-06,
+      "loss": 2.8644,
+      "step": 3541
+    },
+    {
+      "epoch": 0.6814487037660526,
+      "grad_norm": 3.460020280428904,
+      "learning_rate": 4.8692495014840825e-06,
+      "loss": 3.1581,
+      "step": 3542
+    },
+    {
+      "epoch": 0.6816410947044395,
+      "grad_norm": 3.3099515236772508,
+      "learning_rate": 4.863901229358261e-06,
+      "loss": 2.8983,
+      "step": 3543
+    },
+    {
+      "epoch": 0.6818334856428262,
+      "grad_norm": 3.3965793017141377,
+      "learning_rate": 4.858554952034019e-06,
+      "loss": 3.2494,
+      "step": 3544
+    },
+    {
+      "epoch": 0.682025876581213,
+      "grad_norm": 3.3968299287051,
+      "learning_rate": 4.853210671587789e-06,
+      "loss": 2.9608,
+      "step": 3545
+    },
+    {
+      "epoch": 0.6822182675195998,
+      "grad_norm": 3.245271657593338,
+      "learning_rate": 4.847868390095227e-06,
+      "loss": 2.9307,
+      "step": 3546
+    },
+    {
+      "epoch": 0.6824106584579867,
+      "grad_norm": 3.604328002289091,
+      "learning_rate": 4.842528109631218e-06,
+      "loss": 3.0496,
+      "step": 3547
+    },
+    {
+      "epoch": 0.6826030493963734,
+      "grad_norm": 2.5539332134084547,
+      "learning_rate": 4.837189832269858e-06,
+      "loss": 2.5311,
+      "step": 3548
+    },
+    {
+      "epoch": 0.6827954403347603,
+      "grad_norm": 3.305643759668087,
+      "learning_rate": 4.8318535600844775e-06,
+      "loss": 2.9903,
+      "step": 3549
+    },
+    {
+      "epoch": 0.682987831273147,
+      "grad_norm": 3.5705938675742868,
+      "learning_rate": 4.8265192951476206e-06,
+      "loss": 3.0448,
+      "step": 3550
+    },
+    {
+      "epoch": 0.6831802222115338,
+      "grad_norm": 3.2873782487793166,
+      "learning_rate": 4.8211870395310556e-06,
+      "loss": 2.9009,
+      "step": 3551
+    },
+    {
+      "epoch": 0.6833726131499206,
+      "grad_norm": 3.264421046427747,
+      "learning_rate": 4.815856795305772e-06,
+      "loss": 2.9428,
+      "step": 3552
+    },
+    {
+      "epoch": 0.6835650040883074,
+      "grad_norm": 3.3458621499153414,
+      "learning_rate": 4.81052856454197e-06,
+      "loss": 3.2174,
+      "step": 3553
+    },
+    {
+      "epoch": 0.6837573950266942,
+      "grad_norm": 3.5338028080046016,
+      "learning_rate": 4.805202349309074e-06,
+      "loss": 3.2382,
+      "step": 3554
+    },
+    {
+      "epoch": 0.683949785965081,
+      "grad_norm": 3.5272167622613964,
+      "learning_rate": 4.7998781516757295e-06,
+      "loss": 2.7736,
+      "step": 3555
+    },
+    {
+      "epoch": 0.6841421769034678,
+      "grad_norm": 3.420516065511884,
+      "learning_rate": 4.794555973709783e-06,
+      "loss": 2.9855,
+      "step": 3556
+    },
+    {
+      "epoch": 0.6843345678418546,
+      "grad_norm": 3.7802022050398207,
+      "learning_rate": 4.789235817478322e-06,
+      "loss": 3.16,
+      "step": 3557
+    },
+    {
+      "epoch": 0.6845269587802415,
+      "grad_norm": 3.5421191260225253,
+      "learning_rate": 4.783917685047621e-06,
+      "loss": 3.0854,
+      "step": 3558
+    },
+    {
+      "epoch": 0.6847193497186282,
+      "grad_norm": 3.5162457735422104,
+      "learning_rate": 4.778601578483187e-06,
+      "loss": 3.1998,
+      "step": 3559
+    },
+    {
+      "epoch": 0.6849117406570151,
+      "grad_norm": 3.3562717962437,
+      "learning_rate": 4.773287499849737e-06,
+      "loss": 3.0145,
+      "step": 3560
+    },
+    {
+      "epoch": 0.6851041315954018,
+      "grad_norm": 3.6239449837534434,
+      "learning_rate": 4.767975451211191e-06,
+      "loss": 2.9489,
+      "step": 3561
+    },
+    {
+      "epoch": 0.6852965225337887,
+      "grad_norm": 3.307007446694776,
+      "learning_rate": 4.762665434630692e-06,
+      "loss": 2.8603,
+      "step": 3562
+    },
+    {
+      "epoch": 0.6854889134721754,
+      "grad_norm": 2.4234842781078636,
+      "learning_rate": 4.757357452170588e-06,
+      "loss": 2.6896,
+      "step": 3563
+    },
+    {
+      "epoch": 0.6856813044105623,
+      "grad_norm": 3.6910980045097626,
+      "learning_rate": 4.752051505892438e-06,
+      "loss": 2.9605,
+      "step": 3564
+    },
+    {
+      "epoch": 0.685873695348949,
+      "grad_norm": 3.337405308629826,
+      "learning_rate": 4.746747597857014e-06,
+      "loss": 2.9358,
+      "step": 3565
+    },
+    {
+      "epoch": 0.6860660862873359,
+      "grad_norm": 3.123854364870766,
+      "learning_rate": 4.741445730124287e-06,
+      "loss": 2.9758,
+      "step": 3566
+    },
+    {
+      "epoch": 0.6862584772257226,
+      "grad_norm": 3.273872236783852,
+      "learning_rate": 4.736145904753445e-06,
+      "loss": 3.0374,
+      "step": 3567
+    },
+    {
+      "epoch": 0.6864508681641095,
+      "grad_norm": 3.720362579268897,
+      "learning_rate": 4.730848123802877e-06,
+      "loss": 3.173,
+      "step": 3568
+    },
+    {
+      "epoch": 0.6866432591024962,
+      "grad_norm": 3.551469207083505,
+      "learning_rate": 4.725552389330183e-06,
+      "loss": 3.0247,
+      "step": 3569
+    },
+    {
+      "epoch": 0.6868356500408831,
+      "grad_norm": 3.895439135795643,
+      "learning_rate": 4.720258703392161e-06,
+      "loss": 3.0728,
+      "step": 3570
+    },
+    {
+      "epoch": 0.6870280409792698,
+      "grad_norm": 3.3556750468486403,
+      "learning_rate": 4.714967068044826e-06,
+      "loss": 2.8805,
+      "step": 3571
+    },
+    {
+      "epoch": 0.6872204319176567,
+      "grad_norm": 3.3461828843849326,
+      "learning_rate": 4.7096774853433765e-06,
+      "loss": 2.8317,
+      "step": 3572
+    },
+    {
+      "epoch": 0.6874128228560434,
+      "grad_norm": 3.747426724512767,
+      "learning_rate": 4.704389957342237e-06,
+      "loss": 2.8978,
+      "step": 3573
+    },
+    {
+      "epoch": 0.6876052137944303,
+      "grad_norm": 3.426577247002292,
+      "learning_rate": 4.699104486095008e-06,
+      "loss": 2.8034,
+      "step": 3574
+    },
+    {
+      "epoch": 0.687797604732817,
+      "grad_norm": 3.5211269637252354,
+      "learning_rate": 4.69382107365452e-06,
+      "loss": 3.0321,
+      "step": 3575
+    },
+    {
+      "epoch": 0.6879899956712039,
+      "grad_norm": 3.567854924328207,
+      "learning_rate": 4.6885397220727855e-06,
+      "loss": 3.1482,
+      "step": 3576
+    },
+    {
+      "epoch": 0.6881823866095906,
+      "grad_norm": 3.687574483967839,
+      "learning_rate": 4.683260433401016e-06,
+      "loss": 3.0669,
+      "step": 3577
+    },
+    {
+      "epoch": 0.6883747775479775,
+      "grad_norm": 3.2573438581617333,
+      "learning_rate": 4.677983209689631e-06,
+      "loss": 2.8319,
+      "step": 3578
+    },
+    {
+      "epoch": 0.6885671684863643,
+      "grad_norm": 3.93678807845847,
+      "learning_rate": 4.6727080529882394e-06,
+      "loss": 3.0957,
+      "step": 3579
+    },
+    {
+      "epoch": 0.6887595594247511,
+      "grad_norm": 3.230346511925388,
+      "learning_rate": 4.667434965345654e-06,
+      "loss": 2.89,
+      "step": 3580
+    },
+    {
+      "epoch": 0.6889519503631379,
+      "grad_norm": 3.424388966214222,
+      "learning_rate": 4.6621639488098856e-06,
+      "loss": 3.0295,
+      "step": 3581
+    },
+    {
+      "epoch": 0.6891443413015247,
+      "grad_norm": 3.6573997805787792,
+      "learning_rate": 4.656895005428127e-06,
+      "loss": 2.9461,
+      "step": 3582
+    },
+    {
+      "epoch": 0.6893367322399115,
+      "grad_norm": 3.3822137352974324,
+      "learning_rate": 4.651628137246781e-06,
+      "loss": 3.1408,
+      "step": 3583
+    },
+    {
+      "epoch": 0.6895291231782983,
+      "grad_norm": 3.3577380836152892,
+      "learning_rate": 4.6463633463114395e-06,
+      "loss": 2.9559,
+      "step": 3584
+    },
+    {
+      "epoch": 0.6897215141166851,
+      "grad_norm": 3.2240092515042473,
+      "learning_rate": 4.641100634666877e-06,
+      "loss": 2.8423,
+      "step": 3585
+    },
+    {
+      "epoch": 0.6899139050550719,
+      "grad_norm": 3.332645526009152,
+      "learning_rate": 4.635840004357086e-06,
+      "loss": 2.9747,
+      "step": 3586
+    },
+    {
+      "epoch": 0.6901062959934587,
+      "grad_norm": 3.138046086508054,
+      "learning_rate": 4.630581457425222e-06,
+      "loss": 2.9909,
+      "step": 3587
+    },
+    {
+      "epoch": 0.6902986869318455,
+      "grad_norm": 3.529464271838886,
+      "learning_rate": 4.625324995913648e-06,
+      "loss": 2.8553,
+      "step": 3588
+    },
+    {
+      "epoch": 0.6904910778702323,
+      "grad_norm": 3.4097171598106915,
+      "learning_rate": 4.620070621863917e-06,
+      "loss": 3.0716,
+      "step": 3589
+    },
+    {
+      "epoch": 0.6906834688086191,
+      "grad_norm": 3.387571447897359,
+      "learning_rate": 4.614818337316759e-06,
+      "loss": 2.9738,
+      "step": 3590
+    },
+    {
+      "epoch": 0.6908758597470059,
+      "grad_norm": 2.287010189941132,
+      "learning_rate": 4.609568144312107e-06,
+      "loss": 2.6293,
+      "step": 3591
+    },
+    {
+      "epoch": 0.6910682506853927,
+      "grad_norm": 3.222660422630561,
+      "learning_rate": 4.6043200448890724e-06,
+      "loss": 2.96,
+      "step": 3592
+    },
+    {
+      "epoch": 0.6912606416237795,
+      "grad_norm": 2.4594893552313617,
+      "learning_rate": 4.599074041085958e-06,
+      "loss": 2.6802,
+      "step": 3593
+    },
+    {
+      "epoch": 0.6914530325621663,
+      "grad_norm": 3.5311730910658827,
+      "learning_rate": 4.593830134940256e-06,
+      "loss": 2.9039,
+      "step": 3594
+    },
+    {
+      "epoch": 0.6916454235005531,
+      "grad_norm": 3.3891220019220523,
+      "learning_rate": 4.588588328488629e-06,
+      "loss": 2.9644,
+      "step": 3595
+    },
+    {
+      "epoch": 0.6918378144389399,
+      "grad_norm": 3.421592465163304,
+      "learning_rate": 4.5833486237669414e-06,
+      "loss": 2.8493,
+      "step": 3596
+    },
+    {
+      "epoch": 0.6920302053773267,
+      "grad_norm": 3.4727489767562933,
+      "learning_rate": 4.578111022810231e-06,
+      "loss": 3.1785,
+      "step": 3597
+    },
+    {
+      "epoch": 0.6922225963157135,
+      "grad_norm": 3.496575735821459,
+      "learning_rate": 4.5728755276527225e-06,
+      "loss": 2.702,
+      "step": 3598
+    },
+    {
+      "epoch": 0.6924149872541003,
+      "grad_norm": 3.343015277607264,
+      "learning_rate": 4.567642140327823e-06,
+      "loss": 2.7301,
+      "step": 3599
+    },
+    {
+      "epoch": 0.6926073781924872,
+      "grad_norm": 3.5486639164736578,
+      "learning_rate": 4.562410862868123e-06,
+      "loss": 2.9329,
+      "step": 3600
+    },
+    {
+      "epoch": 0.6927997691308739,
+      "grad_norm": 3.1709681660871127,
+      "learning_rate": 4.557181697305383e-06,
+      "loss": 2.8063,
+      "step": 3601
+    },
+    {
+      "epoch": 0.6929921600692608,
+      "grad_norm": 3.2549253280249464,
+      "learning_rate": 4.551954645670557e-06,
+      "loss": 2.7968,
+      "step": 3602
+    },
+    {
+      "epoch": 0.6931845510076475,
+      "grad_norm": 2.261853818565958,
+      "learning_rate": 4.546729709993762e-06,
+      "loss": 2.5981,
+      "step": 3603
+    },
+    {
+      "epoch": 0.6933769419460344,
+      "grad_norm": 2.2066610493916405,
+      "learning_rate": 4.541506892304314e-06,
+      "loss": 2.4699,
+      "step": 3604
+    },
+    {
+      "epoch": 0.6935693328844211,
+      "grad_norm": 3.5526328806463447,
+      "learning_rate": 4.536286194630694e-06,
+      "loss": 3.0668,
+      "step": 3605
+    },
+    {
+      "epoch": 0.693761723822808,
+      "grad_norm": 3.6983598031045295,
+      "learning_rate": 4.531067619000553e-06,
+      "loss": 3.1869,
+      "step": 3606
+    },
+    {
+      "epoch": 0.6939541147611947,
+      "grad_norm": 3.3949520930863266,
+      "learning_rate": 4.525851167440731e-06,
+      "loss": 2.7329,
+      "step": 3607
+    },
+    {
+      "epoch": 0.6941465056995816,
+      "grad_norm": 3.398458986511719,
+      "learning_rate": 4.52063684197724e-06,
+      "loss": 2.9151,
+      "step": 3608
+    },
+    {
+      "epoch": 0.6943388966379683,
+      "grad_norm": 4.0294720537561695,
+      "learning_rate": 4.515424644635254e-06,
+      "loss": 3.0293,
+      "step": 3609
+    },
+    {
+      "epoch": 0.6945312875763552,
+      "grad_norm": 3.5574310211670603,
+      "learning_rate": 4.510214577439146e-06,
+      "loss": 2.961,
+      "step": 3610
+    },
+    {
+      "epoch": 0.6947236785147419,
+      "grad_norm": 3.7010961085707517,
+      "learning_rate": 4.5050066424124324e-06,
+      "loss": 3.113,
+      "step": 3611
+    },
+    {
+      "epoch": 0.6949160694531288,
+      "grad_norm": 3.188877777253753,
+      "learning_rate": 4.49980084157782e-06,
+      "loss": 2.6658,
+      "step": 3612
+    },
+    {
+      "epoch": 0.6951084603915155,
+      "grad_norm": 3.476850713460373,
+      "learning_rate": 4.494597176957186e-06,
+      "loss": 3.1872,
+      "step": 3613
+    },
+    {
+      "epoch": 0.6953008513299024,
+      "grad_norm": 3.5902948422938468,
+      "learning_rate": 4.489395650571562e-06,
+      "loss": 2.8138,
+      "step": 3614
+    },
+    {
+      "epoch": 0.6954932422682891,
+      "grad_norm": 3.4916410265530886,
+      "learning_rate": 4.4841962644411765e-06,
+      "loss": 2.8978,
+      "step": 3615
+    },
+    {
+      "epoch": 0.695685633206676,
+      "grad_norm": 3.633388026452155,
+      "learning_rate": 4.4789990205854e-06,
+      "loss": 3.0204,
+      "step": 3616
+    },
+    {
+      "epoch": 0.6958780241450627,
+      "grad_norm": 3.458657063969127,
+      "learning_rate": 4.473803921022784e-06,
+      "loss": 2.8054,
+      "step": 3617
+    },
+    {
+      "epoch": 0.6960704150834496,
+      "grad_norm": 3.2390624967225032,
+      "learning_rate": 4.468610967771051e-06,
+      "loss": 2.923,
+      "step": 3618
+    },
+    {
+      "epoch": 0.6962628060218363,
+      "grad_norm": 3.2994394038692927,
+      "learning_rate": 4.4634201628470766e-06,
+      "loss": 3.2333,
+      "step": 3619
+    },
+    {
+      "epoch": 0.6964551969602232,
+      "grad_norm": 3.52585352409715,
+      "learning_rate": 4.458231508266912e-06,
+      "loss": 3.1406,
+      "step": 3620
+    },
+    {
+      "epoch": 0.6966475878986099,
+      "grad_norm": 3.271225797871454,
+      "learning_rate": 4.453045006045773e-06,
+      "loss": 3.0575,
+      "step": 3621
+    },
+    {
+      "epoch": 0.6968399788369968,
+      "grad_norm": 3.4424766687960697,
+      "learning_rate": 4.447860658198035e-06,
+      "loss": 3.0237,
+      "step": 3622
+    },
+    {
+      "epoch": 0.6970323697753836,
+      "grad_norm": 3.6395674830042983,
+      "learning_rate": 4.442678466737245e-06,
+      "loss": 3.029,
+      "step": 3623
+    },
+    {
+      "epoch": 0.6972247607137704,
+      "grad_norm": 2.4661112443608046,
+      "learning_rate": 4.4374984336760975e-06,
+      "loss": 2.674,
+      "step": 3624
+    },
+    {
+      "epoch": 0.6974171516521572,
+      "grad_norm": 3.3512777964540392,
+      "learning_rate": 4.432320561026461e-06,
+      "loss": 2.8511,
+      "step": 3625
+    },
+    {
+      "epoch": 0.697609542590544,
+      "grad_norm": 3.7727829212155024,
+      "learning_rate": 4.427144850799363e-06,
+      "loss": 3.1476,
+      "step": 3626
+    },
+    {
+      "epoch": 0.6978019335289308,
+      "grad_norm": 3.482059765796887,
+      "learning_rate": 4.421971305004989e-06,
+      "loss": 3.1318,
+      "step": 3627
+    },
+    {
+      "epoch": 0.6979943244673176,
+      "grad_norm": 3.39987337889336,
+      "learning_rate": 4.416799925652684e-06,
+      "loss": 2.7939,
+      "step": 3628
+    },
+    {
+      "epoch": 0.6981867154057044,
+      "grad_norm": 3.5517546697061078,
+      "learning_rate": 4.411630714750956e-06,
+      "loss": 2.9394,
+      "step": 3629
+    },
+    {
+      "epoch": 0.6983791063440912,
+      "grad_norm": 3.434488208526668,
+      "learning_rate": 4.4064636743074605e-06,
+      "loss": 2.9676,
+      "step": 3630
+    },
+    {
+      "epoch": 0.698571497282478,
+      "grad_norm": 3.474474999733661,
+      "learning_rate": 4.40129880632902e-06,
+      "loss": 3.1736,
+      "step": 3631
+    },
+    {
+      "epoch": 0.6987638882208648,
+      "grad_norm": 3.533973577195402,
+      "learning_rate": 4.396136112821608e-06,
+      "loss": 3.2047,
+      "step": 3632
+    },
+    {
+      "epoch": 0.6989562791592516,
+      "grad_norm": 3.598427519293094,
+      "learning_rate": 4.390975595790358e-06,
+      "loss": 2.873,
+      "step": 3633
+    },
+    {
+      "epoch": 0.6991486700976384,
+      "grad_norm": 3.35872042729811,
+      "learning_rate": 4.385817257239556e-06,
+      "loss": 3.0422,
+      "step": 3634
+    },
+    {
+      "epoch": 0.6993410610360252,
+      "grad_norm": 3.5813853934312574,
+      "learning_rate": 4.380661099172636e-06,
+      "loss": 2.9708,
+      "step": 3635
+    },
+    {
+      "epoch": 0.699533451974412,
+      "grad_norm": 3.421325239043944,
+      "learning_rate": 4.375507123592194e-06,
+      "loss": 3.1751,
+      "step": 3636
+    },
+    {
+      "epoch": 0.6997258429127988,
+      "grad_norm": 3.500154499144847,
+      "learning_rate": 4.370355332499977e-06,
+      "loss": 2.8768,
+      "step": 3637
+    },
+    {
+      "epoch": 0.6999182338511856,
+      "grad_norm": 3.4960778987688714,
+      "learning_rate": 4.365205727896872e-06,
+      "loss": 2.9647,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7001106247895724,
+      "grad_norm": 3.541444260286821,
+      "learning_rate": 4.36005831178294e-06,
+      "loss": 3.1464,
+      "step": 3639
+    },
+    {
+      "epoch": 0.7003030157279592,
+      "grad_norm": 3.6485786089765706,
+      "learning_rate": 4.354913086157367e-06,
+      "loss": 3.1477,
+      "step": 3640
+    },
+    {
+      "epoch": 0.700495406666346,
+      "grad_norm": 3.643559894022245,
+      "learning_rate": 4.349770053018502e-06,
+      "loss": 3.1569,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7006877976047328,
+      "grad_norm": 3.3319079663734485,
+      "learning_rate": 4.344629214363845e-06,
+      "loss": 3.3949,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7008801885431196,
+      "grad_norm": 3.6529843189944313,
+      "learning_rate": 4.339490572190031e-06,
+      "loss": 3.0649,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7010725794815065,
+      "grad_norm": 3.5095756751079303,
+      "learning_rate": 4.334354128492851e-06,
+      "loss": 2.9347,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7012649704198932,
+      "grad_norm": 3.3234548047855883,
+      "learning_rate": 4.329219885267244e-06,
+      "loss": 2.9703,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7014573613582801,
+      "grad_norm": 3.656573131115749,
+      "learning_rate": 4.324087844507289e-06,
+      "loss": 2.9397,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7016497522966668,
+      "grad_norm": 3.6263895020017998,
+      "learning_rate": 4.318958008206214e-06,
+      "loss": 3.0349,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7018421432350537,
+      "grad_norm": 3.514473446321026,
+      "learning_rate": 4.313830378356384e-06,
+      "loss": 2.533,
+      "step": 3648
+    },
+    {
+      "epoch": 0.7020345341734404,
+      "grad_norm": 3.3865580957006194,
+      "learning_rate": 4.3087049569493136e-06,
+      "loss": 2.9391,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7022269251118273,
+      "grad_norm": 3.5966164031401533,
+      "learning_rate": 4.303581745975656e-06,
+      "loss": 2.9661,
+      "step": 3650
+    },
+    {
+      "epoch": 0.702419316050214,
+      "grad_norm": 3.8483153595442654,
+      "learning_rate": 4.2984607474252084e-06,
+      "loss": 2.8499,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7026117069886009,
+      "grad_norm": 3.2434907303310236,
+      "learning_rate": 4.293341963286912e-06,
+      "loss": 3.035,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7028040979269876,
+      "grad_norm": 3.7330611416269006,
+      "learning_rate": 4.288225395548835e-06,
+      "loss": 3.2114,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7029964888653745,
+      "grad_norm": 3.42785478738359,
+      "learning_rate": 4.283111046198198e-06,
+      "loss": 2.7748,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7031888798037612,
+      "grad_norm": 3.467859750698643,
+      "learning_rate": 4.277998917221354e-06,
+      "loss": 3.2721,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7033812707421481,
+      "grad_norm": 3.2687337933205396,
+      "learning_rate": 4.272889010603798e-06,
+      "loss": 2.8283,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7035736616805348,
+      "grad_norm": 3.2297885294061435,
+      "learning_rate": 4.267781328330155e-06,
+      "loss": 3.103,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7037660526189217,
+      "grad_norm": 3.5738803874152127,
+      "learning_rate": 4.262675872384197e-06,
+      "loss": 3.1275,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7039584435573084,
+      "grad_norm": 3.3173990691253272,
+      "learning_rate": 4.257572644748813e-06,
+      "loss": 3.0224,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7041508344956953,
+      "grad_norm": 3.7185086965407597,
+      "learning_rate": 4.252471647406045e-06,
+      "loss": 3.0851,
+      "step": 3660
+    },
+    {
+      "epoch": 0.704343225434082,
+      "grad_norm": 3.9964924379230182,
+      "learning_rate": 4.2473728823370605e-06,
+      "loss": 2.8651,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7045356163724689,
+      "grad_norm": 3.303849135349504,
+      "learning_rate": 4.242276351522161e-06,
+      "loss": 3.2048,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7047280073108556,
+      "grad_norm": 3.920021786700408,
+      "learning_rate": 4.237182056940784e-06,
+      "loss": 3.1795,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7049203982492425,
+      "grad_norm": 3.4237315086398747,
+      "learning_rate": 4.232090000571488e-06,
+      "loss": 3.1124,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7051127891876292,
+      "grad_norm": 3.6147089066523903,
+      "learning_rate": 4.2270001843919714e-06,
+      "loss": 3.1115,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7053051801260161,
+      "grad_norm": 2.5477755151986563,
+      "learning_rate": 4.221912610379065e-06,
+      "loss": 2.8615,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7054975710644029,
+      "grad_norm": 3.432080413683231,
+      "learning_rate": 4.216827280508712e-06,
+      "loss": 3.0423,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7056899620027897,
+      "grad_norm": 3.4913184766925607,
+      "learning_rate": 4.211744196756011e-06,
+      "loss": 3.1848,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 3.169074890819567,
+      "learning_rate": 4.206663361095164e-06,
+      "loss": 2.8883,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7060747438795633,
+      "grad_norm": 3.6685501321028333,
+      "learning_rate": 4.201584775499509e-06,
+      "loss": 3.0153,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7062671348179501,
+      "grad_norm": 3.576490555449451,
+      "learning_rate": 4.196508441941516e-06,
+      "loss": 3.0498,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7064595257563369,
+      "grad_norm": 3.6396592683522866,
+      "learning_rate": 4.191434362392768e-06,
+      "loss": 3.2427,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7066519166947237,
+      "grad_norm": 3.913133622556318,
+      "learning_rate": 4.186362538823981e-06,
+      "loss": 3.1404,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7068443076331105,
+      "grad_norm": 3.548287895185263,
+      "learning_rate": 4.181292973204992e-06,
+      "loss": 2.9921,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7070366985714973,
+      "grad_norm": 3.543464623054729,
+      "learning_rate": 4.1762256675047655e-06,
+      "loss": 2.8131,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7072290895098841,
+      "grad_norm": 3.503657445935293,
+      "learning_rate": 4.171160623691384e-06,
+      "loss": 2.9557,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7074214804482709,
+      "grad_norm": 3.53460607203215,
+      "learning_rate": 4.166097843732048e-06,
+      "loss": 3.147,
+      "step": 3677
+    },
+    {
+      "epoch": 0.7076138713866577,
+      "grad_norm": 3.3788355318239818,
+      "learning_rate": 4.161037329593085e-06,
+      "loss": 2.9136,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7078062623250445,
+      "grad_norm": 3.4104652067773236,
+      "learning_rate": 4.155979083239942e-06,
+      "loss": 2.8964,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7079986532634313,
+      "grad_norm": 2.2448288203569313,
+      "learning_rate": 4.1509231066371815e-06,
+      "loss": 2.4635,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7081910442018181,
+      "grad_norm": 3.2008781745068946,
+      "learning_rate": 4.1458694017484915e-06,
+      "loss": 2.8419,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7083834351402049,
+      "grad_norm": 3.181627249745761,
+      "learning_rate": 4.140817970536664e-06,
+      "loss": 2.8597,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7085758260785917,
+      "grad_norm": 3.626326037081585,
+      "learning_rate": 4.135768814963622e-06,
+      "loss": 2.7496,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7087682170169785,
+      "grad_norm": 3.414005587286497,
+      "learning_rate": 4.130721936990399e-06,
+      "loss": 3.1152,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7089606079553653,
+      "grad_norm": 3.265872664014813,
+      "learning_rate": 4.1256773385771444e-06,
+      "loss": 3.1218,
+      "step": 3685
+    },
+    {
+      "epoch": 0.709152998893752,
+      "grad_norm": 3.678366208302878,
+      "learning_rate": 4.120635021683122e-06,
+      "loss": 3.0664,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7093453898321389,
+      "grad_norm": 3.4946893384532496,
+      "learning_rate": 4.115594988266711e-06,
+      "loss": 2.9932,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7095377807705258,
+      "grad_norm": 4.0326429464365345,
+      "learning_rate": 4.1105572402853976e-06,
+      "loss": 3.2098,
+      "step": 3688
+    },
+    {
+      "epoch": 0.7097301717089125,
+      "grad_norm": 3.7526904522794817,
+      "learning_rate": 4.1055217796957895e-06,
+      "loss": 2.8851,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7099225626472994,
+      "grad_norm": 3.7124186648118087,
+      "learning_rate": 4.100488608453599e-06,
+      "loss": 2.9408,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7101149535856861,
+      "grad_norm": 3.5688848943257008,
+      "learning_rate": 4.095457728513652e-06,
+      "loss": 3.1351,
+      "step": 3691
+    },
+    {
+      "epoch": 0.710307344524073,
+      "grad_norm": 3.3298554384971606,
+      "learning_rate": 4.09042914182989e-06,
+      "loss": 2.9673,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7104997354624597,
+      "grad_norm": 3.2068420627267864,
+      "learning_rate": 4.08540285035535e-06,
+      "loss": 2.843,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7106921264008466,
+      "grad_norm": 3.361898230572475,
+      "learning_rate": 4.0803788560421885e-06,
+      "loss": 2.8687,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7108845173392333,
+      "grad_norm": 3.5372226475114164,
+      "learning_rate": 4.075357160841671e-06,
+      "loss": 3.2281,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7110769082776202,
+      "grad_norm": 3.4091068384457968,
+      "learning_rate": 4.070337766704155e-06,
+      "loss": 2.8439,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7112692992160069,
+      "grad_norm": 3.061682101347099,
+      "learning_rate": 4.065320675579132e-06,
+      "loss": 2.823,
+      "step": 3697
+    },
+    {
+      "epoch": 0.7114616901543938,
+      "grad_norm": 3.5029611505111533,
+      "learning_rate": 4.0603058894151685e-06,
+      "loss": 2.8949,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7116540810927805,
+      "grad_norm": 3.805509383121566,
+      "learning_rate": 4.055293410159954e-06,
+      "loss": 3.092,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7118464720311674,
+      "grad_norm": 3.5076827005995703,
+      "learning_rate": 4.050283239760282e-06,
+      "loss": 3.0458,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7120388629695541,
+      "grad_norm": 3.5953826103728614,
+      "learning_rate": 4.045275380162038e-06,
+      "loss": 3.1713,
+      "step": 3701
+    },
+    {
+      "epoch": 0.712231253907941,
+      "grad_norm": 3.630123788732817,
+      "learning_rate": 4.04026983331022e-06,
+      "loss": 3.0657,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7124236448463277,
+      "grad_norm": 2.1615630572713647,
+      "learning_rate": 4.035266601148924e-06,
+      "loss": 2.4269,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7126160357847146,
+      "grad_norm": 3.4525923008900388,
+      "learning_rate": 4.03026568562135e-06,
+      "loss": 3.074,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7128084267231013,
+      "grad_norm": 3.746316219527655,
+      "learning_rate": 4.025267088669797e-06,
+      "loss": 3.0585,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7130008176614882,
+      "grad_norm": 3.71243978875533,
+      "learning_rate": 4.020270812235656e-06,
+      "loss": 2.952,
+      "step": 3706
+    },
+    {
+      "epoch": 0.7131932085998749,
+      "grad_norm": 3.4043925290443466,
+      "learning_rate": 4.015276858259427e-06,
+      "loss": 2.9399,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7133855995382617,
+      "grad_norm": 3.6511927788402248,
+      "learning_rate": 4.010285228680705e-06,
+      "loss": 2.877,
+      "step": 3708
+    },
+    {
+      "epoch": 0.7135779904766486,
+      "grad_norm": 3.1730348371337107,
+      "learning_rate": 4.005295925438181e-06,
+      "loss": 2.9489,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7137703814150353,
+      "grad_norm": 3.4567302681274303,
+      "learning_rate": 4.000308950469646e-06,
+      "loss": 2.8924,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7139627723534222,
+      "grad_norm": 3.601404311251763,
+      "learning_rate": 3.995324305711976e-06,
+      "loss": 2.9536,
+      "step": 3711
+    },
+    {
+      "epoch": 0.714155163291809,
+      "grad_norm": 3.5309052791694953,
+      "learning_rate": 3.990341993101154e-06,
+      "loss": 3.2293,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7143475542301958,
+      "grad_norm": 3.5610523736076587,
+      "learning_rate": 3.985362014572256e-06,
+      "loss": 2.9644,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7145399451685825,
+      "grad_norm": 3.5074724968767796,
+      "learning_rate": 3.9803843720594385e-06,
+      "loss": 3.1043,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7147323361069694,
+      "grad_norm": 3.4195362239890814,
+      "learning_rate": 3.97540906749597e-06,
+      "loss": 2.9974,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7149247270453561,
+      "grad_norm": 3.4405701276067884,
+      "learning_rate": 3.970436102814203e-06,
+      "loss": 2.5923,
+      "step": 3716
+    },
+    {
+      "epoch": 0.715117117983743,
+      "grad_norm": 3.565275629467232,
+      "learning_rate": 3.965465479945569e-06,
+      "loss": 3.0519,
+      "step": 3717
+    },
+    {
+      "epoch": 0.7153095089221297,
+      "grad_norm": 2.1383281262924383,
+      "learning_rate": 3.9604972008206085e-06,
+      "loss": 2.5067,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7155018998605166,
+      "grad_norm": 3.411705117079131,
+      "learning_rate": 3.955531267368942e-06,
+      "loss": 2.8985,
+      "step": 3719
+    },
+    {
+      "epoch": 0.7156942907989033,
+      "grad_norm": 3.3782734908904164,
+      "learning_rate": 3.950567681519279e-06,
+      "loss": 3.1318,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7158866817372902,
+      "grad_norm": 4.1501307889181485,
+      "learning_rate": 3.945606445199427e-06,
+      "loss": 3.3131,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7160790726756769,
+      "grad_norm": 3.265562442821812,
+      "learning_rate": 3.940647560336262e-06,
+      "loss": 2.8567,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7162714636140638,
+      "grad_norm": 3.4665448845178397,
+      "learning_rate": 3.935691028855763e-06,
+      "loss": 3.1419,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7164638545524505,
+      "grad_norm": 3.4876990878537626,
+      "learning_rate": 3.930736852682993e-06,
+      "loss": 2.8237,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7166562454908374,
+      "grad_norm": 3.6873506783961,
+      "learning_rate": 3.9257850337420856e-06,
+      "loss": 3.1883,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7168486364292241,
+      "grad_norm": 2.375842996481896,
+      "learning_rate": 3.920835573956285e-06,
+      "loss": 2.4882,
+      "step": 3726
+    },
+    {
+      "epoch": 0.717041027367611,
+      "grad_norm": 3.153593675143908,
+      "learning_rate": 3.915888475247894e-06,
+      "loss": 2.7537,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7172334183059977,
+      "grad_norm": 3.352857302260258,
+      "learning_rate": 3.910943739538313e-06,
+      "loss": 3.0866,
+      "step": 3728
+    },
+    {
+      "epoch": 0.7174258092443846,
+      "grad_norm": 3.323189597588503,
+      "learning_rate": 3.906001368748023e-06,
+      "loss": 2.8689,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7176182001827713,
+      "grad_norm": 3.4688249466392214,
+      "learning_rate": 3.901061364796574e-06,
+      "loss": 3.0046,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7178105911211582,
+      "grad_norm": 3.664644833376379,
+      "learning_rate": 3.8961237296026155e-06,
+      "loss": 2.9804,
+      "step": 3731
+    },
+    {
+      "epoch": 0.718002982059545,
+      "grad_norm": 3.4277521335054786,
+      "learning_rate": 3.891188465083865e-06,
+      "loss": 3.3419,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7181953729979318,
+      "grad_norm": 3.428828942292238,
+      "learning_rate": 3.886255573157121e-06,
+      "loss": 3.1042,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7183877639363186,
+      "grad_norm": 3.5110136199783226,
+      "learning_rate": 3.88132505573827e-06,
+      "loss": 2.8305,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7185801548747054,
+      "grad_norm": 3.6429949967381483,
+      "learning_rate": 3.876396914742258e-06,
+      "loss": 2.8965,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7187725458130922,
+      "grad_norm": 3.453775421441825,
+      "learning_rate": 3.871471152083121e-06,
+      "loss": 2.984,
+      "step": 3736
+    },
+    {
+      "epoch": 0.718964936751479,
+      "grad_norm": 3.683688778884585,
+      "learning_rate": 3.866547769673968e-06,
+      "loss": 3.165,
+      "step": 3737
+    },
+    {
+      "epoch": 0.7191573276898658,
+      "grad_norm": 3.825258276469865,
+      "learning_rate": 3.861626769426988e-06,
+      "loss": 3.1238,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7193497186282526,
+      "grad_norm": 3.4485910245157725,
+      "learning_rate": 3.8567081532534374e-06,
+      "loss": 2.9196,
+      "step": 3739
+    },
+    {
+      "epoch": 0.7195421095666394,
+      "grad_norm": 3.247367900165205,
+      "learning_rate": 3.851791923063655e-06,
+      "loss": 2.8501,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7197345005050262,
+      "grad_norm": 3.40776799888761,
+      "learning_rate": 3.846878080767039e-06,
+      "loss": 2.831,
+      "step": 3741
+    },
+    {
+      "epoch": 0.719926891443413,
+      "grad_norm": 3.3715960689496147,
+      "learning_rate": 3.841966628272079e-06,
+      "loss": 2.8235,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7201192823817998,
+      "grad_norm": 3.1839012769889092,
+      "learning_rate": 3.837057567486314e-06,
+      "loss": 2.986,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7203116733201866,
+      "grad_norm": 3.086442117422391,
+      "learning_rate": 3.832150900316377e-06,
+      "loss": 3.1285,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7205040642585734,
+      "grad_norm": 3.624898615803773,
+      "learning_rate": 3.827246628667962e-06,
+      "loss": 2.9636,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7206964551969602,
+      "grad_norm": 3.624233669340631,
+      "learning_rate": 3.822344754445826e-06,
+      "loss": 2.8014,
+      "step": 3746
+    },
+    {
+      "epoch": 0.720888846135347,
+      "grad_norm": 3.3467342423944455,
+      "learning_rate": 3.817445279553801e-06,
+      "loss": 2.9615,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7210812370737338,
+      "grad_norm": 3.3983526459800735,
+      "learning_rate": 3.8125482058947905e-06,
+      "loss": 2.9919,
+      "step": 3748
+    },
+    {
+      "epoch": 0.7212736280121206,
+      "grad_norm": 2.4766977721455574,
+      "learning_rate": 3.8076535353707523e-06,
+      "loss": 2.6984,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7214660189505074,
+      "grad_norm": 3.5729321841581485,
+      "learning_rate": 3.8027612698827344e-06,
+      "loss": 3.1071,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7216584098888942,
+      "grad_norm": 3.3370443644054113,
+      "learning_rate": 3.7978714113308246e-06,
+      "loss": 2.9426,
+      "step": 3751
+    },
+    {
+      "epoch": 0.721850800827281,
+      "grad_norm": 3.560281155154998,
+      "learning_rate": 3.7929839616141917e-06,
+      "loss": 2.9814,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7220431917656679,
+      "grad_norm": 3.678494905005189,
+      "learning_rate": 3.788098922631067e-06,
+      "loss": 2.8248,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7222355827040546,
+      "grad_norm": 3.7523395007527536,
+      "learning_rate": 3.7832162962787355e-06,
+      "loss": 2.9521,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7224279736424415,
+      "grad_norm": 3.2354099093444426,
+      "learning_rate": 3.7783360844535653e-06,
+      "loss": 2.531,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7226203645808282,
+      "grad_norm": 3.3830424401133445,
+      "learning_rate": 3.773458289050963e-06,
+      "loss": 2.7371,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7228127555192151,
+      "grad_norm": 3.64675459859372,
+      "learning_rate": 3.768582911965414e-06,
+      "loss": 3.0891,
+      "step": 3757
+    },
+    {
+      "epoch": 0.7230051464576018,
+      "grad_norm": 3.4168839197473786,
+      "learning_rate": 3.763709955090461e-06,
+      "loss": 2.8588,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7231975373959887,
+      "grad_norm": 3.2561244557945903,
+      "learning_rate": 3.7588394203186963e-06,
+      "loss": 2.8003,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7233899283343754,
+      "grad_norm": 3.2375147661720804,
+      "learning_rate": 3.753971309541784e-06,
+      "loss": 2.5555,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7235823192727623,
+      "grad_norm": 3.493388165429392,
+      "learning_rate": 3.7491056246504433e-06,
+      "loss": 3.1148,
+      "step": 3761
+    },
+    {
+      "epoch": 0.723774710211149,
+      "grad_norm": 3.5508804638761915,
+      "learning_rate": 3.7442423675344474e-06,
+      "loss": 3.0957,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7239671011495359,
+      "grad_norm": 3.2195934805487396,
+      "learning_rate": 3.739381540082635e-06,
+      "loss": 3.0551,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7241594920879226,
+      "grad_norm": 3.460842618908131,
+      "learning_rate": 3.7345231441828876e-06,
+      "loss": 3.1416,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7243518830263095,
+      "grad_norm": 3.6965523387740697,
+      "learning_rate": 3.729667181722154e-06,
+      "loss": 3.0866,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7245442739646962,
+      "grad_norm": 3.647414800664074,
+      "learning_rate": 3.7248136545864345e-06,
+      "loss": 2.9697,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7247366649030831,
+      "grad_norm": 3.4278472999882674,
+      "learning_rate": 3.719962564660783e-06,
+      "loss": 2.9021,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7249290558414698,
+      "grad_norm": 3.480565696261719,
+      "learning_rate": 3.7151139138293056e-06,
+      "loss": 3.2989,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7251214467798567,
+      "grad_norm": 4.082716538826984,
+      "learning_rate": 3.7102677039751667e-06,
+      "loss": 2.854,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7253138377182434,
+      "grad_norm": 3.559972283152064,
+      "learning_rate": 3.705423936980572e-06,
+      "loss": 2.8831,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7255062286566303,
+      "grad_norm": 3.6877791258874604,
+      "learning_rate": 3.700582614726791e-06,
+      "loss": 3.2277,
+      "step": 3771
+    },
+    {
+      "epoch": 0.725698619595017,
+      "grad_norm": 3.465664973263849,
+      "learning_rate": 3.6957437390941274e-06,
+      "loss": 3.0153,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7258910105334039,
+      "grad_norm": 4.019603323805785,
+      "learning_rate": 3.6909073119619555e-06,
+      "loss": 3.1115,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7260834014717906,
+      "grad_norm": 3.651101807919002,
+      "learning_rate": 3.6860733352086866e-06,
+      "loss": 2.9648,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7262757924101775,
+      "grad_norm": 3.7146958855503858,
+      "learning_rate": 3.6812418107117765e-06,
+      "loss": 2.9977,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7264681833485643,
+      "grad_norm": 3.5329448180725427,
+      "learning_rate": 3.6764127403477347e-06,
+      "loss": 2.8862,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7266605742869511,
+      "grad_norm": 3.5735981260100274,
+      "learning_rate": 3.6715861259921226e-06,
+      "loss": 2.9583,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7268529652253379,
+      "grad_norm": 3.915344939671466,
+      "learning_rate": 3.6667619695195287e-06,
+      "loss": 2.9486,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7270453561637247,
+      "grad_norm": 3.6231456029754425,
+      "learning_rate": 3.6619402728036157e-06,
+      "loss": 3.1685,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7272377471021115,
+      "grad_norm": 3.5236335435066657,
+      "learning_rate": 3.657121037717064e-06,
+      "loss": 3.2697,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7274301380404983,
+      "grad_norm": 3.58198566492163,
+      "learning_rate": 3.652304266131612e-06,
+      "loss": 3.3195,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7276225289788851,
+      "grad_norm": 3.5036580261371033,
+      "learning_rate": 3.6474899599180426e-06,
+      "loss": 3.004,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7278149199172719,
+      "grad_norm": 3.832963667426647,
+      "learning_rate": 3.642678120946168e-06,
+      "loss": 2.8861,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7280073108556587,
+      "grad_norm": 3.752470009012331,
+      "learning_rate": 3.6378687510848576e-06,
+      "loss": 3.1039,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7281997017940455,
+      "grad_norm": 3.743334370592176,
+      "learning_rate": 3.6330618522020124e-06,
+      "loss": 2.7957,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7283920927324323,
+      "grad_norm": 3.4093629409723096,
+      "learning_rate": 3.6282574261645776e-06,
+      "loss": 2.8753,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7285844836708191,
+      "grad_norm": 3.5060201436044727,
+      "learning_rate": 3.62345547483854e-06,
+      "loss": 3.1612,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7287768746092059,
+      "grad_norm": 3.5109084688308805,
+      "learning_rate": 3.618656000088916e-06,
+      "loss": 2.8641,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7289692655475927,
+      "grad_norm": 3.832996925186937,
+      "learning_rate": 3.6138590037797695e-06,
+      "loss": 2.9862,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7291616564859795,
+      "grad_norm": 3.3239706434037357,
+      "learning_rate": 3.6090644877741986e-06,
+      "loss": 3.0968,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7293540474243663,
+      "grad_norm": 3.6557656454695895,
+      "learning_rate": 3.6042724539343378e-06,
+      "loss": 3.1375,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7295464383627531,
+      "grad_norm": 3.4712576230520096,
+      "learning_rate": 3.599482904121361e-06,
+      "loss": 2.8069,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7297388293011399,
+      "grad_norm": 3.4831952914978115,
+      "learning_rate": 3.594695840195468e-06,
+      "loss": 3.0464,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7299312202395267,
+      "grad_norm": 3.591077101262401,
+      "learning_rate": 3.5899112640159017e-06,
+      "loss": 3.0688,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7301236111779135,
+      "grad_norm": 2.2748183443546535,
+      "learning_rate": 3.585129177440938e-06,
+      "loss": 2.4783,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7303160021163003,
+      "grad_norm": 3.5951092128781355,
+      "learning_rate": 3.580349582327882e-06,
+      "loss": 3.0621,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7305083930546872,
+      "grad_norm": 3.9189939017367728,
+      "learning_rate": 3.575572480533076e-06,
+      "loss": 3.2015,
+      "step": 3797
+    },
+    {
+      "epoch": 0.7307007839930739,
+      "grad_norm": 3.5126784926174146,
+      "learning_rate": 3.570797873911892e-06,
+      "loss": 3.0929,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7308931749314608,
+      "grad_norm": 3.2515711116710557,
+      "learning_rate": 3.566025764318728e-06,
+      "loss": 2.6242,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7310855658698475,
+      "grad_norm": 3.707928110911965,
+      "learning_rate": 3.5612561536070213e-06,
+      "loss": 2.7474,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7312779568082344,
+      "grad_norm": 3.6817553880384737,
+      "learning_rate": 3.5564890436292243e-06,
+      "loss": 3.315,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7314703477466211,
+      "grad_norm": 3.6208136863956644,
+      "learning_rate": 3.5517244362368363e-06,
+      "loss": 3.0112,
+      "step": 3802
+    },
+    {
+      "epoch": 0.731662738685008,
+      "grad_norm": 3.4624200024962737,
+      "learning_rate": 3.5469623332803795e-06,
+      "loss": 3.0413,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7318551296233947,
+      "grad_norm": 3.5748808495821645,
+      "learning_rate": 3.5422027366093893e-06,
+      "loss": 2.975,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7320475205617816,
+      "grad_norm": 3.63569151650079,
+      "learning_rate": 3.5374456480724427e-06,
+      "loss": 3.0492,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7322399115001683,
+      "grad_norm": 3.910415948873743,
+      "learning_rate": 3.532691069517142e-06,
+      "loss": 2.9934,
+      "step": 3806
+    },
+    {
+      "epoch": 0.7324323024385552,
+      "grad_norm": 2.3096594306411564,
+      "learning_rate": 3.5279390027901004e-06,
+      "loss": 2.444,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7326246933769419,
+      "grad_norm": 3.7162790664309058,
+      "learning_rate": 3.5231894497369802e-06,
+      "loss": 2.997,
+      "step": 3808
+    },
+    {
+      "epoch": 0.7328170843153288,
+      "grad_norm": 3.26782418334384,
+      "learning_rate": 3.5184424122024406e-06,
+      "loss": 2.9541,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7330094752537155,
+      "grad_norm": 2.2974721876191095,
+      "learning_rate": 3.5136978920301822e-06,
+      "loss": 2.6004,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7332018661921024,
+      "grad_norm": 3.7733076012465343,
+      "learning_rate": 3.508955891062924e-06,
+      "loss": 3.0554,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7333942571304891,
+      "grad_norm": 3.463893169499586,
+      "learning_rate": 3.5042164111423983e-06,
+      "loss": 2.9731,
+      "step": 3812
+    },
+    {
+      "epoch": 0.733586648068876,
+      "grad_norm": 3.505457203744759,
+      "learning_rate": 3.4994794541093667e-06,
+      "loss": 2.9318,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7337790390072627,
+      "grad_norm": 3.42770929733296,
+      "learning_rate": 3.4947450218036106e-06,
+      "loss": 2.9747,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7339714299456496,
+      "grad_norm": 3.36728786428956,
+      "learning_rate": 3.4900131160639283e-06,
+      "loss": 3.0208,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7341638208840363,
+      "grad_norm": 3.2534001355903617,
+      "learning_rate": 3.485283738728139e-06,
+      "loss": 3.1223,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7343562118224232,
+      "grad_norm": 3.7255043290751986,
+      "learning_rate": 3.4805568916330747e-06,
+      "loss": 3.0515,
+      "step": 3817
+    },
+    {
+      "epoch": 0.73454860276081,
+      "grad_norm": 3.5178636238437044,
+      "learning_rate": 3.4758325766145896e-06,
+      "loss": 3.2362,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7347409936991968,
+      "grad_norm": 3.4097901374403485,
+      "learning_rate": 3.471110795507554e-06,
+      "loss": 3.0407,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7349333846375836,
+      "grad_norm": 3.2687294176821364,
+      "learning_rate": 3.4663915501458523e-06,
+      "loss": 2.9956,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7351257755759704,
+      "grad_norm": 3.6136652999164305,
+      "learning_rate": 3.4616748423623893e-06,
+      "loss": 3.0262,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7353181665143572,
+      "grad_norm": 3.2240316240250193,
+      "learning_rate": 3.4569606739890737e-06,
+      "loss": 2.9944,
+      "step": 3822
+    },
+    {
+      "epoch": 0.735510557452744,
+      "grad_norm": 2.277178439631655,
+      "learning_rate": 3.452249046856836e-06,
+      "loss": 2.6716,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7357029483911308,
+      "grad_norm": 3.1900720221605643,
+      "learning_rate": 3.4475399627956197e-06,
+      "loss": 2.8969,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7358953393295176,
+      "grad_norm": 3.5573203545297862,
+      "learning_rate": 3.4428334236343774e-06,
+      "loss": 2.9517,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7360877302679044,
+      "grad_norm": 3.3757799823903922,
+      "learning_rate": 3.438129431201075e-06,
+      "loss": 3.1409,
+      "step": 3826
+    },
+    {
+      "epoch": 0.7362801212062912,
+      "grad_norm": 3.306397331040259,
+      "learning_rate": 3.433427987322693e-06,
+      "loss": 3.1856,
+      "step": 3827
+    },
+    {
+      "epoch": 0.736472512144678,
+      "grad_norm": 3.348763357716968,
+      "learning_rate": 3.4287290938252103e-06,
+      "loss": 3.0644,
+      "step": 3828
+    },
+    {
+      "epoch": 0.7366649030830648,
+      "grad_norm": 3.6672765055733367,
+      "learning_rate": 3.424032752533627e-06,
+      "loss": 2.916,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7368572940214516,
+      "grad_norm": 2.395448255135369,
+      "learning_rate": 3.4193389652719478e-06,
+      "loss": 2.6327,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7370496849598384,
+      "grad_norm": 3.586325005895689,
+      "learning_rate": 3.4146477338631856e-06,
+      "loss": 3.0147,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7372420758982252,
+      "grad_norm": 3.3858821919222746,
+      "learning_rate": 3.4099590601293632e-06,
+      "loss": 3.0498,
+      "step": 3832
+    },
+    {
+      "epoch": 0.737434466836612,
+      "grad_norm": 3.4608236859287684,
+      "learning_rate": 3.4052729458915024e-06,
+      "loss": 2.7888,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7376268577749988,
+      "grad_norm": 3.5399802563023055,
+      "learning_rate": 3.4005893929696377e-06,
+      "loss": 2.9825,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7378192487133856,
+      "grad_norm": 3.307453169156825,
+      "learning_rate": 3.3959084031828114e-06,
+      "loss": 2.8831,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7380116396517724,
+      "grad_norm": 2.490288375037646,
+      "learning_rate": 3.3912299783490567e-06,
+      "loss": 2.5871,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7382040305901592,
+      "grad_norm": 3.309111802449627,
+      "learning_rate": 3.3865541202854314e-06,
+      "loss": 3.0341,
+      "step": 3837
+    },
+    {
+      "epoch": 0.738396421528546,
+      "grad_norm": 3.3821387472419797,
+      "learning_rate": 3.3818808308079753e-06,
+      "loss": 3.0987,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7385888124669328,
+      "grad_norm": 3.6734057740421626,
+      "learning_rate": 3.3772101117317437e-06,
+      "loss": 2.888,
+      "step": 3839
+    },
+    {
+      "epoch": 0.7387812034053196,
+      "grad_norm": 3.521301479288793,
+      "learning_rate": 3.372541964870795e-06,
+      "loss": 2.8111,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7389735943437065,
+      "grad_norm": 3.5561932177238837,
+      "learning_rate": 3.367876392038174e-06,
+      "loss": 2.8118,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7391659852820932,
+      "grad_norm": 3.6419523592631196,
+      "learning_rate": 3.363213395045941e-06,
+      "loss": 3.0851,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7393583762204801,
+      "grad_norm": 3.493978761920176,
+      "learning_rate": 3.3585529757051504e-06,
+      "loss": 3.285,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7395507671588668,
+      "grad_norm": 3.278746764137485,
+      "learning_rate": 3.353895135825854e-06,
+      "loss": 3.0424,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7397431580972537,
+      "grad_norm": 3.351366709057618,
+      "learning_rate": 3.3492398772171074e-06,
+      "loss": 3.1276,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7399355490356404,
+      "grad_norm": 3.3431728824319995,
+      "learning_rate": 3.344587201686952e-06,
+      "loss": 3.0671,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7401279399740273,
+      "grad_norm": 3.552575396112952,
+      "learning_rate": 3.3399371110424372e-06,
+      "loss": 3.0624,
+      "step": 3847
+    },
+    {
+      "epoch": 0.740320330912414,
+      "grad_norm": 3.8154037559362903,
+      "learning_rate": 3.3352896070896057e-06,
+      "loss": 3.1513,
+      "step": 3848
+    },
+    {
+      "epoch": 0.7405127218508009,
+      "grad_norm": 3.6688808917134432,
+      "learning_rate": 3.330644691633492e-06,
+      "loss": 3.0229,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7407051127891876,
+      "grad_norm": 3.69866238531236,
+      "learning_rate": 3.3260023664781326e-06,
+      "loss": 3.1121,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7408975037275745,
+      "grad_norm": 3.6053590686389625,
+      "learning_rate": 3.321362633426547e-06,
+      "loss": 2.8333,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7410898946659612,
+      "grad_norm": 2.5418232010460127,
+      "learning_rate": 3.316725494280757e-06,
+      "loss": 2.6513,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7412822856043481,
+      "grad_norm": 3.3231458346742095,
+      "learning_rate": 3.3120909508417754e-06,
+      "loss": 2.9646,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7414746765427348,
+      "grad_norm": 3.4056993678013634,
+      "learning_rate": 3.307459004909599e-06,
+      "loss": 2.8959,
+      "step": 3854
+    },
+    {
+      "epoch": 0.7416670674811217,
+      "grad_norm": 3.611312775582591,
+      "learning_rate": 3.3028296582832285e-06,
+      "loss": 2.6722,
+      "step": 3855
+    },
+    {
+      "epoch": 0.7418594584195084,
+      "grad_norm": 3.8122989912284204,
+      "learning_rate": 3.2982029127606517e-06,
+      "loss": 3.0818,
+      "step": 3856
+    },
+    {
+      "epoch": 0.7420518493578953,
+      "grad_norm": 3.4184717006789924,
+      "learning_rate": 3.2935787701388346e-06,
+      "loss": 2.7794,
+      "step": 3857
+    },
+    {
+      "epoch": 0.742244240296282,
+      "grad_norm": 3.675525596802545,
+      "learning_rate": 3.2889572322137454e-06,
+      "loss": 2.8722,
+      "step": 3858
+    },
+    {
+      "epoch": 0.7424366312346689,
+      "grad_norm": 3.4019393454209634,
+      "learning_rate": 3.2843383007803364e-06,
+      "loss": 2.9659,
+      "step": 3859
+    },
+    {
+      "epoch": 0.7426290221730556,
+      "grad_norm": 3.2305776900594405,
+      "learning_rate": 3.279721977632546e-06,
+      "loss": 2.8105,
+      "step": 3860
+    },
+    {
+      "epoch": 0.7428214131114425,
+      "grad_norm": 3.5773892164186054,
+      "learning_rate": 3.275108264563306e-06,
+      "loss": 3.2415,
+      "step": 3861
+    },
+    {
+      "epoch": 0.7430138040498293,
+      "grad_norm": 3.284337252203322,
+      "learning_rate": 3.270497163364521e-06,
+      "loss": 2.8886,
+      "step": 3862
+    },
+    {
+      "epoch": 0.743206194988216,
+      "grad_norm": 3.3849794768674855,
+      "learning_rate": 3.2658886758270947e-06,
+      "loss": 2.9173,
+      "step": 3863
+    },
+    {
+      "epoch": 0.7433985859266029,
+      "grad_norm": 3.5266870971907025,
+      "learning_rate": 3.2612828037409116e-06,
+      "loss": 2.8811,
+      "step": 3864
+    },
+    {
+      "epoch": 0.7435909768649897,
+      "grad_norm": 3.272914132269951,
+      "learning_rate": 3.256679548894831e-06,
+      "loss": 3.2237,
+      "step": 3865
+    },
+    {
+      "epoch": 0.7437833678033765,
+      "grad_norm": 3.4342481517359933,
+      "learning_rate": 3.252078913076718e-06,
+      "loss": 2.6668,
+      "step": 3866
+    },
+    {
+      "epoch": 0.7439757587417632,
+      "grad_norm": 3.387462059504962,
+      "learning_rate": 3.247480898073395e-06,
+      "loss": 2.6393,
+      "step": 3867
+    },
+    {
+      "epoch": 0.7441681496801501,
+      "grad_norm": 3.3865254440657813,
+      "learning_rate": 3.242885505670681e-06,
+      "loss": 2.8145,
+      "step": 3868
+    },
+    {
+      "epoch": 0.7443605406185368,
+      "grad_norm": 3.2627326995371373,
+      "learning_rate": 3.238292737653379e-06,
+      "loss": 2.9189,
+      "step": 3869
+    },
+    {
+      "epoch": 0.7445529315569237,
+      "grad_norm": 3.3719912615470453,
+      "learning_rate": 3.233702595805258e-06,
+      "loss": 2.9979,
+      "step": 3870
+    },
+    {
+      "epoch": 0.7447453224953104,
+      "grad_norm": 3.661773307432398,
+      "learning_rate": 3.229115081909082e-06,
+      "loss": 3.0035,
+      "step": 3871
+    },
+    {
+      "epoch": 0.7449377134336973,
+      "grad_norm": 3.1728220700241305,
+      "learning_rate": 3.224530197746587e-06,
+      "loss": 2.7623,
+      "step": 3872
+    },
+    {
+      "epoch": 0.745130104372084,
+      "grad_norm": 3.483946739278625,
+      "learning_rate": 3.2199479450984892e-06,
+      "loss": 3.0739,
+      "step": 3873
+    },
+    {
+      "epoch": 0.7453224953104709,
+      "grad_norm": 3.7983209780840834,
+      "learning_rate": 3.2153683257444856e-06,
+      "loss": 3.034,
+      "step": 3874
+    },
+    {
+      "epoch": 0.7455148862488576,
+      "grad_norm": 3.490951743240452,
+      "learning_rate": 3.210791341463243e-06,
+      "loss": 2.7673,
+      "step": 3875
+    },
+    {
+      "epoch": 0.7457072771872445,
+      "grad_norm": 2.5774493898500324,
+      "learning_rate": 3.206216994032411e-06,
+      "loss": 2.5984,
+      "step": 3876
+    },
+    {
+      "epoch": 0.7458996681256312,
+      "grad_norm": 3.444571649528615,
+      "learning_rate": 3.2016452852286127e-06,
+      "loss": 2.918,
+      "step": 3877
+    },
+    {
+      "epoch": 0.7460920590640181,
+      "grad_norm": 5.528089435168836,
+      "learning_rate": 3.1970762168274495e-06,
+      "loss": 2.5919,
+      "step": 3878
+    },
+    {
+      "epoch": 0.7462844500024048,
+      "grad_norm": 4.040266015220688,
+      "learning_rate": 3.1925097906034962e-06,
+      "loss": 2.9198,
+      "step": 3879
+    },
+    {
+      "epoch": 0.7464768409407917,
+      "grad_norm": 3.692651533846078,
+      "learning_rate": 3.187946008330295e-06,
+      "loss": 2.8447,
+      "step": 3880
+    },
+    {
+      "epoch": 0.7466692318791784,
+      "grad_norm": 3.510154719647963,
+      "learning_rate": 3.1833848717803674e-06,
+      "loss": 2.954,
+      "step": 3881
+    },
+    {
+      "epoch": 0.7468616228175653,
+      "grad_norm": 3.661305362338092,
+      "learning_rate": 3.178826382725212e-06,
+      "loss": 3.0264,
+      "step": 3882
+    },
+    {
+      "epoch": 0.747054013755952,
+      "grad_norm": 3.361581866494499,
+      "learning_rate": 3.1742705429352827e-06,
+      "loss": 2.9668,
+      "step": 3883
+    },
+    {
+      "epoch": 0.7472464046943389,
+      "grad_norm": 3.4766127539580562,
+      "learning_rate": 3.1697173541800254e-06,
+      "loss": 3.0585,
+      "step": 3884
+    },
+    {
+      "epoch": 0.7474387956327257,
+      "grad_norm": 3.2300685779847917,
+      "learning_rate": 3.165166818227845e-06,
+      "loss": 2.6724,
+      "step": 3885
+    },
+    {
+      "epoch": 0.7476311865711125,
+      "grad_norm": 3.5955122135332527,
+      "learning_rate": 3.1606189368461117e-06,
+      "loss": 3.1503,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7478235775094993,
+      "grad_norm": 3.3009312442868346,
+      "learning_rate": 3.156073711801172e-06,
+      "loss": 2.938,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7480159684478861,
+      "grad_norm": 3.7283584073002443,
+      "learning_rate": 3.151531144858344e-06,
+      "loss": 2.8626,
+      "step": 3888
+    },
+    {
+      "epoch": 0.748208359386273,
+      "grad_norm": 3.4475422467665044,
+      "learning_rate": 3.146991237781899e-06,
+      "loss": 3.2175,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7484007503246597,
+      "grad_norm": 3.6205226004205002,
+      "learning_rate": 3.142453992335096e-06,
+      "loss": 2.9409,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7485931412630465,
+      "grad_norm": 3.5553591669355415,
+      "learning_rate": 3.137919410280139e-06,
+      "loss": 2.8753,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7487855322014333,
+      "grad_norm": 2.354154375170172,
+      "learning_rate": 3.1333874933782114e-06,
+      "loss": 2.6572,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7489779231398201,
+      "grad_norm": 3.5563940479750054,
+      "learning_rate": 3.128858243389461e-06,
+      "loss": 3.0607,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7491703140782069,
+      "grad_norm": 3.2280205069948744,
+      "learning_rate": 3.124331662072987e-06,
+      "loss": 2.9407,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7493627050165937,
+      "grad_norm": 3.5669122292215207,
+      "learning_rate": 3.119807751186872e-06,
+      "loss": 3.1654,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7495550959549805,
+      "grad_norm": 3.88790409900022,
+      "learning_rate": 3.1152865124881436e-06,
+      "loss": 2.8716,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7497474868933673,
+      "grad_norm": 3.4985380720616766,
+      "learning_rate": 3.110767947732801e-06,
+      "loss": 2.8108,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7499398778317541,
+      "grad_norm": 3.7133714584882576,
+      "learning_rate": 3.106252058675806e-06,
+      "loss": 3.1482,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7501322687701409,
+      "grad_norm": 3.2180024661013964,
+      "learning_rate": 3.101738847071072e-06,
+      "loss": 2.7213,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7503246597085277,
+      "grad_norm": 3.782031985771673,
+      "learning_rate": 3.097228314671481e-06,
+      "loss": 3.1622,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7505170506469145,
+      "grad_norm": 3.5560118772935327,
+      "learning_rate": 3.092720463228872e-06,
+      "loss": 2.9933,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7507094415853013,
+      "grad_norm": 3.445833437596797,
+      "learning_rate": 3.0882152944940423e-06,
+      "loss": 2.7963,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7509018325236881,
+      "grad_norm": 3.3794275820792947,
+      "learning_rate": 3.0837128102167514e-06,
+      "loss": 2.913,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7510942234620749,
+      "grad_norm": 3.764441279098627,
+      "learning_rate": 3.079213012145705e-06,
+      "loss": 2.8977,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7512866144004617,
+      "grad_norm": 3.428930689658525,
+      "learning_rate": 3.0747159020285766e-06,
+      "loss": 3.0842,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7514790053388486,
+      "grad_norm": 3.3954077074605986,
+      "learning_rate": 3.0702214816119925e-06,
+      "loss": 3.0084,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7516713962772353,
+      "grad_norm": 3.4773550836998046,
+      "learning_rate": 3.065729752641532e-06,
+      "loss": 2.902,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7518637872156222,
+      "grad_norm": 3.280914110430288,
+      "learning_rate": 3.0612407168617352e-06,
+      "loss": 2.8062,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7520561781540089,
+      "grad_norm": 3.3717618480233393,
+      "learning_rate": 3.0567543760160866e-06,
+      "loss": 3.0213,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7522485690923958,
+      "grad_norm": 3.461696562217283,
+      "learning_rate": 3.05227073184703e-06,
+      "loss": 2.7998,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7524409600307825,
+      "grad_norm": 3.543158664087604,
+      "learning_rate": 3.047789786095967e-06,
+      "loss": 2.7747,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7526333509691694,
+      "grad_norm": 3.2694921066467857,
+      "learning_rate": 3.0433115405032334e-06,
+      "loss": 2.6808,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7528257419075561,
+      "grad_norm": 3.4908210682408107,
+      "learning_rate": 3.0388359968081395e-06,
+      "loss": 3.14,
+      "step": 3913
+    },
+    {
+      "epoch": 0.753018132845943,
+      "grad_norm": 3.5265316070530908,
+      "learning_rate": 3.034363156748933e-06,
+      "loss": 3.2269,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7532105237843297,
+      "grad_norm": 3.538082166660351,
+      "learning_rate": 3.0298930220628086e-06,
+      "loss": 3.2014,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7534029147227166,
+      "grad_norm": 3.504611702996879,
+      "learning_rate": 3.025425594485919e-06,
+      "loss": 3.0084,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7535953056611033,
+      "grad_norm": 3.2471300129131344,
+      "learning_rate": 3.0209608757533626e-06,
+      "loss": 2.7073,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7537876965994902,
+      "grad_norm": 3.2854257730129075,
+      "learning_rate": 3.0164988675991768e-06,
+      "loss": 2.9402,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7539800875378769,
+      "grad_norm": 3.494342851013085,
+      "learning_rate": 3.0120395717563655e-06,
+      "loss": 2.9,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7541724784762638,
+      "grad_norm": 3.4481621837805125,
+      "learning_rate": 3.00758298995686e-06,
+      "loss": 2.7177,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7543648694146505,
+      "grad_norm": 3.111082629319539,
+      "learning_rate": 3.0031291239315473e-06,
+      "loss": 2.6863,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7545572603530374,
+      "grad_norm": 3.5359451042158345,
+      "learning_rate": 2.9986779754102613e-06,
+      "loss": 2.889,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7547496512914241,
+      "grad_norm": 3.400180861587929,
+      "learning_rate": 2.9942295461217698e-06,
+      "loss": 2.9318,
+      "step": 3923
+    },
+    {
+      "epoch": 0.754942042229811,
+      "grad_norm": 2.9889710332401953,
+      "learning_rate": 2.9897838377937947e-06,
+      "loss": 2.5184,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7551344331681977,
+      "grad_norm": 3.4098497567063024,
+      "learning_rate": 2.985340852152999e-06,
+      "loss": 2.8074,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7553268241065846,
+      "grad_norm": 3.330790582168974,
+      "learning_rate": 2.9809005909249866e-06,
+      "loss": 2.9029,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7555192150449714,
+      "grad_norm": 2.3588179031948218,
+      "learning_rate": 2.9764630558343064e-06,
+      "loss": 2.4571,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7557116059833582,
+      "grad_norm": 3.3937032889149523,
+      "learning_rate": 2.9720282486044407e-06,
+      "loss": 2.9729,
+      "step": 3928
+    },
+    {
+      "epoch": 0.755903996921745,
+      "grad_norm": 3.2438166414073715,
+      "learning_rate": 2.9675961709578194e-06,
+      "loss": 3.0243,
+      "step": 3929
+    },
+    {
+      "epoch": 0.7560963878601318,
+      "grad_norm": 3.555382219259726,
+      "learning_rate": 2.9631668246158105e-06,
+      "loss": 2.7654,
+      "step": 3930
+    },
+    {
+      "epoch": 0.7562887787985186,
+      "grad_norm": 3.469517550456272,
+      "learning_rate": 2.958740211298722e-06,
+      "loss": 2.7918,
+      "step": 3931
+    },
+    {
+      "epoch": 0.7564811697369054,
+      "grad_norm": 3.495512850858629,
+      "learning_rate": 2.9543163327258016e-06,
+      "loss": 3.0765,
+      "step": 3932
+    },
+    {
+      "epoch": 0.7566735606752922,
+      "grad_norm": 3.480057119586825,
+      "learning_rate": 2.949895190615227e-06,
+      "loss": 2.7491,
+      "step": 3933
+    },
+    {
+      "epoch": 0.756865951613679,
+      "grad_norm": 3.4102456519798863,
+      "learning_rate": 2.9454767866841225e-06,
+      "loss": 2.9126,
+      "step": 3934
+    },
+    {
+      "epoch": 0.7570583425520658,
+      "grad_norm": 2.3126044670485473,
+      "learning_rate": 2.941061122648545e-06,
+      "loss": 2.3828,
+      "step": 3935
+    },
+    {
+      "epoch": 0.7572507334904526,
+      "grad_norm": 3.646082914104923,
+      "learning_rate": 2.9366482002234874e-06,
+      "loss": 2.776,
+      "step": 3936
+    },
+    {
+      "epoch": 0.7574431244288394,
+      "grad_norm": 3.43092254875137,
+      "learning_rate": 2.932238021122877e-06,
+      "loss": 3.1702,
+      "step": 3937
+    },
+    {
+      "epoch": 0.7576355153672262,
+      "grad_norm": 3.748877137660498,
+      "learning_rate": 2.9278305870595814e-06,
+      "loss": 3.0241,
+      "step": 3938
+    },
+    {
+      "epoch": 0.757827906305613,
+      "grad_norm": 3.3088738729587037,
+      "learning_rate": 2.92342589974539e-06,
+      "loss": 3.0875,
+      "step": 3939
+    },
+    {
+      "epoch": 0.7580202972439998,
+      "grad_norm": 3.3080434382567008,
+      "learning_rate": 2.919023960891039e-06,
+      "loss": 2.9984,
+      "step": 3940
+    },
+    {
+      "epoch": 0.7582126881823866,
+      "grad_norm": 3.7581955831765397,
+      "learning_rate": 2.9146247722061806e-06,
+      "loss": 2.8886,
+      "step": 3941
+    },
+    {
+      "epoch": 0.7584050791207734,
+      "grad_norm": 3.6348499709728253,
+      "learning_rate": 2.910228335399419e-06,
+      "loss": 2.8481,
+      "step": 3942
+    },
+    {
+      "epoch": 0.7585974700591602,
+      "grad_norm": 3.4596663379510386,
+      "learning_rate": 2.90583465217828e-06,
+      "loss": 2.9775,
+      "step": 3943
+    },
+    {
+      "epoch": 0.758789860997547,
+      "grad_norm": 3.2722829188259066,
+      "learning_rate": 2.9014437242492133e-06,
+      "loss": 2.96,
+      "step": 3944
+    },
+    {
+      "epoch": 0.7589822519359338,
+      "grad_norm": 3.4268152300069885,
+      "learning_rate": 2.897055553317607e-06,
+      "loss": 3.1358,
+      "step": 3945
+    },
+    {
+      "epoch": 0.7591746428743206,
+      "grad_norm": 3.5517828007760603,
+      "learning_rate": 2.89267014108778e-06,
+      "loss": 3.0992,
+      "step": 3946
+    },
+    {
+      "epoch": 0.7593670338127074,
+      "grad_norm": 3.1520454721441467,
+      "learning_rate": 2.8882874892629654e-06,
+      "loss": 3.0634,
+      "step": 3947
+    },
+    {
+      "epoch": 0.7595594247510942,
+      "grad_norm": 2.328804522081028,
+      "learning_rate": 2.883907599545348e-06,
+      "loss": 2.5196,
+      "step": 3948
+    },
+    {
+      "epoch": 0.759751815689481,
+      "grad_norm": 3.4837205626759853,
+      "learning_rate": 2.8795304736360184e-06,
+      "loss": 2.9591,
+      "step": 3949
+    },
+    {
+      "epoch": 0.7599442066278679,
+      "grad_norm": 3.3906589162074967,
+      "learning_rate": 2.8751561132350025e-06,
+      "loss": 2.677,
+      "step": 3950
+    },
+    {
+      "epoch": 0.7601365975662546,
+      "grad_norm": 3.851292614631259,
+      "learning_rate": 2.8707845200412567e-06,
+      "loss": 3.0258,
+      "step": 3951
+    },
+    {
+      "epoch": 0.7603289885046415,
+      "grad_norm": 2.2622862526661787,
+      "learning_rate": 2.866415695752649e-06,
+      "loss": 2.523,
+      "step": 3952
+    },
+    {
+      "epoch": 0.7605213794430282,
+      "grad_norm": 3.4207025648484874,
+      "learning_rate": 2.862049642065986e-06,
+      "loss": 3.1439,
+      "step": 3953
+    },
+    {
+      "epoch": 0.7607137703814151,
+      "grad_norm": 3.3484255056842933,
+      "learning_rate": 2.857686360676991e-06,
+      "loss": 3.0789,
+      "step": 3954
+    },
+    {
+      "epoch": 0.7609061613198018,
+      "grad_norm": 3.469356191129252,
+      "learning_rate": 2.853325853280312e-06,
+      "loss": 3.0308,
+      "step": 3955
+    },
+    {
+      "epoch": 0.7610985522581887,
+      "grad_norm": 3.4627904775912626,
+      "learning_rate": 2.8489681215695242e-06,
+      "loss": 2.9439,
+      "step": 3956
+    },
+    {
+      "epoch": 0.7612909431965754,
+      "grad_norm": 3.424840450724874,
+      "learning_rate": 2.8446131672371136e-06,
+      "loss": 2.7007,
+      "step": 3957
+    },
+    {
+      "epoch": 0.7614833341349623,
+      "grad_norm": 3.686633147545663,
+      "learning_rate": 2.840260991974497e-06,
+      "loss": 2.8207,
+      "step": 3958
+    },
+    {
+      "epoch": 0.761675725073349,
+      "grad_norm": 3.478986903360685,
+      "learning_rate": 2.83591159747201e-06,
+      "loss": 2.9923,
+      "step": 3959
+    },
+    {
+      "epoch": 0.7618681160117359,
+      "grad_norm": 3.4092238620158595,
+      "learning_rate": 2.8315649854189066e-06,
+      "loss": 2.7892,
+      "step": 3960
+    },
+    {
+      "epoch": 0.7620605069501226,
+      "grad_norm": 3.7810803801674817,
+      "learning_rate": 2.8272211575033635e-06,
+      "loss": 3.144,
+      "step": 3961
+    },
+    {
+      "epoch": 0.7622528978885095,
+      "grad_norm": 3.609182982968076,
+      "learning_rate": 2.8228801154124687e-06,
+      "loss": 3.1234,
+      "step": 3962
+    },
+    {
+      "epoch": 0.7624452888268962,
+      "grad_norm": 3.4652830599771685,
+      "learning_rate": 2.8185418608322344e-06,
+      "loss": 3.0117,
+      "step": 3963
+    },
+    {
+      "epoch": 0.7626376797652831,
+      "grad_norm": 3.6560165217532252,
+      "learning_rate": 2.814206395447593e-06,
+      "loss": 2.7763,
+      "step": 3964
+    },
+    {
+      "epoch": 0.7628300707036698,
+      "grad_norm": 3.681010880911594,
+      "learning_rate": 2.80987372094238e-06,
+      "loss": 3.2069,
+      "step": 3965
+    },
+    {
+      "epoch": 0.7630224616420567,
+      "grad_norm": 3.504715445748165,
+      "learning_rate": 2.805543838999364e-06,
+      "loss": 3.3046,
+      "step": 3966
+    },
+    {
+      "epoch": 0.7632148525804434,
+      "grad_norm": 3.690889828701401,
+      "learning_rate": 2.801216751300223e-06,
+      "loss": 2.7931,
+      "step": 3967
+    },
+    {
+      "epoch": 0.7634072435188303,
+      "grad_norm": 3.9781591841713118,
+      "learning_rate": 2.796892459525541e-06,
+      "loss": 3.0231,
+      "step": 3968
+    },
+    {
+      "epoch": 0.763599634457217,
+      "grad_norm": 3.425979819517581,
+      "learning_rate": 2.7925709653548295e-06,
+      "loss": 2.8973,
+      "step": 3969
+    },
+    {
+      "epoch": 0.7637920253956039,
+      "grad_norm": 3.875049551171286,
+      "learning_rate": 2.788252270466497e-06,
+      "loss": 2.8003,
+      "step": 3970
+    },
+    {
+      "epoch": 0.7639844163339907,
+      "grad_norm": 2.3978158304638875,
+      "learning_rate": 2.783936376537886e-06,
+      "loss": 2.5317,
+      "step": 3971
+    },
+    {
+      "epoch": 0.7641768072723775,
+      "grad_norm": 3.338468073943056,
+      "learning_rate": 2.7796232852452378e-06,
+      "loss": 2.9776,
+      "step": 3972
+    },
+    {
+      "epoch": 0.7643691982107643,
+      "grad_norm": 3.073568127901513,
+      "learning_rate": 2.775312998263703e-06,
+      "loss": 2.9204,
+      "step": 3973
+    },
+    {
+      "epoch": 0.7645615891491511,
+      "grad_norm": 3.5747572034336397,
+      "learning_rate": 2.771005517267349e-06,
+      "loss": 2.8232,
+      "step": 3974
+    },
+    {
+      "epoch": 0.7647539800875379,
+      "grad_norm": 3.3471948290452898,
+      "learning_rate": 2.7667008439291552e-06,
+      "loss": 2.9676,
+      "step": 3975
+    },
+    {
+      "epoch": 0.7649463710259247,
+      "grad_norm": 3.622750545372289,
+      "learning_rate": 2.762398979920998e-06,
+      "loss": 3.1124,
+      "step": 3976
+    },
+    {
+      "epoch": 0.7651387619643115,
+      "grad_norm": 3.1926326453969387,
+      "learning_rate": 2.7580999269136854e-06,
+      "loss": 3.0902,
+      "step": 3977
+    },
+    {
+      "epoch": 0.7653311529026983,
+      "grad_norm": 3.452235151265305,
+      "learning_rate": 2.7538036865769093e-06,
+      "loss": 3.1408,
+      "step": 3978
+    },
+    {
+      "epoch": 0.7655235438410851,
+      "grad_norm": 3.452545071852768,
+      "learning_rate": 2.749510260579282e-06,
+      "loss": 3.2963,
+      "step": 3979
+    },
+    {
+      "epoch": 0.7657159347794719,
+      "grad_norm": 3.1789436314889703,
+      "learning_rate": 2.7452196505883265e-06,
+      "loss": 2.8971,
+      "step": 3980
+    },
+    {
+      "epoch": 0.7659083257178587,
+      "grad_norm": 3.6458710765799744,
+      "learning_rate": 2.7409318582704594e-06,
+      "loss": 2.8935,
+      "step": 3981
+    },
+    {
+      "epoch": 0.7661007166562455,
+      "grad_norm": 3.3456155656293975,
+      "learning_rate": 2.736646885291011e-06,
+      "loss": 2.9973,
+      "step": 3982
+    },
+    {
+      "epoch": 0.7662931075946323,
+      "grad_norm": 3.4953730263064506,
+      "learning_rate": 2.7323647333142176e-06,
+      "loss": 3.0403,
+      "step": 3983
+    },
+    {
+      "epoch": 0.7664854985330191,
+      "grad_norm": 3.4815767827099617,
+      "learning_rate": 2.728085404003217e-06,
+      "loss": 3.0857,
+      "step": 3984
+    },
+    {
+      "epoch": 0.7666778894714059,
+      "grad_norm": 3.3163685937485337,
+      "learning_rate": 2.7238088990200538e-06,
+      "loss": 2.6984,
+      "step": 3985
+    },
+    {
+      "epoch": 0.7668702804097927,
+      "grad_norm": 3.5539874214192686,
+      "learning_rate": 2.7195352200256675e-06,
+      "loss": 3.0885,
+      "step": 3986
+    },
+    {
+      "epoch": 0.7670626713481795,
+      "grad_norm": 3.7221038548972247,
+      "learning_rate": 2.7152643686799095e-06,
+      "loss": 3.0906,
+      "step": 3987
+    },
+    {
+      "epoch": 0.7672550622865663,
+      "grad_norm": 2.394053957050498,
+      "learning_rate": 2.710996346641528e-06,
+      "loss": 2.6083,
+      "step": 3988
+    },
+    {
+      "epoch": 0.7674474532249531,
+      "grad_norm": 3.5743338297208163,
+      "learning_rate": 2.706731155568175e-06,
+      "loss": 3.0764,
+      "step": 3989
+    },
+    {
+      "epoch": 0.7676398441633399,
+      "grad_norm": 3.469618767505044,
+      "learning_rate": 2.7024687971164032e-06,
+      "loss": 2.9981,
+      "step": 3990
+    },
+    {
+      "epoch": 0.7678322351017267,
+      "grad_norm": 3.4915268430689803,
+      "learning_rate": 2.698209272941659e-06,
+      "loss": 3.0941,
+      "step": 3991
+    },
+    {
+      "epoch": 0.7680246260401135,
+      "grad_norm": 3.4139056347612415,
+      "learning_rate": 2.693952584698294e-06,
+      "loss": 2.9858,
+      "step": 3992
+    },
+    {
+      "epoch": 0.7682170169785003,
+      "grad_norm": 3.6083893470509087,
+      "learning_rate": 2.689698734039561e-06,
+      "loss": 3.0954,
+      "step": 3993
+    },
+    {
+      "epoch": 0.7684094079168872,
+      "grad_norm": 3.317479472766701,
+      "learning_rate": 2.685447722617597e-06,
+      "loss": 3.057,
+      "step": 3994
+    },
+    {
+      "epoch": 0.7686017988552739,
+      "grad_norm": 3.629873544253391,
+      "learning_rate": 2.6811995520834543e-06,
+      "loss": 2.868,
+      "step": 3995
+    },
+    {
+      "epoch": 0.7687941897936608,
+      "grad_norm": 3.4607066657105694,
+      "learning_rate": 2.676954224087075e-06,
+      "loss": 2.9172,
+      "step": 3996
+    },
+    {
+      "epoch": 0.7689865807320475,
+      "grad_norm": 2.28650813836515,
+      "learning_rate": 2.6727117402772886e-06,
+      "loss": 2.4436,
+      "step": 3997
+    },
+    {
+      "epoch": 0.7691789716704344,
+      "grad_norm": 3.4238260981801947,
+      "learning_rate": 2.668472102301829e-06,
+      "loss": 2.9168,
+      "step": 3998
+    },
+    {
+      "epoch": 0.7693713626088211,
+      "grad_norm": 3.626650007449124,
+      "learning_rate": 2.664235311807327e-06,
+      "loss": 3.1391,
+      "step": 3999
+    },
+    {
+      "epoch": 0.769563753547208,
+      "grad_norm": 3.552919114081789,
+      "learning_rate": 2.6600013704392946e-06,
+      "loss": 3.0866,
+      "step": 4000
+    },
+    {
+      "epoch": 0.7697561444855947,
+      "grad_norm": 3.767578947951143,
+      "learning_rate": 2.655770279842157e-06,
+      "loss": 3.1421,
+      "step": 4001
+    },
+    {
+      "epoch": 0.7699485354239816,
+      "grad_norm": 3.3293679638807823,
+      "learning_rate": 2.651542041659211e-06,
+      "loss": 2.9329,
+      "step": 4002
+    },
+    {
+      "epoch": 0.7701409263623683,
+      "grad_norm": 3.4217398109795742,
+      "learning_rate": 2.6473166575326603e-06,
+      "loss": 2.9873,
+      "step": 4003
+    },
+    {
+      "epoch": 0.7703333173007552,
+      "grad_norm": 3.5383934396794237,
+      "learning_rate": 2.6430941291035984e-06,
+      "loss": 3.008,
+      "step": 4004
+    },
+    {
+      "epoch": 0.7705257082391419,
+      "grad_norm": 3.842602157522853,
+      "learning_rate": 2.6388744580119975e-06,
+      "loss": 3.0388,
+      "step": 4005
+    },
+    {
+      "epoch": 0.7707180991775288,
+      "grad_norm": 3.2438370782096593,
+      "learning_rate": 2.6346576458967397e-06,
+      "loss": 3.0173,
+      "step": 4006
+    },
+    {
+      "epoch": 0.7709104901159155,
+      "grad_norm": 3.608471252221908,
+      "learning_rate": 2.630443694395579e-06,
+      "loss": 3.0113,
+      "step": 4007
+    },
+    {
+      "epoch": 0.7711028810543024,
+      "grad_norm": 3.5686323935258137,
+      "learning_rate": 2.6262326051451683e-06,
+      "loss": 3.0306,
+      "step": 4008
+    },
+    {
+      "epoch": 0.7712952719926891,
+      "grad_norm": 3.3822518484068804,
+      "learning_rate": 2.6220243797810483e-06,
+      "loss": 3.2071,
+      "step": 4009
+    },
+    {
+      "epoch": 0.771487662931076,
+      "grad_norm": 3.489847857681779,
+      "learning_rate": 2.6178190199376394e-06,
+      "loss": 3.1544,
+      "step": 4010
+    },
+    {
+      "epoch": 0.7716800538694627,
+      "grad_norm": 2.304930361336288,
+      "learning_rate": 2.6136165272482596e-06,
+      "loss": 2.7221,
+      "step": 4011
+    },
+    {
+      "epoch": 0.7718724448078496,
+      "grad_norm": 3.5186128216411054,
+      "learning_rate": 2.6094169033451066e-06,
+      "loss": 3.218,
+      "step": 4012
+    },
+    {
+      "epoch": 0.7720648357462363,
+      "grad_norm": 3.626181541091798,
+      "learning_rate": 2.605220149859267e-06,
+      "loss": 2.9387,
+      "step": 4013
+    },
+    {
+      "epoch": 0.7722572266846232,
+      "grad_norm": 3.2926100443772994,
+      "learning_rate": 2.6010262684207134e-06,
+      "loss": 2.9708,
+      "step": 4014
+    },
+    {
+      "epoch": 0.77244961762301,
+      "grad_norm": 3.562630049946964,
+      "learning_rate": 2.596835260658297e-06,
+      "loss": 2.9292,
+      "step": 4015
+    },
+    {
+      "epoch": 0.7726420085613968,
+      "grad_norm": 3.180054747186241,
+      "learning_rate": 2.5926471281997577e-06,
+      "loss": 2.8185,
+      "step": 4016
+    },
+    {
+      "epoch": 0.7728343994997836,
+      "grad_norm": 3.1045398987444712,
+      "learning_rate": 2.588461872671719e-06,
+      "loss": 2.5555,
+      "step": 4017
+    },
+    {
+      "epoch": 0.7730267904381704,
+      "grad_norm": 3.8619409662049367,
+      "learning_rate": 2.5842794956996863e-06,
+      "loss": 3.0746,
+      "step": 4018
+    },
+    {
+      "epoch": 0.7732191813765572,
+      "grad_norm": 3.4815845786703674,
+      "learning_rate": 2.580099998908049e-06,
+      "loss": 2.8301,
+      "step": 4019
+    },
+    {
+      "epoch": 0.773411572314944,
+      "grad_norm": 3.625498358211811,
+      "learning_rate": 2.575923383920069e-06,
+      "loss": 2.9163,
+      "step": 4020
+    },
+    {
+      "epoch": 0.7736039632533308,
+      "grad_norm": 3.711289497860947,
+      "learning_rate": 2.5717496523578998e-06,
+      "loss": 2.8617,
+      "step": 4021
+    },
+    {
+      "epoch": 0.7737963541917176,
+      "grad_norm": 3.327363731702211,
+      "learning_rate": 2.5675788058425723e-06,
+      "loss": 3.1584,
+      "step": 4022
+    },
+    {
+      "epoch": 0.7739887451301044,
+      "grad_norm": 2.165014306254779,
+      "learning_rate": 2.563410845993988e-06,
+      "loss": 2.5996,
+      "step": 4023
+    },
+    {
+      "epoch": 0.7741811360684911,
+      "grad_norm": 3.408443753573168,
+      "learning_rate": 2.5592457744309405e-06,
+      "loss": 2.8451,
+      "step": 4024
+    },
+    {
+      "epoch": 0.774373527006878,
+      "grad_norm": 3.3988703806893157,
+      "learning_rate": 2.5550835927710982e-06,
+      "loss": 3.1277,
+      "step": 4025
+    },
+    {
+      "epoch": 0.7745659179452647,
+      "grad_norm": 3.2260271593954104,
+      "learning_rate": 2.5509243026309983e-06,
+      "loss": 2.8971,
+      "step": 4026
+    },
+    {
+      "epoch": 0.7747583088836516,
+      "grad_norm": 3.5454220043174436,
+      "learning_rate": 2.546767905626063e-06,
+      "loss": 2.9303,
+      "step": 4027
+    },
+    {
+      "epoch": 0.7749506998220383,
+      "grad_norm": 3.391556590858669,
+      "learning_rate": 2.5426144033705937e-06,
+      "loss": 2.95,
+      "step": 4028
+    },
+    {
+      "epoch": 0.7751430907604252,
+      "grad_norm": 3.6173977897091585,
+      "learning_rate": 2.5384637974777513e-06,
+      "loss": 3.018,
+      "step": 4029
+    },
+    {
+      "epoch": 0.775335481698812,
+      "grad_norm": 3.523408811864249,
+      "learning_rate": 2.5343160895595977e-06,
+      "loss": 2.9729,
+      "step": 4030
+    },
+    {
+      "epoch": 0.7755278726371988,
+      "grad_norm": 3.417266456395185,
+      "learning_rate": 2.530171281227044e-06,
+      "loss": 3.0477,
+      "step": 4031
+    },
+    {
+      "epoch": 0.7757202635755855,
+      "grad_norm": 3.7101588740589957,
+      "learning_rate": 2.52602937408989e-06,
+      "loss": 3.1335,
+      "step": 4032
+    },
+    {
+      "epoch": 0.7759126545139724,
+      "grad_norm": 3.519415944658491,
+      "learning_rate": 2.5218903697568075e-06,
+      "loss": 3.0881,
+      "step": 4033
+    },
+    {
+      "epoch": 0.7761050454523591,
+      "grad_norm": 3.392651755364916,
+      "learning_rate": 2.517754269835332e-06,
+      "loss": 2.7876,
+      "step": 4034
+    },
+    {
+      "epoch": 0.776297436390746,
+      "grad_norm": 3.3154843616299994,
+      "learning_rate": 2.5136210759318814e-06,
+      "loss": 2.872,
+      "step": 4035
+    },
+    {
+      "epoch": 0.7764898273291327,
+      "grad_norm": 3.310881050391088,
+      "learning_rate": 2.5094907896517383e-06,
+      "loss": 3.0939,
+      "step": 4036
+    },
+    {
+      "epoch": 0.7766822182675196,
+      "grad_norm": 3.5338953939212043,
+      "learning_rate": 2.505363412599059e-06,
+      "loss": 2.9138,
+      "step": 4037
+    },
+    {
+      "epoch": 0.7768746092059065,
+      "grad_norm": 3.5338013709195826,
+      "learning_rate": 2.5012389463768737e-06,
+      "loss": 2.6494,
+      "step": 4038
+    },
+    {
+      "epoch": 0.7770670001442932,
+      "grad_norm": 3.3339750000832264,
+      "learning_rate": 2.4971173925870694e-06,
+      "loss": 3.0342,
+      "step": 4039
+    },
+    {
+      "epoch": 0.77725939108268,
+      "grad_norm": 3.3985174369855717,
+      "learning_rate": 2.4929987528304144e-06,
+      "loss": 3.0002,
+      "step": 4040
+    },
+    {
+      "epoch": 0.7774517820210668,
+      "grad_norm": 3.396681881774369,
+      "learning_rate": 2.4888830287065414e-06,
+      "loss": 2.7097,
+      "step": 4041
+    },
+    {
+      "epoch": 0.7776441729594537,
+      "grad_norm": 3.4224379075458464,
+      "learning_rate": 2.4847702218139493e-06,
+      "loss": 2.8049,
+      "step": 4042
+    },
+    {
+      "epoch": 0.7778365638978404,
+      "grad_norm": 3.4924797118636692,
+      "learning_rate": 2.480660333750007e-06,
+      "loss": 2.9919,
+      "step": 4043
+    },
+    {
+      "epoch": 0.7780289548362272,
+      "grad_norm": 3.3790613852538685,
+      "learning_rate": 2.476553366110944e-06,
+      "loss": 2.9564,
+      "step": 4044
+    },
+    {
+      "epoch": 0.778221345774614,
+      "grad_norm": 3.4451208444379833,
+      "learning_rate": 2.4724493204918598e-06,
+      "loss": 2.9099,
+      "step": 4045
+    },
+    {
+      "epoch": 0.7784137367130008,
+      "grad_norm": 3.574049909499038,
+      "learning_rate": 2.4683481984867207e-06,
+      "loss": 2.9909,
+      "step": 4046
+    },
+    {
+      "epoch": 0.7786061276513876,
+      "grad_norm": 3.3700230770398645,
+      "learning_rate": 2.4642500016883532e-06,
+      "loss": 2.6162,
+      "step": 4047
+    },
+    {
+      "epoch": 0.7787985185897744,
+      "grad_norm": 3.532185810410236,
+      "learning_rate": 2.4601547316884544e-06,
+      "loss": 2.7972,
+      "step": 4048
+    },
+    {
+      "epoch": 0.7789909095281612,
+      "grad_norm": 3.8012440629320325,
+      "learning_rate": 2.4560623900775728e-06,
+      "loss": 3.2367,
+      "step": 4049
+    },
+    {
+      "epoch": 0.779183300466548,
+      "grad_norm": 3.438066449289783,
+      "learning_rate": 2.45197297844513e-06,
+      "loss": 2.9525,
+      "step": 4050
+    },
+    {
+      "epoch": 0.7793756914049348,
+      "grad_norm": 3.2689860315959978,
+      "learning_rate": 2.447886498379409e-06,
+      "loss": 2.6533,
+      "step": 4051
+    },
+    {
+      "epoch": 0.7795680823433216,
+      "grad_norm": 3.9432255013377175,
+      "learning_rate": 2.4438029514675444e-06,
+      "loss": 2.9633,
+      "step": 4052
+    },
+    {
+      "epoch": 0.7797604732817084,
+      "grad_norm": 3.543325636450491,
+      "learning_rate": 2.439722339295545e-06,
+      "loss": 2.9367,
+      "step": 4053
+    },
+    {
+      "epoch": 0.7799528642200952,
+      "grad_norm": 3.6522356346804408,
+      "learning_rate": 2.4356446634482756e-06,
+      "loss": 2.737,
+      "step": 4054
+    },
+    {
+      "epoch": 0.780145255158482,
+      "grad_norm": 3.6141770704250287,
+      "learning_rate": 2.4315699255094516e-06,
+      "loss": 2.9751,
+      "step": 4055
+    },
+    {
+      "epoch": 0.7803376460968688,
+      "grad_norm": 3.593790106297318,
+      "learning_rate": 2.427498127061658e-06,
+      "loss": 3.1623,
+      "step": 4056
+    },
+    {
+      "epoch": 0.7805300370352556,
+      "grad_norm": 2.328547040445898,
+      "learning_rate": 2.423429269686336e-06,
+      "loss": 2.5917,
+      "step": 4057
+    },
+    {
+      "epoch": 0.7807224279736424,
+      "grad_norm": 3.4687648352398575,
+      "learning_rate": 2.4193633549637765e-06,
+      "loss": 2.7956,
+      "step": 4058
+    },
+    {
+      "epoch": 0.7809148189120293,
+      "grad_norm": 2.4507177911229854,
+      "learning_rate": 2.4153003844731425e-06,
+      "loss": 2.5623,
+      "step": 4059
+    },
+    {
+      "epoch": 0.781107209850416,
+      "grad_norm": 3.2965751341476395,
+      "learning_rate": 2.411240359792438e-06,
+      "loss": 3.0291,
+      "step": 4060
+    },
+    {
+      "epoch": 0.7812996007888029,
+      "grad_norm": 3.7227867712299685,
+      "learning_rate": 2.407183282498534e-06,
+      "loss": 3.0271,
+      "step": 4061
+    },
+    {
+      "epoch": 0.7814919917271896,
+      "grad_norm": 3.4618753139357783,
+      "learning_rate": 2.403129154167153e-06,
+      "loss": 3.0154,
+      "step": 4062
+    },
+    {
+      "epoch": 0.7816843826655765,
+      "grad_norm": 3.545809964116786,
+      "learning_rate": 2.3990779763728666e-06,
+      "loss": 3.1173,
+      "step": 4063
+    },
+    {
+      "epoch": 0.7818767736039632,
+      "grad_norm": 3.663256551382548,
+      "learning_rate": 2.3950297506891084e-06,
+      "loss": 2.9976,
+      "step": 4064
+    },
+    {
+      "epoch": 0.7820691645423501,
+      "grad_norm": 3.5908308970380176,
+      "learning_rate": 2.390984478688164e-06,
+      "loss": 2.9781,
+      "step": 4065
+    },
+    {
+      "epoch": 0.7822615554807368,
+      "grad_norm": 3.4292267464915054,
+      "learning_rate": 2.386942161941169e-06,
+      "loss": 3.1255,
+      "step": 4066
+    },
+    {
+      "epoch": 0.7824539464191237,
+      "grad_norm": 3.2984279603381985,
+      "learning_rate": 2.3829028020181154e-06,
+      "loss": 2.7311,
+      "step": 4067
+    },
+    {
+      "epoch": 0.7826463373575104,
+      "grad_norm": 3.34380380452495,
+      "learning_rate": 2.3788664004878405e-06,
+      "loss": 2.986,
+      "step": 4068
+    },
+    {
+      "epoch": 0.7828387282958973,
+      "grad_norm": 3.348037678298487,
+      "learning_rate": 2.374832958918035e-06,
+      "loss": 2.7614,
+      "step": 4069
+    },
+    {
+      "epoch": 0.783031119234284,
+      "grad_norm": 3.360954061053967,
+      "learning_rate": 2.3708024788752448e-06,
+      "loss": 2.9937,
+      "step": 4070
+    },
+    {
+      "epoch": 0.7832235101726709,
+      "grad_norm": 3.523595592034383,
+      "learning_rate": 2.3667749619248614e-06,
+      "loss": 2.7754,
+      "step": 4071
+    },
+    {
+      "epoch": 0.7834159011110576,
+      "grad_norm": 3.314545154462919,
+      "learning_rate": 2.3627504096311273e-06,
+      "loss": 2.9747,
+      "step": 4072
+    },
+    {
+      "epoch": 0.7836082920494445,
+      "grad_norm": 3.80029347666686,
+      "learning_rate": 2.3587288235571258e-06,
+      "loss": 2.9679,
+      "step": 4073
+    },
+    {
+      "epoch": 0.7838006829878312,
+      "grad_norm": 3.6094026818040756,
+      "learning_rate": 2.354710205264801e-06,
+      "loss": 3.0729,
+      "step": 4074
+    },
+    {
+      "epoch": 0.7839930739262181,
+      "grad_norm": 3.444122400885414,
+      "learning_rate": 2.350694556314934e-06,
+      "loss": 2.6979,
+      "step": 4075
+    },
+    {
+      "epoch": 0.7841854648646048,
+      "grad_norm": 3.304348281155039,
+      "learning_rate": 2.3466818782671597e-06,
+      "loss": 3.0106,
+      "step": 4076
+    },
+    {
+      "epoch": 0.7843778558029917,
+      "grad_norm": 3.3628069469409145,
+      "learning_rate": 2.3426721726799573e-06,
+      "loss": 2.5985,
+      "step": 4077
+    },
+    {
+      "epoch": 0.7845702467413784,
+      "grad_norm": 3.5012292500859274,
+      "learning_rate": 2.3386654411106446e-06,
+      "loss": 2.9014,
+      "step": 4078
+    },
+    {
+      "epoch": 0.7847626376797653,
+      "grad_norm": 3.7626212913384953,
+      "learning_rate": 2.3346616851153935e-06,
+      "loss": 3.1817,
+      "step": 4079
+    },
+    {
+      "epoch": 0.7849550286181521,
+      "grad_norm": 3.612019183275409,
+      "learning_rate": 2.330660906249218e-06,
+      "loss": 3.1392,
+      "step": 4080
+    },
+    {
+      "epoch": 0.7851474195565389,
+      "grad_norm": 3.535939082632351,
+      "learning_rate": 2.3266631060659685e-06,
+      "loss": 2.7837,
+      "step": 4081
+    },
+    {
+      "epoch": 0.7853398104949257,
+      "grad_norm": 3.5172587835447375,
+      "learning_rate": 2.32266828611835e-06,
+      "loss": 2.9427,
+      "step": 4082
+    },
+    {
+      "epoch": 0.7855322014333125,
+      "grad_norm": 3.6512484274328325,
+      "learning_rate": 2.318676447957907e-06,
+      "loss": 2.9284,
+      "step": 4083
+    },
+    {
+      "epoch": 0.7857245923716993,
+      "grad_norm": 3.213343924866923,
+      "learning_rate": 2.3146875931350165e-06,
+      "loss": 3.0496,
+      "step": 4084
+    },
+    {
+      "epoch": 0.7859169833100861,
+      "grad_norm": 3.406929514174529,
+      "learning_rate": 2.310701723198908e-06,
+      "loss": 2.6821,
+      "step": 4085
+    },
+    {
+      "epoch": 0.7861093742484729,
+      "grad_norm": 3.507944572012812,
+      "learning_rate": 2.3067188396976482e-06,
+      "loss": 2.9075,
+      "step": 4086
+    },
+    {
+      "epoch": 0.7863017651868597,
+      "grad_norm": 3.703087645076483,
+      "learning_rate": 2.3027389441781368e-06,
+      "loss": 3.1477,
+      "step": 4087
+    },
+    {
+      "epoch": 0.7864941561252465,
+      "grad_norm": 3.870926225802243,
+      "learning_rate": 2.2987620381861288e-06,
+      "loss": 2.7948,
+      "step": 4088
+    },
+    {
+      "epoch": 0.7866865470636333,
+      "grad_norm": 3.543287500366523,
+      "learning_rate": 2.2947881232662007e-06,
+      "loss": 2.9182,
+      "step": 4089
+    },
+    {
+      "epoch": 0.7868789380020201,
+      "grad_norm": 3.4746972477783356,
+      "learning_rate": 2.290817200961779e-06,
+      "loss": 2.8827,
+      "step": 4090
+    },
+    {
+      "epoch": 0.7870713289404069,
+      "grad_norm": 3.5995483869962275,
+      "learning_rate": 2.286849272815126e-06,
+      "loss": 2.6038,
+      "step": 4091
+    },
+    {
+      "epoch": 0.7872637198787937,
+      "grad_norm": 3.3405495839947386,
+      "learning_rate": 2.2828843403673338e-06,
+      "loss": 3.0012,
+      "step": 4092
+    },
+    {
+      "epoch": 0.7874561108171805,
+      "grad_norm": 3.2392096250140074,
+      "learning_rate": 2.2789224051583403e-06,
+      "loss": 2.8616,
+      "step": 4093
+    },
+    {
+      "epoch": 0.7876485017555673,
+      "grad_norm": 3.319906164463606,
+      "learning_rate": 2.274963468726914e-06,
+      "loss": 2.9369,
+      "step": 4094
+    },
+    {
+      "epoch": 0.7878408926939541,
+      "grad_norm": 3.571805602687379,
+      "learning_rate": 2.2710075326106618e-06,
+      "loss": 2.8092,
+      "step": 4095
+    },
+    {
+      "epoch": 0.7880332836323409,
+      "grad_norm": 3.935259620204544,
+      "learning_rate": 2.2670545983460245e-06,
+      "loss": 2.8446,
+      "step": 4096
+    },
+    {
+      "epoch": 0.7882256745707277,
+      "grad_norm": 3.5674723685499705,
+      "learning_rate": 2.263104667468272e-06,
+      "loss": 3.061,
+      "step": 4097
+    },
+    {
+      "epoch": 0.7884180655091145,
+      "grad_norm": 3.1964727924771803,
+      "learning_rate": 2.259157741511515e-06,
+      "loss": 2.6905,
+      "step": 4098
+    },
+    {
+      "epoch": 0.7886104564475013,
+      "grad_norm": 3.343701787428628,
+      "learning_rate": 2.2552138220086927e-06,
+      "loss": 2.9983,
+      "step": 4099
+    },
+    {
+      "epoch": 0.7888028473858881,
+      "grad_norm": 3.541160713684125,
+      "learning_rate": 2.2512729104915787e-06,
+      "loss": 2.6647,
+      "step": 4100
+    },
+    {
+      "epoch": 0.7889952383242749,
+      "grad_norm": 3.3812360180159526,
+      "learning_rate": 2.2473350084907806e-06,
+      "loss": 2.7751,
+      "step": 4101
+    },
+    {
+      "epoch": 0.7891876292626617,
+      "grad_norm": 3.3815738369078683,
+      "learning_rate": 2.243400117535729e-06,
+      "loss": 3.0204,
+      "step": 4102
+    },
+    {
+      "epoch": 0.7893800202010486,
+      "grad_norm": 2.3861151649781083,
+      "learning_rate": 2.2394682391546928e-06,
+      "loss": 2.4876,
+      "step": 4103
+    },
+    {
+      "epoch": 0.7895724111394353,
+      "grad_norm": 3.712144089429207,
+      "learning_rate": 2.2355393748747702e-06,
+      "loss": 2.7488,
+      "step": 4104
+    },
+    {
+      "epoch": 0.7897648020778222,
+      "grad_norm": 3.6683686941655567,
+      "learning_rate": 2.2316135262218787e-06,
+      "loss": 2.9788,
+      "step": 4105
+    },
+    {
+      "epoch": 0.7899571930162089,
+      "grad_norm": 2.23028981213147,
+      "learning_rate": 2.227690694720784e-06,
+      "loss": 2.4404,
+      "step": 4106
+    },
+    {
+      "epoch": 0.7901495839545958,
+      "grad_norm": 3.590974619061298,
+      "learning_rate": 2.223770881895061e-06,
+      "loss": 3.1223,
+      "step": 4107
+    },
+    {
+      "epoch": 0.7903419748929825,
+      "grad_norm": 3.2920301225424056,
+      "learning_rate": 2.2198540892671215e-06,
+      "loss": 2.9191,
+      "step": 4108
+    },
+    {
+      "epoch": 0.7905343658313694,
+      "grad_norm": 3.380355337858891,
+      "learning_rate": 2.215940318358206e-06,
+      "loss": 2.8961,
+      "step": 4109
+    },
+    {
+      "epoch": 0.7907267567697561,
+      "grad_norm": 3.4863096296390275,
+      "learning_rate": 2.2120295706883698e-06,
+      "loss": 2.7907,
+      "step": 4110
+    },
+    {
+      "epoch": 0.790919147708143,
+      "grad_norm": 3.4369997076461285,
+      "learning_rate": 2.2081218477765097e-06,
+      "loss": 2.7647,
+      "step": 4111
+    },
+    {
+      "epoch": 0.7911115386465297,
+      "grad_norm": 3.4895895958664562,
+      "learning_rate": 2.204217151140342e-06,
+      "loss": 2.9532,
+      "step": 4112
+    },
+    {
+      "epoch": 0.7913039295849166,
+      "grad_norm": 3.6727639516865875,
+      "learning_rate": 2.200315482296398e-06,
+      "loss": 2.9155,
+      "step": 4113
+    },
+    {
+      "epoch": 0.7914963205233033,
+      "grad_norm": 3.291929942053213,
+      "learning_rate": 2.1964168427600462e-06,
+      "loss": 2.9108,
+      "step": 4114
+    },
+    {
+      "epoch": 0.7916887114616902,
+      "grad_norm": 3.5295912669054763,
+      "learning_rate": 2.1925212340454737e-06,
+      "loss": 2.9903,
+      "step": 4115
+    },
+    {
+      "epoch": 0.7918811024000769,
+      "grad_norm": 3.4204232338569387,
+      "learning_rate": 2.1886286576656834e-06,
+      "loss": 2.8507,
+      "step": 4116
+    },
+    {
+      "epoch": 0.7920734933384638,
+      "grad_norm": 3.585155420260069,
+      "learning_rate": 2.184739115132517e-06,
+      "loss": 2.7722,
+      "step": 4117
+    },
+    {
+      "epoch": 0.7922658842768505,
+      "grad_norm": 3.315673572154263,
+      "learning_rate": 2.1808526079566215e-06,
+      "loss": 2.6865,
+      "step": 4118
+    },
+    {
+      "epoch": 0.7924582752152374,
+      "grad_norm": 3.5611416971391,
+      "learning_rate": 2.1769691376474722e-06,
+      "loss": 3.1528,
+      "step": 4119
+    },
+    {
+      "epoch": 0.7926506661536241,
+      "grad_norm": 3.5698674611044257,
+      "learning_rate": 2.1730887057133678e-06,
+      "loss": 3.2416,
+      "step": 4120
+    },
+    {
+      "epoch": 0.792843057092011,
+      "grad_norm": 3.586142639546891,
+      "learning_rate": 2.1692113136614177e-06,
+      "loss": 3.2559,
+      "step": 4121
+    },
+    {
+      "epoch": 0.7930354480303977,
+      "grad_norm": 3.534051220853263,
+      "learning_rate": 2.1653369629975595e-06,
+      "loss": 2.7472,
+      "step": 4122
+    },
+    {
+      "epoch": 0.7932278389687846,
+      "grad_norm": 3.4313073858250465,
+      "learning_rate": 2.1614656552265457e-06,
+      "loss": 3.1796,
+      "step": 4123
+    },
+    {
+      "epoch": 0.7934202299071714,
+      "grad_norm": 3.5613182328227913,
+      "learning_rate": 2.157597391851949e-06,
+      "loss": 3.0339,
+      "step": 4124
+    },
+    {
+      "epoch": 0.7936126208455582,
+      "grad_norm": 3.61387151118246,
+      "learning_rate": 2.1537321743761587e-06,
+      "loss": 3.088,
+      "step": 4125
+    },
+    {
+      "epoch": 0.793805011783945,
+      "grad_norm": 3.3615025656897632,
+      "learning_rate": 2.1498700043003773e-06,
+      "loss": 2.9608,
+      "step": 4126
+    },
+    {
+      "epoch": 0.7939974027223318,
+      "grad_norm": 3.4199142458249367,
+      "learning_rate": 2.1460108831246295e-06,
+      "loss": 3.0516,
+      "step": 4127
+    },
+    {
+      "epoch": 0.7941897936607186,
+      "grad_norm": 3.361792690535717,
+      "learning_rate": 2.142154812347753e-06,
+      "loss": 3.1628,
+      "step": 4128
+    },
+    {
+      "epoch": 0.7943821845991054,
+      "grad_norm": 3.510169533872196,
+      "learning_rate": 2.1383017934674012e-06,
+      "loss": 2.9325,
+      "step": 4129
+    },
+    {
+      "epoch": 0.7945745755374922,
+      "grad_norm": 3.5018024093312206,
+      "learning_rate": 2.1344518279800454e-06,
+      "loss": 3.0986,
+      "step": 4130
+    },
+    {
+      "epoch": 0.794766966475879,
+      "grad_norm": 3.2106533863321016,
+      "learning_rate": 2.130604917380962e-06,
+      "loss": 2.8102,
+      "step": 4131
+    },
+    {
+      "epoch": 0.7949593574142658,
+      "grad_norm": 3.333796250812083,
+      "learning_rate": 2.12676106316425e-06,
+      "loss": 2.9027,
+      "step": 4132
+    },
+    {
+      "epoch": 0.7951517483526526,
+      "grad_norm": 2.251500743016271,
+      "learning_rate": 2.1229202668228197e-06,
+      "loss": 2.2838,
+      "step": 4133
+    },
+    {
+      "epoch": 0.7953441392910394,
+      "grad_norm": 3.618970364178117,
+      "learning_rate": 2.1190825298483855e-06,
+      "loss": 2.9725,
+      "step": 4134
+    },
+    {
+      "epoch": 0.7955365302294262,
+      "grad_norm": 2.3484768317219347,
+      "learning_rate": 2.115247853731488e-06,
+      "loss": 2.5419,
+      "step": 4135
+    },
+    {
+      "epoch": 0.795728921167813,
+      "grad_norm": 3.6244208202496644,
+      "learning_rate": 2.11141623996147e-06,
+      "loss": 2.8243,
+      "step": 4136
+    },
+    {
+      "epoch": 0.7959213121061998,
+      "grad_norm": 3.170618768857132,
+      "learning_rate": 2.107587690026481e-06,
+      "loss": 3.0556,
+      "step": 4137
+    },
+    {
+      "epoch": 0.7961137030445866,
+      "grad_norm": 3.4428743020730264,
+      "learning_rate": 2.103762205413493e-06,
+      "loss": 3.0442,
+      "step": 4138
+    },
+    {
+      "epoch": 0.7963060939829734,
+      "grad_norm": 3.5712055985992386,
+      "learning_rate": 2.0999397876082726e-06,
+      "loss": 2.9676,
+      "step": 4139
+    },
+    {
+      "epoch": 0.7964984849213602,
+      "grad_norm": 3.7436450307631657,
+      "learning_rate": 2.096120438095404e-06,
+      "loss": 2.9045,
+      "step": 4140
+    },
+    {
+      "epoch": 0.796690875859747,
+      "grad_norm": 3.136947851061208,
+      "learning_rate": 2.092304158358286e-06,
+      "loss": 2.8631,
+      "step": 4141
+    },
+    {
+      "epoch": 0.7968832667981338,
+      "grad_norm": 3.330431682881438,
+      "learning_rate": 2.0884909498791106e-06,
+      "loss": 2.9655,
+      "step": 4142
+    },
+    {
+      "epoch": 0.7970756577365206,
+      "grad_norm": 3.774710773306019,
+      "learning_rate": 2.0846808141388852e-06,
+      "loss": 2.8566,
+      "step": 4143
+    },
+    {
+      "epoch": 0.7972680486749074,
+      "grad_norm": 3.3354116887794634,
+      "learning_rate": 2.080873752617426e-06,
+      "loss": 3.1185,
+      "step": 4144
+    },
+    {
+      "epoch": 0.7974604396132942,
+      "grad_norm": 3.5410764864232966,
+      "learning_rate": 2.0770697667933436e-06,
+      "loss": 3.1533,
+      "step": 4145
+    },
+    {
+      "epoch": 0.797652830551681,
+      "grad_norm": 3.5413093891037692,
+      "learning_rate": 2.073268858144074e-06,
+      "loss": 3.109,
+      "step": 4146
+    },
+    {
+      "epoch": 0.7978452214900679,
+      "grad_norm": 3.0543007840530754,
+      "learning_rate": 2.0694710281458372e-06,
+      "loss": 2.9398,
+      "step": 4147
+    },
+    {
+      "epoch": 0.7980376124284546,
+      "grad_norm": 3.534370756710516,
+      "learning_rate": 2.0656762782736693e-06,
+      "loss": 2.9865,
+      "step": 4148
+    },
+    {
+      "epoch": 0.7982300033668415,
+      "grad_norm": 3.4431229479719634,
+      "learning_rate": 2.061884610001411e-06,
+      "loss": 3.0415,
+      "step": 4149
+    },
+    {
+      "epoch": 0.7984223943052282,
+      "grad_norm": 3.304197670257059,
+      "learning_rate": 2.0580960248016966e-06,
+      "loss": 2.9479,
+      "step": 4150
+    },
+    {
+      "epoch": 0.7986147852436151,
+      "grad_norm": 3.7315036800967176,
+      "learning_rate": 2.0543105241459713e-06,
+      "loss": 2.9158,
+      "step": 4151
+    },
+    {
+      "epoch": 0.7988071761820018,
+      "grad_norm": 3.5113746990143646,
+      "learning_rate": 2.0505281095044804e-06,
+      "loss": 2.9221,
+      "step": 4152
+    },
+    {
+      "epoch": 0.7989995671203887,
+      "grad_norm": 3.6435632264941273,
+      "learning_rate": 2.0467487823462696e-06,
+      "loss": 2.8937,
+      "step": 4153
+    },
+    {
+      "epoch": 0.7991919580587754,
+      "grad_norm": 3.493200064942973,
+      "learning_rate": 2.042972544139189e-06,
+      "loss": 2.8652,
+      "step": 4154
+    },
+    {
+      "epoch": 0.7993843489971623,
+      "grad_norm": 3.728005809975312,
+      "learning_rate": 2.039199396349881e-06,
+      "loss": 2.6643,
+      "step": 4155
+    },
+    {
+      "epoch": 0.799576739935549,
+      "grad_norm": 3.2059576560850913,
+      "learning_rate": 2.0354293404437963e-06,
+      "loss": 3.1,
+      "step": 4156
+    },
+    {
+      "epoch": 0.7997691308739359,
+      "grad_norm": 3.3705993519087305,
+      "learning_rate": 2.0316623778851784e-06,
+      "loss": 2.5699,
+      "step": 4157
+    },
+    {
+      "epoch": 0.7999615218123226,
+      "grad_norm": 3.4523779096700418,
+      "learning_rate": 2.0278985101370753e-06,
+      "loss": 3.0584,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8001539127507095,
+      "grad_norm": 3.4690099822005482,
+      "learning_rate": 2.024137738661329e-06,
+      "loss": 2.8534,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8003463036890962,
+      "grad_norm": 3.451975501804431,
+      "learning_rate": 2.020380064918579e-06,
+      "loss": 2.9116,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8005386946274831,
+      "grad_norm": 3.467680528857577,
+      "learning_rate": 2.01662549036826e-06,
+      "loss": 2.9443,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8007310855658698,
+      "grad_norm": 3.3192286656553236,
+      "learning_rate": 2.0128740164686134e-06,
+      "loss": 2.9336,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8009234765042567,
+      "grad_norm": 3.5273821678419015,
+      "learning_rate": 2.009125644676656e-06,
+      "loss": 2.8225,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8011158674426434,
+      "grad_norm": 3.4993455796069544,
+      "learning_rate": 2.0053803764482226e-06,
+      "loss": 3.0003,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8013082583810303,
+      "grad_norm": 3.7294688747287874,
+      "learning_rate": 2.001638213237932e-06,
+      "loss": 2.9679,
+      "step": 4165
+    },
+    {
+      "epoch": 0.801500649319417,
+      "grad_norm": 3.7614554567612424,
+      "learning_rate": 1.997899156499191e-06,
+      "loss": 2.9237,
+      "step": 4166
+    },
+    {
+      "epoch": 0.8016930402578039,
+      "grad_norm": 3.404254741729379,
+      "learning_rate": 1.994163207684212e-06,
+      "loss": 2.9859,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8018854311961907,
+      "grad_norm": 3.683161876922934,
+      "learning_rate": 1.9904303682439896e-06,
+      "loss": 2.8505,
+      "step": 4168
+    },
+    {
+      "epoch": 0.8020778221345775,
+      "grad_norm": 3.5994558022535883,
+      "learning_rate": 1.986700639628316e-06,
+      "loss": 2.787,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8022702130729643,
+      "grad_norm": 2.311427361319493,
+      "learning_rate": 1.9829740232857807e-06,
+      "loss": 2.6409,
+      "step": 4170
+    },
+    {
+      "epoch": 0.802462604011351,
+      "grad_norm": 3.525885560469133,
+      "learning_rate": 1.9792505206637523e-06,
+      "loss": 3.05,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8026549949497379,
+      "grad_norm": 3.428174689781946,
+      "learning_rate": 1.9755301332083997e-06,
+      "loss": 2.8544,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8028473858881247,
+      "grad_norm": 2.2902289694374525,
+      "learning_rate": 1.9718128623646792e-06,
+      "loss": 2.7084,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8030397768265115,
+      "grad_norm": 3.644701874351,
+      "learning_rate": 1.9680987095763315e-06,
+      "loss": 3.2633,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8032321677648983,
+      "grad_norm": 3.3246195426815155,
+      "learning_rate": 1.964387676285894e-06,
+      "loss": 2.7362,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8034245587032851,
+      "grad_norm": 3.452275454050276,
+      "learning_rate": 1.9606797639346874e-06,
+      "loss": 3.1654,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8036169496416719,
+      "grad_norm": 3.6421108968658826,
+      "learning_rate": 1.9569749739628243e-06,
+      "loss": 2.8913,
+      "step": 4177
+    },
+    {
+      "epoch": 0.8038093405800587,
+      "grad_norm": 3.6896801312791285,
+      "learning_rate": 1.9532733078092034e-06,
+      "loss": 3.0575,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8040017315184455,
+      "grad_norm": 3.3408450381027017,
+      "learning_rate": 1.9495747669115062e-06,
+      "loss": 2.857,
+      "step": 4179
+    },
+    {
+      "epoch": 0.8041941224568323,
+      "grad_norm": 2.3114403774994288,
+      "learning_rate": 1.9458793527062035e-06,
+      "loss": 2.4369,
+      "step": 4180
+    },
+    {
+      "epoch": 0.804386513395219,
+      "grad_norm": 3.4886141786932634,
+      "learning_rate": 1.9421870666285523e-06,
+      "loss": 2.8463,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8045789043336059,
+      "grad_norm": 3.8741999983834075,
+      "learning_rate": 1.9384979101125944e-06,
+      "loss": 2.9125,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8047712952719926,
+      "grad_norm": 3.263714260717895,
+      "learning_rate": 1.934811884591159e-06,
+      "loss": 2.6962,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8049636862103795,
+      "grad_norm": 3.485264848649117,
+      "learning_rate": 1.93112899149585e-06,
+      "loss": 2.8614,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8051560771487662,
+      "grad_norm": 3.3692976714692877,
+      "learning_rate": 1.9274492322570616e-06,
+      "loss": 2.8633,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8053484680871531,
+      "grad_norm": 3.4425799397036254,
+      "learning_rate": 1.923772608303972e-06,
+      "loss": 2.8511,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8055408590255398,
+      "grad_norm": 3.546734155195182,
+      "learning_rate": 1.9200991210645394e-06,
+      "loss": 2.9643,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8057332499639267,
+      "grad_norm": 3.3933617056716514,
+      "learning_rate": 1.916428771965506e-06,
+      "loss": 3.18,
+      "step": 4188
+    },
+    {
+      "epoch": 0.8059256409023136,
+      "grad_norm": 3.427646018548786,
+      "learning_rate": 1.912761562432388e-06,
+      "loss": 2.7746,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8061180318407003,
+      "grad_norm": 3.4310513735771506,
+      "learning_rate": 1.9090974938894902e-06,
+      "loss": 2.989,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8063104227790872,
+      "grad_norm": 3.535273328283496,
+      "learning_rate": 1.9054365677598963e-06,
+      "loss": 2.9597,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8065028137174739,
+      "grad_norm": 3.455781876476265,
+      "learning_rate": 1.9017787854654613e-06,
+      "loss": 2.9617,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8066952046558608,
+      "grad_norm": 3.4868660644980602,
+      "learning_rate": 1.898124148426832e-06,
+      "loss": 2.976,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8068875955942475,
+      "grad_norm": 3.5218051736000433,
+      "learning_rate": 1.8944726580634287e-06,
+      "loss": 3.0264,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8070799865326344,
+      "grad_norm": 3.3259184231541155,
+      "learning_rate": 1.8908243157934424e-06,
+      "loss": 3.0047,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8072723774710211,
+      "grad_norm": 3.251953741660502,
+      "learning_rate": 1.8871791230338499e-06,
+      "loss": 2.8462,
+      "step": 4196
+    },
+    {
+      "epoch": 0.807464768409408,
+      "grad_norm": 3.601363009200978,
+      "learning_rate": 1.883537081200404e-06,
+      "loss": 3.0154,
+      "step": 4197
+    },
+    {
+      "epoch": 0.8076571593477947,
+      "grad_norm": 3.2696597971986705,
+      "learning_rate": 1.8798981917076254e-06,
+      "loss": 2.8857,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8078495502861816,
+      "grad_norm": 3.4478514079971685,
+      "learning_rate": 1.876262455968826e-06,
+      "loss": 2.9129,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8080419412245683,
+      "grad_norm": 3.7054880726312227,
+      "learning_rate": 1.872629875396076e-06,
+      "loss": 3.0026,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8082343321629551,
+      "grad_norm": 3.55720658789402,
+      "learning_rate": 1.8690004514002314e-06,
+      "loss": 2.9215,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8084267231013419,
+      "grad_norm": 3.56245224416241,
+      "learning_rate": 1.8653741853909201e-06,
+      "loss": 2.7583,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8086191140397287,
+      "grad_norm": 3.3378440785655137,
+      "learning_rate": 1.861751078776538e-06,
+      "loss": 2.6116,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8088115049781155,
+      "grad_norm": 3.7361808645996892,
+      "learning_rate": 1.8581311329642592e-06,
+      "loss": 3.1911,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8090038959165023,
+      "grad_norm": 3.2991213326202202,
+      "learning_rate": 1.8545143493600293e-06,
+      "loss": 3.2431,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8091962868548891,
+      "grad_norm": 3.630303660373292,
+      "learning_rate": 1.8509007293685666e-06,
+      "loss": 3.1685,
+      "step": 4206
+    },
+    {
+      "epoch": 0.809388677793276,
+      "grad_norm": 3.3619477149483505,
+      "learning_rate": 1.8472902743933608e-06,
+      "loss": 2.8168,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8095810687316627,
+      "grad_norm": 3.584685912375191,
+      "learning_rate": 1.8436829858366655e-06,
+      "loss": 2.7536,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8097734596700495,
+      "grad_norm": 3.1271535369198555,
+      "learning_rate": 1.8400788650995137e-06,
+      "loss": 2.9241,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8099658506084363,
+      "grad_norm": 3.4745499134172317,
+      "learning_rate": 1.8364779135817045e-06,
+      "loss": 2.8079,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8101582415468231,
+      "grad_norm": 3.3167273795835195,
+      "learning_rate": 1.8328801326818045e-06,
+      "loss": 2.8396,
+      "step": 4211
+    },
+    {
+      "epoch": 0.81035063248521,
+      "grad_norm": 3.712895174925625,
+      "learning_rate": 1.829285523797155e-06,
+      "loss": 3.0624,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8105430234235967,
+      "grad_norm": 3.505878657481498,
+      "learning_rate": 1.8256940883238538e-06,
+      "loss": 3.1142,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8107354143619836,
+      "grad_norm": 3.6640457226910965,
+      "learning_rate": 1.822105827656776e-06,
+      "loss": 3.1235,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8109278053003703,
+      "grad_norm": 3.647671663060882,
+      "learning_rate": 1.8185207431895613e-06,
+      "loss": 2.9944,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8111201962387572,
+      "grad_norm": 3.3811714170392695,
+      "learning_rate": 1.8149388363146148e-06,
+      "loss": 3.0361,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8113125871771439,
+      "grad_norm": 2.397006594435446,
+      "learning_rate": 1.8113601084231091e-06,
+      "loss": 2.5615,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8115049781155308,
+      "grad_norm": 3.4249813343095323,
+      "learning_rate": 1.8077845609049782e-06,
+      "loss": 2.9998,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8116973690539175,
+      "grad_norm": 3.4489762222320226,
+      "learning_rate": 1.8042121951489254e-06,
+      "loss": 2.87,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8118897599923044,
+      "grad_norm": 3.515539286017614,
+      "learning_rate": 1.800643012542418e-06,
+      "loss": 2.8588,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8120821509306911,
+      "grad_norm": 3.569771363177175,
+      "learning_rate": 1.7970770144716777e-06,
+      "loss": 3.1813,
+      "step": 4221
+    },
+    {
+      "epoch": 0.812274541869078,
+      "grad_norm": 3.2974625244125533,
+      "learning_rate": 1.7935142023217056e-06,
+      "loss": 2.9723,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8124669328074647,
+      "grad_norm": 2.4834486860328835,
+      "learning_rate": 1.7899545774762573e-06,
+      "loss": 2.661,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8126593237458516,
+      "grad_norm": 3.6327122670122636,
+      "learning_rate": 1.7863981413178433e-06,
+      "loss": 2.8472,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8128517146842383,
+      "grad_norm": 2.359356893350355,
+      "learning_rate": 1.7828448952277456e-06,
+      "loss": 2.6041,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8130441056226252,
+      "grad_norm": 3.6974116501568344,
+      "learning_rate": 1.7792948405860079e-06,
+      "loss": 3.247,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8132364965610119,
+      "grad_norm": 3.5796356768747897,
+      "learning_rate": 1.7757479787714217e-06,
+      "loss": 2.8097,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8134288874993988,
+      "grad_norm": 3.7097425483416595,
+      "learning_rate": 1.7722043111615572e-06,
+      "loss": 2.9759,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8136212784377855,
+      "grad_norm": 3.4149022104189144,
+      "learning_rate": 1.768663839132727e-06,
+      "loss": 2.8082,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8138136693761724,
+      "grad_norm": 3.3585667607417533,
+      "learning_rate": 1.7651265640600113e-06,
+      "loss": 3.0905,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8140060603145591,
+      "grad_norm": 3.127221327853468,
+      "learning_rate": 1.7615924873172506e-06,
+      "loss": 2.6096,
+      "step": 4231
+    },
+    {
+      "epoch": 0.814198451252946,
+      "grad_norm": 3.8130344787785053,
+      "learning_rate": 1.7580616102770353e-06,
+      "loss": 2.977,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8143908421913328,
+      "grad_norm": 4.033840900140363,
+      "learning_rate": 1.754533934310717e-06,
+      "loss": 2.9495,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8145832331297196,
+      "grad_norm": 3.5688431778034735,
+      "learning_rate": 1.7510094607884075e-06,
+      "loss": 3.0382,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8147756240681064,
+      "grad_norm": 3.459931051941258,
+      "learning_rate": 1.7474881910789698e-06,
+      "loss": 3.125,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8149680150064932,
+      "grad_norm": 3.7510254274937163,
+      "learning_rate": 1.7439701265500274e-06,
+      "loss": 2.8863,
+      "step": 4236
+    },
+    {
+      "epoch": 0.81516040594488,
+      "grad_norm": 3.4713317419416985,
+      "learning_rate": 1.740455268567951e-06,
+      "loss": 2.8646,
+      "step": 4237
+    },
+    {
+      "epoch": 0.8153527968832668,
+      "grad_norm": 2.2449302129763242,
+      "learning_rate": 1.7369436184978738e-06,
+      "loss": 2.5982,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8155451878216536,
+      "grad_norm": 3.755949363708671,
+      "learning_rate": 1.7334351777036807e-06,
+      "loss": 3.0266,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8157375787600404,
+      "grad_norm": 3.7547654110873054,
+      "learning_rate": 1.729929947548008e-06,
+      "loss": 3.1089,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8159299696984272,
+      "grad_norm": 3.4031054567879493,
+      "learning_rate": 1.7264279293922503e-06,
+      "loss": 2.8589,
+      "step": 4241
+    },
+    {
+      "epoch": 0.816122360636814,
+      "grad_norm": 3.8310833691074615,
+      "learning_rate": 1.7229291245965463e-06,
+      "loss": 3.0137,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8163147515752008,
+      "grad_norm": 3.3868794572553464,
+      "learning_rate": 1.7194335345197933e-06,
+      "loss": 2.843,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8165071425135876,
+      "grad_norm": 2.3166101979163765,
+      "learning_rate": 1.7159411605196407e-06,
+      "loss": 2.4137,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8166995334519744,
+      "grad_norm": 3.7687709862347605,
+      "learning_rate": 1.7124520039524805e-06,
+      "loss": 3.1449,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8168919243903612,
+      "grad_norm": 2.2603394841110562,
+      "learning_rate": 1.7089660661734685e-06,
+      "loss": 2.4246,
+      "step": 4246
+    },
+    {
+      "epoch": 0.817084315328748,
+      "grad_norm": 3.1604223032386165,
+      "learning_rate": 1.7054833485364962e-06,
+      "loss": 3.0653,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8172767062671348,
+      "grad_norm": 3.5470483619629807,
+      "learning_rate": 1.702003852394214e-06,
+      "loss": 3.0148,
+      "step": 4248
+    },
+    {
+      "epoch": 0.8174690972055216,
+      "grad_norm": 3.300946947402923,
+      "learning_rate": 1.6985275790980205e-06,
+      "loss": 2.8852,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8176614881439084,
+      "grad_norm": 3.1575268589428096,
+      "learning_rate": 1.6950545299980526e-06,
+      "loss": 2.8346,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8178538790822952,
+      "grad_norm": 3.709301345713615,
+      "learning_rate": 1.691584706443209e-06,
+      "loss": 2.87,
+      "step": 4251
+    },
+    {
+      "epoch": 0.818046270020682,
+      "grad_norm": 3.6139926571660106,
+      "learning_rate": 1.6881181097811305e-06,
+      "loss": 3.1306,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8182386609590688,
+      "grad_norm": 3.606905114054193,
+      "learning_rate": 1.6846547413581981e-06,
+      "loss": 3.017,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8184310518974556,
+      "grad_norm": 3.389178214382058,
+      "learning_rate": 1.6811946025195459e-06,
+      "loss": 2.8221,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8186234428358424,
+      "grad_norm": 3.8000884181542114,
+      "learning_rate": 1.6777376946090552e-06,
+      "loss": 2.8956,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8188158337742293,
+      "grad_norm": 3.236266383620102,
+      "learning_rate": 1.674284018969342e-06,
+      "loss": 2.907,
+      "step": 4256
+    },
+    {
+      "epoch": 0.819008224712616,
+      "grad_norm": 3.31680972305381,
+      "learning_rate": 1.6708335769417827e-06,
+      "loss": 2.8862,
+      "step": 4257
+    },
+    {
+      "epoch": 0.8192006156510029,
+      "grad_norm": 2.2361483519965955,
+      "learning_rate": 1.667386369866484e-06,
+      "loss": 2.3754,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8193930065893896,
+      "grad_norm": 3.7496796312114773,
+      "learning_rate": 1.6639423990823011e-06,
+      "loss": 3.1277,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8195853975277765,
+      "grad_norm": 3.4497777601869966,
+      "learning_rate": 1.660501665926838e-06,
+      "loss": 3.1096,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8197777884661632,
+      "grad_norm": 3.57215361351059,
+      "learning_rate": 1.6570641717364277e-06,
+      "loss": 2.9827,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8199701794045501,
+      "grad_norm": 3.4781691998337156,
+      "learning_rate": 1.653629917846159e-06,
+      "loss": 3.0291,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8201625703429368,
+      "grad_norm": 3.600064973450613,
+      "learning_rate": 1.6501989055898537e-06,
+      "loss": 3.1207,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8203549612813237,
+      "grad_norm": 3.446434317406245,
+      "learning_rate": 1.6467711363000794e-06,
+      "loss": 2.9354,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8205473522197104,
+      "grad_norm": 3.3282248558991725,
+      "learning_rate": 1.6433466113081442e-06,
+      "loss": 2.8259,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8207397431580973,
+      "grad_norm": 3.3097693179796894,
+      "learning_rate": 1.6399253319440888e-06,
+      "loss": 2.6278,
+      "step": 4266
+    },
+    {
+      "epoch": 0.820932134096484,
+      "grad_norm": 3.250376623150259,
+      "learning_rate": 1.6365072995367004e-06,
+      "loss": 2.904,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8211245250348709,
+      "grad_norm": 3.17942659957951,
+      "learning_rate": 1.6330925154135057e-06,
+      "loss": 3.0184,
+      "step": 4268
+    },
+    {
+      "epoch": 0.8213169159732576,
+      "grad_norm": 3.36619315422832,
+      "learning_rate": 1.6296809809007652e-06,
+      "loss": 2.9053,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8215093069116445,
+      "grad_norm": 3.2949070076518954,
+      "learning_rate": 1.6262726973234844e-06,
+      "loss": 3.0786,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8217016978500312,
+      "grad_norm": 4.0548071838941775,
+      "learning_rate": 1.6228676660053932e-06,
+      "loss": 3.1557,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8218940887884181,
+      "grad_norm": 3.49998758282229,
+      "learning_rate": 1.6194658882689718e-06,
+      "loss": 2.9164,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8220864797268048,
+      "grad_norm": 3.3024712266889864,
+      "learning_rate": 1.6160673654354331e-06,
+      "loss": 3.1869,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8222788706651917,
+      "grad_norm": 3.8651933282644135,
+      "learning_rate": 1.6126720988247168e-06,
+      "loss": 3.2116,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8224712616035784,
+      "grad_norm": 3.414053755980381,
+      "learning_rate": 1.609280089755515e-06,
+      "loss": 3.1941,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8226636525419653,
+      "grad_norm": 3.4368337271541645,
+      "learning_rate": 1.605891339545237e-06,
+      "loss": 2.9048,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8228560434803521,
+      "grad_norm": 3.3869847407584395,
+      "learning_rate": 1.6025058495100388e-06,
+      "loss": 2.8989,
+      "step": 4277
+    },
+    {
+      "epoch": 0.8230484344187389,
+      "grad_norm": 3.1127067269669104,
+      "learning_rate": 1.5991236209648052e-06,
+      "loss": 2.7846,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8232408253571257,
+      "grad_norm": 3.642966943322272,
+      "learning_rate": 1.5957446552231526e-06,
+      "loss": 2.9356,
+      "step": 4279
+    },
+    {
+      "epoch": 0.8234332162955125,
+      "grad_norm": 3.042355538314279,
+      "learning_rate": 1.5923689535974307e-06,
+      "loss": 2.6349,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8236256072338993,
+      "grad_norm": 3.2729777319996685,
+      "learning_rate": 1.588996517398731e-06,
+      "loss": 2.9137,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8238179981722861,
+      "grad_norm": 3.2478799477399254,
+      "learning_rate": 1.5856273479368611e-06,
+      "loss": 2.9087,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8240103891106729,
+      "grad_norm": 2.0518617107008015,
+      "learning_rate": 1.582261446520371e-06,
+      "loss": 2.528,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8242027800490597,
+      "grad_norm": 3.46661607078232,
+      "learning_rate": 1.5788988144565397e-06,
+      "loss": 2.8207,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8243951709874465,
+      "grad_norm": 2.310009774612975,
+      "learning_rate": 1.575539453051369e-06,
+      "loss": 2.5933,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8245875619258333,
+      "grad_norm": 3.2918556624756157,
+      "learning_rate": 1.572183363609603e-06,
+      "loss": 3.0311,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8247799528642201,
+      "grad_norm": 3.28391006912501,
+      "learning_rate": 1.568830547434703e-06,
+      "loss": 2.7921,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8249723438026069,
+      "grad_norm": 3.256917102971025,
+      "learning_rate": 1.5654810058288661e-06,
+      "loss": 2.9304,
+      "step": 4288
+    },
+    {
+      "epoch": 0.8251647347409937,
+      "grad_norm": 3.6921108424844946,
+      "learning_rate": 1.5621347400930176e-06,
+      "loss": 3.0559,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8253571256793805,
+      "grad_norm": 3.374277806867043,
+      "learning_rate": 1.5587917515268048e-06,
+      "loss": 2.7801,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8255495166177673,
+      "grad_norm": 3.2857037637281885,
+      "learning_rate": 1.5554520414286067e-06,
+      "loss": 3.1138,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8257419075561541,
+      "grad_norm": 3.3932363535241747,
+      "learning_rate": 1.5521156110955293e-06,
+      "loss": 2.8016,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8259342984945409,
+      "grad_norm": 3.299311220234353,
+      "learning_rate": 1.5487824618234049e-06,
+      "loss": 2.9588,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8261266894329277,
+      "grad_norm": 3.3332799826656476,
+      "learning_rate": 1.54545259490679e-06,
+      "loss": 2.8269,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8263190803713145,
+      "grad_norm": 3.564281768866658,
+      "learning_rate": 1.5421260116389636e-06,
+      "loss": 3.0444,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8265114713097013,
+      "grad_norm": 3.594423199069475,
+      "learning_rate": 1.5388027133119343e-06,
+      "loss": 3.0291,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8267038622480881,
+      "grad_norm": 3.4252186706850587,
+      "learning_rate": 1.535482701216433e-06,
+      "loss": 2.7365,
+      "step": 4297
+    },
+    {
+      "epoch": 0.826896253186475,
+      "grad_norm": 3.4883520429071155,
+      "learning_rate": 1.5321659766419129e-06,
+      "loss": 3.102,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8270886441248617,
+      "grad_norm": 3.556287938665445,
+      "learning_rate": 1.5288525408765564e-06,
+      "loss": 3.029,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8272810350632486,
+      "grad_norm": 3.6206973242754095,
+      "learning_rate": 1.5255423952072567e-06,
+      "loss": 3.1304,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8274734260016353,
+      "grad_norm": 3.388794961313886,
+      "learning_rate": 1.52223554091964e-06,
+      "loss": 2.9977,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8276658169400222,
+      "grad_norm": 3.553465513592159,
+      "learning_rate": 1.5189319792980517e-06,
+      "loss": 2.819,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8278582078784089,
+      "grad_norm": 3.1944557785175727,
+      "learning_rate": 1.5156317116255515e-06,
+      "loss": 3.0251,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8280505988167958,
+      "grad_norm": 3.2859061487613133,
+      "learning_rate": 1.5123347391839305e-06,
+      "loss": 2.7917,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8282429897551825,
+      "grad_norm": 3.842404856516146,
+      "learning_rate": 1.5090410632536968e-06,
+      "loss": 3.0553,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8284353806935694,
+      "grad_norm": 3.426696633136265,
+      "learning_rate": 1.5057506851140701e-06,
+      "loss": 2.766,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8286277716319561,
+      "grad_norm": 3.3503700136932637,
+      "learning_rate": 1.5024636060429998e-06,
+      "loss": 2.9843,
+      "step": 4307
+    },
+    {
+      "epoch": 0.828820162570343,
+      "grad_norm": 3.64056912787485,
+      "learning_rate": 1.4991798273171465e-06,
+      "loss": 2.7796,
+      "step": 4308
+    },
+    {
+      "epoch": 0.8290125535087297,
+      "grad_norm": 3.5690673320560866,
+      "learning_rate": 1.49589935021189e-06,
+      "loss": 2.9275,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8292049444471166,
+      "grad_norm": 3.647325128882531,
+      "learning_rate": 1.4926221760013393e-06,
+      "loss": 2.9143,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8293973353855033,
+      "grad_norm": 3.54889170887107,
+      "learning_rate": 1.4893483059583014e-06,
+      "loss": 2.9649,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8295897263238902,
+      "grad_norm": 3.3827773767766214,
+      "learning_rate": 1.4860777413543138e-06,
+      "loss": 2.975,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8297821172622769,
+      "grad_norm": 3.342282395566732,
+      "learning_rate": 1.4828104834596268e-06,
+      "loss": 2.8181,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8299745082006638,
+      "grad_norm": 3.273267239406805,
+      "learning_rate": 1.4795465335432036e-06,
+      "loss": 2.7229,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8301668991390505,
+      "grad_norm": 3.691960703939425,
+      "learning_rate": 1.4762858928727241e-06,
+      "loss": 3.0475,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8303592900774374,
+      "grad_norm": 3.5582706525565366,
+      "learning_rate": 1.4730285627145858e-06,
+      "loss": 3.1519,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8305516810158241,
+      "grad_norm": 3.5162222870478375,
+      "learning_rate": 1.4697745443338984e-06,
+      "loss": 3.0411,
+      "step": 4317
+    },
+    {
+      "epoch": 0.830744071954211,
+      "grad_norm": 3.3402515502396977,
+      "learning_rate": 1.4665238389944859e-06,
+      "loss": 2.7791,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8309364628925977,
+      "grad_norm": 3.6864389329712233,
+      "learning_rate": 1.46327644795888e-06,
+      "loss": 2.995,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8311288538309846,
+      "grad_norm": 3.711874316983118,
+      "learning_rate": 1.4600323724883337e-06,
+      "loss": 2.8965,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8313212447693714,
+      "grad_norm": 3.521180341948043,
+      "learning_rate": 1.4567916138428072e-06,
+      "loss": 3.0456,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8315136357077582,
+      "grad_norm": 3.587439873684547,
+      "learning_rate": 1.4535541732809755e-06,
+      "loss": 2.7117,
+      "step": 4322
+    },
+    {
+      "epoch": 0.831706026646145,
+      "grad_norm": 3.5353869902890813,
+      "learning_rate": 1.4503200520602245e-06,
+      "loss": 2.9699,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8318984175845318,
+      "grad_norm": 3.3259120659062327,
+      "learning_rate": 1.4470892514366442e-06,
+      "loss": 2.8045,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8320908085229186,
+      "grad_norm": 3.7457042079979592,
+      "learning_rate": 1.443861772665044e-06,
+      "loss": 3.0259,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8322831994613054,
+      "grad_norm": 3.4768459724237455,
+      "learning_rate": 1.4406376169989389e-06,
+      "loss": 2.9387,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8324755903996922,
+      "grad_norm": 3.3265507612199396,
+      "learning_rate": 1.4374167856905542e-06,
+      "loss": 3.1168,
+      "step": 4327
+    },
+    {
+      "epoch": 0.832667981338079,
+      "grad_norm": 3.3017503844237,
+      "learning_rate": 1.4341992799908255e-06,
+      "loss": 2.942,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8328603722764658,
+      "grad_norm": 3.4641471197831915,
+      "learning_rate": 1.4309851011493903e-06,
+      "loss": 2.9742,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8330527632148526,
+      "grad_norm": 3.447446522843586,
+      "learning_rate": 1.427774250414601e-06,
+      "loss": 3.0799,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8332451541532394,
+      "grad_norm": 3.5362345368173846,
+      "learning_rate": 1.4245667290335175e-06,
+      "loss": 3.1877,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8334375450916262,
+      "grad_norm": 3.3715764470957916,
+      "learning_rate": 1.421362538251897e-06,
+      "loss": 2.9987,
+      "step": 4332
+    },
+    {
+      "epoch": 0.833629936030013,
+      "grad_norm": 3.6147241980314293,
+      "learning_rate": 1.4181616793142173e-06,
+      "loss": 2.9078,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8338223269683998,
+      "grad_norm": 3.3919424950003876,
+      "learning_rate": 1.414964153463655e-06,
+      "loss": 2.9678,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8340147179067866,
+      "grad_norm": 3.610980928627593,
+      "learning_rate": 1.4117699619420878e-06,
+      "loss": 2.9263,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8342071088451734,
+      "grad_norm": 3.369072593344744,
+      "learning_rate": 1.4085791059901077e-06,
+      "loss": 3.1365,
+      "step": 4336
+    },
+    {
+      "epoch": 0.8343994997835602,
+      "grad_norm": 3.7946688251433898,
+      "learning_rate": 1.4053915868470013e-06,
+      "loss": 3.023,
+      "step": 4337
+    },
+    {
+      "epoch": 0.834591890721947,
+      "grad_norm": 3.3563852404460985,
+      "learning_rate": 1.402207405750765e-06,
+      "loss": 2.8013,
+      "step": 4338
+    },
+    {
+      "epoch": 0.8347842816603338,
+      "grad_norm": 3.6130375035032203,
+      "learning_rate": 1.399026563938105e-06,
+      "loss": 2.7703,
+      "step": 4339
+    },
+    {
+      "epoch": 0.8349766725987205,
+      "grad_norm": 3.4554338077599795,
+      "learning_rate": 1.3958490626444154e-06,
+      "loss": 3.0442,
+      "step": 4340
+    },
+    {
+      "epoch": 0.8351690635371074,
+      "grad_norm": 3.6437831463520967,
+      "learning_rate": 1.3926749031038055e-06,
+      "loss": 2.8598,
+      "step": 4341
+    },
+    {
+      "epoch": 0.8353614544754943,
+      "grad_norm": 3.274533128936126,
+      "learning_rate": 1.3895040865490817e-06,
+      "loss": 2.8546,
+      "step": 4342
+    },
+    {
+      "epoch": 0.835553845413881,
+      "grad_norm": 3.2527026971061566,
+      "learning_rate": 1.3863366142117506e-06,
+      "loss": 2.9082,
+      "step": 4343
+    },
+    {
+      "epoch": 0.8357462363522679,
+      "grad_norm": 3.474781337518833,
+      "learning_rate": 1.383172487322023e-06,
+      "loss": 2.9555,
+      "step": 4344
+    },
+    {
+      "epoch": 0.8359386272906546,
+      "grad_norm": 3.28877224701534,
+      "learning_rate": 1.3800117071088104e-06,
+      "loss": 3.0692,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8361310182290415,
+      "grad_norm": 3.387623494159824,
+      "learning_rate": 1.3768542747997215e-06,
+      "loss": 2.9499,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8363234091674282,
+      "grad_norm": 3.3072189477329035,
+      "learning_rate": 1.3737001916210713e-06,
+      "loss": 3.0397,
+      "step": 4347
+    },
+    {
+      "epoch": 0.836515800105815,
+      "grad_norm": 3.5691145626135063,
+      "learning_rate": 1.370549458797863e-06,
+      "loss": 2.8432,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8367081910442018,
+      "grad_norm": 3.7827613644348013,
+      "learning_rate": 1.3674020775538078e-06,
+      "loss": 3.159,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8369005819825887,
+      "grad_norm": 3.6587914040267364,
+      "learning_rate": 1.3642580491113122e-06,
+      "loss": 3.0704,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8370929729209754,
+      "grad_norm": 3.3683286755737747,
+      "learning_rate": 1.3611173746914797e-06,
+      "loss": 3.0867,
+      "step": 4351
+    },
+    {
+      "epoch": 0.8372853638593623,
+      "grad_norm": 3.2940228690395212,
+      "learning_rate": 1.3579800555141165e-06,
+      "loss": 2.9674,
+      "step": 4352
+    },
+    {
+      "epoch": 0.837477754797749,
+      "grad_norm": 3.4059145233785424,
+      "learning_rate": 1.3548460927977158e-06,
+      "loss": 3.0523,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8376701457361359,
+      "grad_norm": 3.6070538568571697,
+      "learning_rate": 1.351715487759474e-06,
+      "loss": 2.9256,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8378625366745226,
+      "grad_norm": 3.4981438624383667,
+      "learning_rate": 1.3485882416152819e-06,
+      "loss": 2.8614,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8380549276129095,
+      "grad_norm": 3.2485317988247333,
+      "learning_rate": 1.3454643555797276e-06,
+      "loss": 2.9636,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8382473185512962,
+      "grad_norm": 3.5344363262953498,
+      "learning_rate": 1.3423438308660929e-06,
+      "loss": 3.2541,
+      "step": 4357
+    },
+    {
+      "epoch": 0.838439709489683,
+      "grad_norm": 3.6262963693528523,
+      "learning_rate": 1.3392266686863508e-06,
+      "loss": 2.9714,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8386321004280698,
+      "grad_norm": 3.705666772112271,
+      "learning_rate": 1.3361128702511716e-06,
+      "loss": 3.128,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8388244913664566,
+      "grad_norm": 3.666921708156158,
+      "learning_rate": 1.3330024367699224e-06,
+      "loss": 3.1845,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8390168823048434,
+      "grad_norm": 2.232580617054783,
+      "learning_rate": 1.3298953694506522e-06,
+      "loss": 2.4399,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8392092732432302,
+      "grad_norm": 3.295028221765404,
+      "learning_rate": 1.3267916695001172e-06,
+      "loss": 3.0458,
+      "step": 4362
+    },
+    {
+      "epoch": 0.839401664181617,
+      "grad_norm": 3.471905429610491,
+      "learning_rate": 1.3236913381237592e-06,
+      "loss": 2.876,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8395940551200038,
+      "grad_norm": 3.4891769081665793,
+      "learning_rate": 1.3205943765257057e-06,
+      "loss": 3.0273,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8397864460583907,
+      "grad_norm": 3.461003353440141,
+      "learning_rate": 1.317500785908783e-06,
+      "loss": 3.0524,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8399788369967774,
+      "grad_norm": 3.6326060661466317,
+      "learning_rate": 1.31441056747451e-06,
+      "loss": 3.3353,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8401712279351643,
+      "grad_norm": 3.1035116473914273,
+      "learning_rate": 1.3113237224230836e-06,
+      "loss": 3.0131,
+      "step": 4367
+    },
+    {
+      "epoch": 0.840363618873551,
+      "grad_norm": 3.6247339288037566,
+      "learning_rate": 1.3082402519534076e-06,
+      "loss": 2.9173,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8405560098119379,
+      "grad_norm": 3.863303054922617,
+      "learning_rate": 1.3051601572630611e-06,
+      "loss": 2.833,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8407484007503246,
+      "grad_norm": 3.6506160760528434,
+      "learning_rate": 1.3020834395483195e-06,
+      "loss": 3.218,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8409407916887115,
+      "grad_norm": 3.284285604609615,
+      "learning_rate": 1.2990101000041445e-06,
+      "loss": 2.9043,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8411331826270982,
+      "grad_norm": 3.471073058464638,
+      "learning_rate": 1.2959401398241844e-06,
+      "loss": 2.7385,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8413255735654851,
+      "grad_norm": 2.2250548941737813,
+      "learning_rate": 1.2928735602007768e-06,
+      "loss": 2.5785,
+      "step": 4373
+    },
+    {
+      "epoch": 0.8415179645038718,
+      "grad_norm": 3.309776405258926,
+      "learning_rate": 1.2898103623249458e-06,
+      "loss": 2.8489,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8417103554422587,
+      "grad_norm": 3.380588839119822,
+      "learning_rate": 1.2867505473864029e-06,
+      "loss": 2.7934,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8419027463806454,
+      "grad_norm": 3.672816840354642,
+      "learning_rate": 1.283694116573546e-06,
+      "loss": 2.863,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8420951373190323,
+      "grad_norm": 3.523714998976942,
+      "learning_rate": 1.2806410710734552e-06,
+      "loss": 2.5103,
+      "step": 4377
+    },
+    {
+      "epoch": 0.842287528257419,
+      "grad_norm": 3.5929041614762918,
+      "learning_rate": 1.2775914120718992e-06,
+      "loss": 3.013,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8424799191958059,
+      "grad_norm": 3.36270470180739,
+      "learning_rate": 1.2745451407533294e-06,
+      "loss": 2.9703,
+      "step": 4379
+    },
+    {
+      "epoch": 0.8426723101341926,
+      "grad_norm": 3.5611308022923085,
+      "learning_rate": 1.2715022583008851e-06,
+      "loss": 3.2413,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8428647010725795,
+      "grad_norm": 3.731845605544363,
+      "learning_rate": 1.2684627658963865e-06,
+      "loss": 2.8613,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8430570920109662,
+      "grad_norm": 3.3147717742887477,
+      "learning_rate": 1.265426664720334e-06,
+      "loss": 2.9502,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8432494829493531,
+      "grad_norm": 3.3593901138310716,
+      "learning_rate": 1.2623939559519161e-06,
+      "loss": 2.836,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8434418738877398,
+      "grad_norm": 3.750030885961316,
+      "learning_rate": 1.2593646407690051e-06,
+      "loss": 3.0839,
+      "step": 4384
+    },
+    {
+      "epoch": 0.8436342648261267,
+      "grad_norm": 3.64367362077206,
+      "learning_rate": 1.2563387203481447e-06,
+      "loss": 3.0744,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8438266557645135,
+      "grad_norm": 3.6435472120583734,
+      "learning_rate": 1.2533161958645755e-06,
+      "loss": 2.8741,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8440190467029003,
+      "grad_norm": 2.410298430386821,
+      "learning_rate": 1.2502970684922067e-06,
+      "loss": 2.4799,
+      "step": 4387
+    },
+    {
+      "epoch": 0.8442114376412871,
+      "grad_norm": 3.6688571363403155,
+      "learning_rate": 1.2472813394036344e-06,
+      "loss": 2.9001,
+      "step": 4388
+    },
+    {
+      "epoch": 0.8444038285796739,
+      "grad_norm": 3.4339401173158945,
+      "learning_rate": 1.2442690097701327e-06,
+      "loss": 2.8759,
+      "step": 4389
+    },
+    {
+      "epoch": 0.8445962195180607,
+      "grad_norm": 3.4156018403952633,
+      "learning_rate": 1.2412600807616526e-06,
+      "loss": 2.9257,
+      "step": 4390
+    },
+    {
+      "epoch": 0.8447886104564475,
+      "grad_norm": 3.180922018047438,
+      "learning_rate": 1.2382545535468316e-06,
+      "loss": 2.6351,
+      "step": 4391
+    },
+    {
+      "epoch": 0.8449810013948343,
+      "grad_norm": 3.4778728787066573,
+      "learning_rate": 1.2352524292929823e-06,
+      "loss": 2.8338,
+      "step": 4392
+    },
+    {
+      "epoch": 0.8451733923332211,
+      "grad_norm": 3.3117326740634785,
+      "learning_rate": 1.2322537091660912e-06,
+      "loss": 2.5585,
+      "step": 4393
+    },
+    {
+      "epoch": 0.8453657832716079,
+      "grad_norm": 3.7151341818508183,
+      "learning_rate": 1.22925839433083e-06,
+      "loss": 2.9787,
+      "step": 4394
+    },
+    {
+      "epoch": 0.8455581742099947,
+      "grad_norm": 3.7144377658885577,
+      "learning_rate": 1.2262664859505434e-06,
+      "loss": 2.4667,
+      "step": 4395
+    },
+    {
+      "epoch": 0.8457505651483815,
+      "grad_norm": 3.6040217374091537,
+      "learning_rate": 1.2232779851872511e-06,
+      "loss": 3.0923,
+      "step": 4396
+    },
+    {
+      "epoch": 0.8459429560867683,
+      "grad_norm": 3.676228754551918,
+      "learning_rate": 1.2202928932016588e-06,
+      "loss": 2.9187,
+      "step": 4397
+    },
+    {
+      "epoch": 0.8461353470251551,
+      "grad_norm": 3.136822119902534,
+      "learning_rate": 1.217311211153137e-06,
+      "loss": 2.604,
+      "step": 4398
+    },
+    {
+      "epoch": 0.8463277379635419,
+      "grad_norm": 3.324307504104822,
+      "learning_rate": 1.2143329401997372e-06,
+      "loss": 2.9809,
+      "step": 4399
+    },
+    {
+      "epoch": 0.8465201289019287,
+      "grad_norm": 2.5287433430632085,
+      "learning_rate": 1.2113580814981884e-06,
+      "loss": 2.6182,
+      "step": 4400
+    },
+    {
+      "epoch": 0.8467125198403155,
+      "grad_norm": 3.204711965021837,
+      "learning_rate": 1.2083866362038865e-06,
+      "loss": 2.8456,
+      "step": 4401
+    },
+    {
+      "epoch": 0.8469049107787023,
+      "grad_norm": 3.681990582396036,
+      "learning_rate": 1.2054186054709105e-06,
+      "loss": 3.0818,
+      "step": 4402
+    },
+    {
+      "epoch": 0.8470973017170891,
+      "grad_norm": 3.6923321737465593,
+      "learning_rate": 1.2024539904520072e-06,
+      "loss": 3.1717,
+      "step": 4403
+    },
+    {
+      "epoch": 0.8472896926554759,
+      "grad_norm": 3.739083951731486,
+      "learning_rate": 1.1994927922985999e-06,
+      "loss": 3.0718,
+      "step": 4404
+    },
+    {
+      "epoch": 0.8474820835938627,
+      "grad_norm": 3.30776802886442,
+      "learning_rate": 1.1965350121607866e-06,
+      "loss": 2.8218,
+      "step": 4405
+    },
+    {
+      "epoch": 0.8476744745322495,
+      "grad_norm": 3.6198104796891117,
+      "learning_rate": 1.1935806511873306e-06,
+      "loss": 2.8817,
+      "step": 4406
+    },
+    {
+      "epoch": 0.8478668654706364,
+      "grad_norm": 3.4313883932390983,
+      "learning_rate": 1.1906297105256725e-06,
+      "loss": 3.014,
+      "step": 4407
+    },
+    {
+      "epoch": 0.8480592564090231,
+      "grad_norm": 3.630910405728805,
+      "learning_rate": 1.187682191321925e-06,
+      "loss": 2.8697,
+      "step": 4408
+    },
+    {
+      "epoch": 0.84825164734741,
+      "grad_norm": 3.160592453654707,
+      "learning_rate": 1.1847380947208697e-06,
+      "loss": 2.9504,
+      "step": 4409
+    },
+    {
+      "epoch": 0.8484440382857967,
+      "grad_norm": 3.442869476723721,
+      "learning_rate": 1.1817974218659621e-06,
+      "loss": 3.0056,
+      "step": 4410
+    },
+    {
+      "epoch": 0.8486364292241836,
+      "grad_norm": 2.2769806082273933,
+      "learning_rate": 1.178860173899321e-06,
+      "loss": 2.5838,
+      "step": 4411
+    },
+    {
+      "epoch": 0.8488288201625703,
+      "grad_norm": 3.395902454852852,
+      "learning_rate": 1.1759263519617437e-06,
+      "loss": 2.6377,
+      "step": 4412
+    },
+    {
+      "epoch": 0.8490212111009572,
+      "grad_norm": 3.340351642980907,
+      "learning_rate": 1.172995957192693e-06,
+      "loss": 2.8523,
+      "step": 4413
+    },
+    {
+      "epoch": 0.8492136020393439,
+      "grad_norm": 3.30511919010743,
+      "learning_rate": 1.1700689907302953e-06,
+      "loss": 2.684,
+      "step": 4414
+    },
+    {
+      "epoch": 0.8494059929777308,
+      "grad_norm": 3.477019173746138,
+      "learning_rate": 1.167145453711358e-06,
+      "loss": 2.768,
+      "step": 4415
+    },
+    {
+      "epoch": 0.8495983839161175,
+      "grad_norm": 3.461492426269982,
+      "learning_rate": 1.1642253472713427e-06,
+      "loss": 2.9089,
+      "step": 4416
+    },
+    {
+      "epoch": 0.8497907748545044,
+      "grad_norm": 3.351046794557673,
+      "learning_rate": 1.161308672544389e-06,
+      "loss": 2.9132,
+      "step": 4417
+    },
+    {
+      "epoch": 0.8499831657928911,
+      "grad_norm": 3.5287428198252857,
+      "learning_rate": 1.1583954306633004e-06,
+      "loss": 2.7212,
+      "step": 4418
+    },
+    {
+      "epoch": 0.850175556731278,
+      "grad_norm": 3.507329857941864,
+      "learning_rate": 1.1554856227595435e-06,
+      "loss": 3.0449,
+      "step": 4419
+    },
+    {
+      "epoch": 0.8503679476696647,
+      "grad_norm": 3.541640305736372,
+      "learning_rate": 1.1525792499632526e-06,
+      "loss": 3.0974,
+      "step": 4420
+    },
+    {
+      "epoch": 0.8505603386080516,
+      "grad_norm": 3.5565489290999146,
+      "learning_rate": 1.1496763134032363e-06,
+      "loss": 2.8524,
+      "step": 4421
+    },
+    {
+      "epoch": 0.8507527295464383,
+      "grad_norm": 3.4355477307375475,
+      "learning_rate": 1.1467768142069546e-06,
+      "loss": 3.1687,
+      "step": 4422
+    },
+    {
+      "epoch": 0.8509451204848252,
+      "grad_norm": 3.4266167038156374,
+      "learning_rate": 1.1438807535005437e-06,
+      "loss": 3.0735,
+      "step": 4423
+    },
+    {
+      "epoch": 0.8511375114232119,
+      "grad_norm": 3.6966490566791332,
+      "learning_rate": 1.1409881324088013e-06,
+      "loss": 3.09,
+      "step": 4424
+    },
+    {
+      "epoch": 0.8513299023615988,
+      "grad_norm": 3.603354781550197,
+      "learning_rate": 1.138098952055181e-06,
+      "loss": 2.7076,
+      "step": 4425
+    },
+    {
+      "epoch": 0.8515222932999855,
+      "grad_norm": 3.294391468511948,
+      "learning_rate": 1.1352132135618165e-06,
+      "loss": 2.796,
+      "step": 4426
+    },
+    {
+      "epoch": 0.8517146842383724,
+      "grad_norm": 3.430211578322317,
+      "learning_rate": 1.13233091804949e-06,
+      "loss": 2.736,
+      "step": 4427
+    },
+    {
+      "epoch": 0.8519070751767591,
+      "grad_norm": 3.515517736086178,
+      "learning_rate": 1.1294520666376518e-06,
+      "loss": 2.893,
+      "step": 4428
+    },
+    {
+      "epoch": 0.852099466115146,
+      "grad_norm": 3.4385918065009653,
+      "learning_rate": 1.1265766604444172e-06,
+      "loss": 2.7803,
+      "step": 4429
+    },
+    {
+      "epoch": 0.8522918570535328,
+      "grad_norm": 3.7240083624948985,
+      "learning_rate": 1.1237047005865576e-06,
+      "loss": 2.7949,
+      "step": 4430
+    },
+    {
+      "epoch": 0.8524842479919196,
+      "grad_norm": 3.3490921018572415,
+      "learning_rate": 1.1208361881795116e-06,
+      "loss": 2.8902,
+      "step": 4431
+    },
+    {
+      "epoch": 0.8526766389303064,
+      "grad_norm": 3.7883965184950297,
+      "learning_rate": 1.1179711243373736e-06,
+      "loss": 2.8223,
+      "step": 4432
+    },
+    {
+      "epoch": 0.8528690298686932,
+      "grad_norm": 3.275699803404194,
+      "learning_rate": 1.1151095101729047e-06,
+      "loss": 2.8861,
+      "step": 4433
+    },
+    {
+      "epoch": 0.85306142080708,
+      "grad_norm": 3.7462408968587444,
+      "learning_rate": 1.1122513467975237e-06,
+      "loss": 3.0354,
+      "step": 4434
+    },
+    {
+      "epoch": 0.8532538117454668,
+      "grad_norm": 3.5872398157491,
+      "learning_rate": 1.1093966353213036e-06,
+      "loss": 2.9787,
+      "step": 4435
+    },
+    {
+      "epoch": 0.8534462026838536,
+      "grad_norm": 3.266030006843339,
+      "learning_rate": 1.1065453768529844e-06,
+      "loss": 3.0376,
+      "step": 4436
+    },
+    {
+      "epoch": 0.8536385936222404,
+      "grad_norm": 3.659330521923791,
+      "learning_rate": 1.103697572499961e-06,
+      "loss": 2.766,
+      "step": 4437
+    },
+    {
+      "epoch": 0.8538309845606272,
+      "grad_norm": 3.889483978686789,
+      "learning_rate": 1.1008532233682878e-06,
+      "loss": 2.9837,
+      "step": 4438
+    },
+    {
+      "epoch": 0.854023375499014,
+      "grad_norm": 3.5376910860337114,
+      "learning_rate": 1.0980123305626812e-06,
+      "loss": 2.9937,
+      "step": 4439
+    },
+    {
+      "epoch": 0.8542157664374008,
+      "grad_norm": 3.4043930093218866,
+      "learning_rate": 1.0951748951865048e-06,
+      "loss": 2.9065,
+      "step": 4440
+    },
+    {
+      "epoch": 0.8544081573757876,
+      "grad_norm": 3.4355531939791226,
+      "learning_rate": 1.0923409183417887e-06,
+      "loss": 2.7626,
+      "step": 4441
+    },
+    {
+      "epoch": 0.8546005483141744,
+      "grad_norm": 3.2107986728423206,
+      "learning_rate": 1.0895104011292202e-06,
+      "loss": 2.9758,
+      "step": 4442
+    },
+    {
+      "epoch": 0.8547929392525612,
+      "grad_norm": 3.6268436846985668,
+      "learning_rate": 1.0866833446481317e-06,
+      "loss": 2.8982,
+      "step": 4443
+    },
+    {
+      "epoch": 0.854985330190948,
+      "grad_norm": 3.5530465962870914,
+      "learning_rate": 1.0838597499965276e-06,
+      "loss": 2.7382,
+      "step": 4444
+    },
+    {
+      "epoch": 0.8551777211293348,
+      "grad_norm": 3.4782015932516974,
+      "learning_rate": 1.0810396182710535e-06,
+      "loss": 2.8516,
+      "step": 4445
+    },
+    {
+      "epoch": 0.8553701120677216,
+      "grad_norm": 3.3549649559674144,
+      "learning_rate": 1.0782229505670195e-06,
+      "loss": 2.8338,
+      "step": 4446
+    },
+    {
+      "epoch": 0.8555625030061084,
+      "grad_norm": 3.4868656502472546,
+      "learning_rate": 1.0754097479783876e-06,
+      "loss": 2.7392,
+      "step": 4447
+    },
+    {
+      "epoch": 0.8557548939444952,
+      "grad_norm": 3.4489709655668146,
+      "learning_rate": 1.0726000115977696e-06,
+      "loss": 2.7519,
+      "step": 4448
+    },
+    {
+      "epoch": 0.855947284882882,
+      "grad_norm": 3.34054116442144,
+      "learning_rate": 1.069793742516435e-06,
+      "loss": 2.8659,
+      "step": 4449
+    },
+    {
+      "epoch": 0.8561396758212688,
+      "grad_norm": 3.843958970654417,
+      "learning_rate": 1.066990941824312e-06,
+      "loss": 3.0103,
+      "step": 4450
+    },
+    {
+      "epoch": 0.8563320667596557,
+      "grad_norm": 3.5036793205432075,
+      "learning_rate": 1.0641916106099691e-06,
+      "loss": 3.0088,
+      "step": 4451
+    },
+    {
+      "epoch": 0.8565244576980424,
+      "grad_norm": 3.4147784565057933,
+      "learning_rate": 1.0613957499606388e-06,
+      "loss": 2.8081,
+      "step": 4452
+    },
+    {
+      "epoch": 0.8567168486364293,
+      "grad_norm": 3.349820762018809,
+      "learning_rate": 1.0586033609622004e-06,
+      "loss": 3.0659,
+      "step": 4453
+    },
+    {
+      "epoch": 0.856909239574816,
+      "grad_norm": 3.479306747889778,
+      "learning_rate": 1.0558144446991836e-06,
+      "loss": 3.0806,
+      "step": 4454
+    },
+    {
+      "epoch": 0.8571016305132029,
+      "grad_norm": 3.727934953858174,
+      "learning_rate": 1.053029002254773e-06,
+      "loss": 2.9545,
+      "step": 4455
+    },
+    {
+      "epoch": 0.8572940214515896,
+      "grad_norm": 3.5345880362374236,
+      "learning_rate": 1.0502470347108017e-06,
+      "loss": 3.08,
+      "step": 4456
+    },
+    {
+      "epoch": 0.8574864123899765,
+      "grad_norm": 3.4194271690679,
+      "learning_rate": 1.0474685431477537e-06,
+      "loss": 3.0607,
+      "step": 4457
+    },
+    {
+      "epoch": 0.8576788033283632,
+      "grad_norm": 3.4220968035847084,
+      "learning_rate": 1.0446935286447657e-06,
+      "loss": 2.727,
+      "step": 4458
+    },
+    {
+      "epoch": 0.8578711942667501,
+      "grad_norm": 2.5159849335975135,
+      "learning_rate": 1.0419219922796175e-06,
+      "loss": 2.8422,
+      "step": 4459
+    },
+    {
+      "epoch": 0.8580635852051368,
+      "grad_norm": 3.5309465440979646,
+      "learning_rate": 1.039153935128744e-06,
+      "loss": 3.2422,
+      "step": 4460
+    },
+    {
+      "epoch": 0.8582559761435237,
+      "grad_norm": 3.4980021382373625,
+      "learning_rate": 1.0363893582672246e-06,
+      "loss": 2.7808,
+      "step": 4461
+    },
+    {
+      "epoch": 0.8584483670819104,
+      "grad_norm": 3.521412933479982,
+      "learning_rate": 1.033628262768792e-06,
+      "loss": 2.9594,
+      "step": 4462
+    },
+    {
+      "epoch": 0.8586407580202973,
+      "grad_norm": 3.485025513523634,
+      "learning_rate": 1.0308706497058252e-06,
+      "loss": 3.0921,
+      "step": 4463
+    },
+    {
+      "epoch": 0.858833148958684,
+      "grad_norm": 3.4815871792149813,
+      "learning_rate": 1.0281165201493437e-06,
+      "loss": 2.9019,
+      "step": 4464
+    },
+    {
+      "epoch": 0.8590255398970709,
+      "grad_norm": 3.5560574580510167,
+      "learning_rate": 1.0253658751690232e-06,
+      "loss": 2.8465,
+      "step": 4465
+    },
+    {
+      "epoch": 0.8592179308354576,
+      "grad_norm": 3.8346540648157585,
+      "learning_rate": 1.0226187158331825e-06,
+      "loss": 2.8016,
+      "step": 4466
+    },
+    {
+      "epoch": 0.8594103217738445,
+      "grad_norm": 3.5150143794677655,
+      "learning_rate": 1.0198750432087855e-06,
+      "loss": 2.9252,
+      "step": 4467
+    },
+    {
+      "epoch": 0.8596027127122312,
+      "grad_norm": 3.548330152146221,
+      "learning_rate": 1.017134858361446e-06,
+      "loss": 3.0284,
+      "step": 4468
+    },
+    {
+      "epoch": 0.8597951036506181,
+      "grad_norm": 3.73522170779034,
+      "learning_rate": 1.0143981623554155e-06,
+      "loss": 2.7087,
+      "step": 4469
+    },
+    {
+      "epoch": 0.8599874945890048,
+      "grad_norm": 3.7371028000082087,
+      "learning_rate": 1.0116649562535984e-06,
+      "loss": 2.9421,
+      "step": 4470
+    },
+    {
+      "epoch": 0.8601798855273917,
+      "grad_norm": 3.452202229066665,
+      "learning_rate": 1.0089352411175424e-06,
+      "loss": 2.9877,
+      "step": 4471
+    },
+    {
+      "epoch": 0.8603722764657784,
+      "grad_norm": 3.2759988382655116,
+      "learning_rate": 1.006209018007429e-06,
+      "loss": 3.0041,
+      "step": 4472
+    },
+    {
+      "epoch": 0.8605646674041653,
+      "grad_norm": 3.4484033341200915,
+      "learning_rate": 1.0034862879821029e-06,
+      "loss": 2.6991,
+      "step": 4473
+    },
+    {
+      "epoch": 0.8607570583425521,
+      "grad_norm": 3.541273565676268,
+      "learning_rate": 1.0007670520990331e-06,
+      "loss": 2.7307,
+      "step": 4474
+    },
+    {
+      "epoch": 0.8609494492809389,
+      "grad_norm": 3.5937191046476222,
+      "learning_rate": 9.98051311414342e-07,
+      "loss": 3.0169,
+      "step": 4475
+    },
+    {
+      "epoch": 0.8611418402193257,
+      "grad_norm": 3.67830100602681,
+      "learning_rate": 9.953390669827944e-07,
+      "loss": 2.8309,
+      "step": 4476
+    },
+    {
+      "epoch": 0.8613342311577125,
+      "grad_norm": 3.4035417653329882,
+      "learning_rate": 9.926303198577913e-07,
+      "loss": 3.0756,
+      "step": 4477
+    },
+    {
+      "epoch": 0.8615266220960993,
+      "grad_norm": 3.4753905113679266,
+      "learning_rate": 9.899250710913767e-07,
+      "loss": 2.9822,
+      "step": 4478
+    },
+    {
+      "epoch": 0.8617190130344861,
+      "grad_norm": 3.4407776897526308,
+      "learning_rate": 9.872233217342463e-07,
+      "loss": 2.933,
+      "step": 4479
+    },
+    {
+      "epoch": 0.8619114039728729,
+      "grad_norm": 3.8753685871939267,
+      "learning_rate": 9.845250728357214e-07,
+      "loss": 2.9069,
+      "step": 4480
+    },
+    {
+      "epoch": 0.8621037949112597,
+      "grad_norm": 3.3691043074373717,
+      "learning_rate": 9.818303254437723e-07,
+      "loss": 2.4958,
+      "step": 4481
+    },
+    {
+      "epoch": 0.8622961858496465,
+      "grad_norm": 3.4726495256530976,
+      "learning_rate": 9.791390806050117e-07,
+      "loss": 3.0645,
+      "step": 4482
+    },
+    {
+      "epoch": 0.8624885767880333,
+      "grad_norm": 3.3157526949003713,
+      "learning_rate": 9.764513393646812e-07,
+      "loss": 2.8075,
+      "step": 4483
+    },
+    {
+      "epoch": 0.8626809677264201,
+      "grad_norm": 3.498357447579884,
+      "learning_rate": 9.737671027666728e-07,
+      "loss": 2.9813,
+      "step": 4484
+    },
+    {
+      "epoch": 0.8628733586648069,
+      "grad_norm": 3.697160659381192,
+      "learning_rate": 9.710863718535135e-07,
+      "loss": 2.9969,
+      "step": 4485
+    },
+    {
+      "epoch": 0.8630657496031937,
+      "grad_norm": 3.316438134482743,
+      "learning_rate": 9.684091476663659e-07,
+      "loss": 2.8074,
+      "step": 4486
+    },
+    {
+      "epoch": 0.8632581405415805,
+      "grad_norm": 3.48574162838506,
+      "learning_rate": 9.657354312450363e-07,
+      "loss": 2.6614,
+      "step": 4487
+    },
+    {
+      "epoch": 0.8634505314799673,
+      "grad_norm": 3.62288309000448,
+      "learning_rate": 9.630652236279626e-07,
+      "loss": 2.8067,
+      "step": 4488
+    },
+    {
+      "epoch": 0.863642922418354,
+      "grad_norm": 3.4972298852873207,
+      "learning_rate": 9.603985258522219e-07,
+      "loss": 3.0001,
+      "step": 4489
+    },
+    {
+      "epoch": 0.8638353133567409,
+      "grad_norm": 3.5478785018699526,
+      "learning_rate": 9.577353389535315e-07,
+      "loss": 2.8861,
+      "step": 4490
+    },
+    {
+      "epoch": 0.8640277042951277,
+      "grad_norm": 3.2472003079801706,
+      "learning_rate": 9.550756639662417e-07,
+      "loss": 2.7827,
+      "step": 4491
+    },
+    {
+      "epoch": 0.8642200952335145,
+      "grad_norm": 3.6862014347142043,
+      "learning_rate": 9.524195019233407e-07,
+      "loss": 2.7684,
+      "step": 4492
+    },
+    {
+      "epoch": 0.8644124861719013,
+      "grad_norm": 3.4136970038672527,
+      "learning_rate": 9.497668538564475e-07,
+      "loss": 2.9308,
+      "step": 4493
+    },
+    {
+      "epoch": 0.8646048771102881,
+      "grad_norm": 3.424556817708788,
+      "learning_rate": 9.471177207958238e-07,
+      "loss": 2.9723,
+      "step": 4494
+    },
+    {
+      "epoch": 0.864797268048675,
+      "grad_norm": 3.889973436038229,
+      "learning_rate": 9.444721037703597e-07,
+      "loss": 3.1017,
+      "step": 4495
+    },
+    {
+      "epoch": 0.8649896589870617,
+      "grad_norm": 3.6221173760444603,
+      "learning_rate": 9.418300038075845e-07,
+      "loss": 2.8525,
+      "step": 4496
+    },
+    {
+      "epoch": 0.8651820499254486,
+      "grad_norm": 3.2619130733957644,
+      "learning_rate": 9.391914219336606e-07,
+      "loss": 2.7337,
+      "step": 4497
+    },
+    {
+      "epoch": 0.8653744408638353,
+      "grad_norm": 3.4741094259178387,
+      "learning_rate": 9.365563591733784e-07,
+      "loss": 2.919,
+      "step": 4498
+    },
+    {
+      "epoch": 0.8655668318022222,
+      "grad_norm": 2.3536416354473455,
+      "learning_rate": 9.33924816550168e-07,
+      "loss": 2.5332,
+      "step": 4499
+    },
+    {
+      "epoch": 0.8657592227406089,
+      "grad_norm": 3.3693912551272502,
+      "learning_rate": 9.31296795086093e-07,
+      "loss": 3.0725,
+      "step": 4500
+    },
+    {
+      "epoch": 0.8659516136789958,
+      "grad_norm": 3.45455120220468,
+      "learning_rate": 9.286722958018391e-07,
+      "loss": 2.9101,
+      "step": 4501
+    },
+    {
+      "epoch": 0.8661440046173825,
+      "grad_norm": 3.4717267742223257,
+      "learning_rate": 9.260513197167398e-07,
+      "loss": 3.2146,
+      "step": 4502
+    },
+    {
+      "epoch": 0.8663363955557694,
+      "grad_norm": 3.394600476715909,
+      "learning_rate": 9.234338678487509e-07,
+      "loss": 2.8844,
+      "step": 4503
+    },
+    {
+      "epoch": 0.8665287864941561,
+      "grad_norm": 3.324367143937873,
+      "learning_rate": 9.208199412144559e-07,
+      "loss": 2.9919,
+      "step": 4504
+    },
+    {
+      "epoch": 0.866721177432543,
+      "grad_norm": 3.225586401314347,
+      "learning_rate": 9.182095408290781e-07,
+      "loss": 2.737,
+      "step": 4505
+    },
+    {
+      "epoch": 0.8669135683709297,
+      "grad_norm": 3.660443643710109,
+      "learning_rate": 9.156026677064633e-07,
+      "loss": 2.8604,
+      "step": 4506
+    },
+    {
+      "epoch": 0.8671059593093166,
+      "grad_norm": 3.558719547691315,
+      "learning_rate": 9.129993228590917e-07,
+      "loss": 2.9963,
+      "step": 4507
+    },
+    {
+      "epoch": 0.8672983502477033,
+      "grad_norm": 3.233451514764446,
+      "learning_rate": 9.103995072980765e-07,
+      "loss": 2.6742,
+      "step": 4508
+    },
+    {
+      "epoch": 0.8674907411860902,
+      "grad_norm": 3.6010778858994343,
+      "learning_rate": 9.078032220331523e-07,
+      "loss": 3.1123,
+      "step": 4509
+    },
+    {
+      "epoch": 0.8676831321244769,
+      "grad_norm": 3.4619045891857807,
+      "learning_rate": 9.052104680726859e-07,
+      "loss": 3.1434,
+      "step": 4510
+    },
+    {
+      "epoch": 0.8678755230628638,
+      "grad_norm": 3.680000323446017,
+      "learning_rate": 9.026212464236772e-07,
+      "loss": 2.971,
+      "step": 4511
+    },
+    {
+      "epoch": 0.8680679140012505,
+      "grad_norm": 3.5327080396051564,
+      "learning_rate": 9.000355580917464e-07,
+      "loss": 2.8418,
+      "step": 4512
+    },
+    {
+      "epoch": 0.8682603049396374,
+      "grad_norm": 3.4289567687973226,
+      "learning_rate": 8.974534040811444e-07,
+      "loss": 2.9581,
+      "step": 4513
+    },
+    {
+      "epoch": 0.8684526958780241,
+      "grad_norm": 3.4230943234183107,
+      "learning_rate": 8.948747853947526e-07,
+      "loss": 2.888,
+      "step": 4514
+    },
+    {
+      "epoch": 0.868645086816411,
+      "grad_norm": 3.678741186062204,
+      "learning_rate": 8.922997030340752e-07,
+      "loss": 2.941,
+      "step": 4515
+    },
+    {
+      "epoch": 0.8688374777547978,
+      "grad_norm": 3.5817566727268453,
+      "learning_rate": 8.897281579992467e-07,
+      "loss": 3.0069,
+      "step": 4516
+    },
+    {
+      "epoch": 0.8690298686931845,
+      "grad_norm": 3.3759838136497176,
+      "learning_rate": 8.871601512890238e-07,
+      "loss": 3.0906,
+      "step": 4517
+    },
+    {
+      "epoch": 0.8692222596315714,
+      "grad_norm": 3.528980935411663,
+      "learning_rate": 8.845956839007897e-07,
+      "loss": 3.1817,
+      "step": 4518
+    },
+    {
+      "epoch": 0.8694146505699581,
+      "grad_norm": 3.6191981323898283,
+      "learning_rate": 8.820347568305543e-07,
+      "loss": 2.8785,
+      "step": 4519
+    },
+    {
+      "epoch": 0.869607041508345,
+      "grad_norm": 3.346027729625762,
+      "learning_rate": 8.794773710729543e-07,
+      "loss": 2.8475,
+      "step": 4520
+    },
+    {
+      "epoch": 0.8697994324467317,
+      "grad_norm": 3.381810443478603,
+      "learning_rate": 8.769235276212496e-07,
+      "loss": 2.8156,
+      "step": 4521
+    },
+    {
+      "epoch": 0.8699918233851186,
+      "grad_norm": 3.203153560650709,
+      "learning_rate": 8.743732274673189e-07,
+      "loss": 2.998,
+      "step": 4522
+    },
+    {
+      "epoch": 0.8701842143235053,
+      "grad_norm": 3.354957674729755,
+      "learning_rate": 8.718264716016722e-07,
+      "loss": 2.7546,
+      "step": 4523
+    },
+    {
+      "epoch": 0.8703766052618922,
+      "grad_norm": 3.6503555629582847,
+      "learning_rate": 8.692832610134428e-07,
+      "loss": 3.174,
+      "step": 4524
+    },
+    {
+      "epoch": 0.8705689962002789,
+      "grad_norm": 3.2892002897101666,
+      "learning_rate": 8.66743596690377e-07,
+      "loss": 3.0455,
+      "step": 4525
+    },
+    {
+      "epoch": 0.8707613871386658,
+      "grad_norm": 3.554313553166474,
+      "learning_rate": 8.642074796188594e-07,
+      "loss": 3.1063,
+      "step": 4526
+    },
+    {
+      "epoch": 0.8709537780770525,
+      "grad_norm": 3.463817954526927,
+      "learning_rate": 8.61674910783884e-07,
+      "loss": 3.0681,
+      "step": 4527
+    },
+    {
+      "epoch": 0.8711461690154394,
+      "grad_norm": 3.4706822117509124,
+      "learning_rate": 8.59145891169072e-07,
+      "loss": 2.9145,
+      "step": 4528
+    },
+    {
+      "epoch": 0.8713385599538261,
+      "grad_norm": 3.8254723467286924,
+      "learning_rate": 8.566204217566664e-07,
+      "loss": 3.0548,
+      "step": 4529
+    },
+    {
+      "epoch": 0.871530950892213,
+      "grad_norm": 3.195432388716343,
+      "learning_rate": 8.540985035275273e-07,
+      "loss": 2.5568,
+      "step": 4530
+    },
+    {
+      "epoch": 0.8717233418305997,
+      "grad_norm": 3.4480977652745217,
+      "learning_rate": 8.515801374611432e-07,
+      "loss": 3.1263,
+      "step": 4531
+    },
+    {
+      "epoch": 0.8719157327689866,
+      "grad_norm": 3.8556914726009923,
+      "learning_rate": 8.490653245356184e-07,
+      "loss": 3.1073,
+      "step": 4532
+    },
+    {
+      "epoch": 0.8721081237073733,
+      "grad_norm": 3.4011136774266295,
+      "learning_rate": 8.465540657276728e-07,
+      "loss": 2.8493,
+      "step": 4533
+    },
+    {
+      "epoch": 0.8723005146457602,
+      "grad_norm": 2.3355440998460986,
+      "learning_rate": 8.44046362012656e-07,
+      "loss": 2.594,
+      "step": 4534
+    },
+    {
+      "epoch": 0.8724929055841469,
+      "grad_norm": 3.58720030318679,
+      "learning_rate": 8.415422143645247e-07,
+      "loss": 3.1639,
+      "step": 4535
+    },
+    {
+      "epoch": 0.8726852965225338,
+      "grad_norm": 3.6125889154218327,
+      "learning_rate": 8.390416237558641e-07,
+      "loss": 2.8768,
+      "step": 4536
+    },
+    {
+      "epoch": 0.8728776874609205,
+      "grad_norm": 3.542140719144097,
+      "learning_rate": 8.365445911578785e-07,
+      "loss": 2.91,
+      "step": 4537
+    },
+    {
+      "epoch": 0.8730700783993074,
+      "grad_norm": 3.340733186584088,
+      "learning_rate": 8.340511175403809e-07,
+      "loss": 2.8592,
+      "step": 4538
+    },
+    {
+      "epoch": 0.8732624693376942,
+      "grad_norm": 3.4238610001093446,
+      "learning_rate": 8.315612038718101e-07,
+      "loss": 2.8729,
+      "step": 4539
+    },
+    {
+      "epoch": 0.873454860276081,
+      "grad_norm": 3.56393241286816,
+      "learning_rate": 8.290748511192214e-07,
+      "loss": 3.0675,
+      "step": 4540
+    },
+    {
+      "epoch": 0.8736472512144678,
+      "grad_norm": 3.3744931973799,
+      "learning_rate": 8.265920602482825e-07,
+      "loss": 2.9389,
+      "step": 4541
+    },
+    {
+      "epoch": 0.8738396421528546,
+      "grad_norm": 3.614236203722177,
+      "learning_rate": 8.241128322232816e-07,
+      "loss": 2.9182,
+      "step": 4542
+    },
+    {
+      "epoch": 0.8740320330912414,
+      "grad_norm": 3.455205239140729,
+      "learning_rate": 8.216371680071244e-07,
+      "loss": 2.9059,
+      "step": 4543
+    },
+    {
+      "epoch": 0.8742244240296282,
+      "grad_norm": 3.7145365968393866,
+      "learning_rate": 8.191650685613273e-07,
+      "loss": 2.7196,
+      "step": 4544
+    },
+    {
+      "epoch": 0.874416814968015,
+      "grad_norm": 3.616651685608736,
+      "learning_rate": 8.166965348460298e-07,
+      "loss": 2.9741,
+      "step": 4545
+    },
+    {
+      "epoch": 0.8746092059064018,
+      "grad_norm": 3.70196401417676,
+      "learning_rate": 8.142315678199764e-07,
+      "loss": 3.0041,
+      "step": 4546
+    },
+    {
+      "epoch": 0.8748015968447886,
+      "grad_norm": 3.415257446401183,
+      "learning_rate": 8.117701684405343e-07,
+      "loss": 3.0647,
+      "step": 4547
+    },
+    {
+      "epoch": 0.8749939877831754,
+      "grad_norm": 3.433519472357566,
+      "learning_rate": 8.093123376636836e-07,
+      "loss": 2.8544,
+      "step": 4548
+    },
+    {
+      "epoch": 0.8751863787215622,
+      "grad_norm": 3.299064929921268,
+      "learning_rate": 8.06858076444017e-07,
+      "loss": 2.8564,
+      "step": 4549
+    },
+    {
+      "epoch": 0.875378769659949,
+      "grad_norm": 3.213973525389161,
+      "learning_rate": 8.044073857347423e-07,
+      "loss": 2.8439,
+      "step": 4550
+    },
+    {
+      "epoch": 0.8755711605983358,
+      "grad_norm": 3.326105307442767,
+      "learning_rate": 8.019602664876758e-07,
+      "loss": 2.9722,
+      "step": 4551
+    },
+    {
+      "epoch": 0.8757635515367226,
+      "grad_norm": 3.446133038664924,
+      "learning_rate": 7.995167196532527e-07,
+      "loss": 2.9203,
+      "step": 4552
+    },
+    {
+      "epoch": 0.8759559424751094,
+      "grad_norm": 3.672770602421332,
+      "learning_rate": 7.970767461805218e-07,
+      "loss": 2.7397,
+      "step": 4553
+    },
+    {
+      "epoch": 0.8761483334134962,
+      "grad_norm": 3.1594662670613776,
+      "learning_rate": 7.946403470171326e-07,
+      "loss": 2.8093,
+      "step": 4554
+    },
+    {
+      "epoch": 0.876340724351883,
+      "grad_norm": 3.323011358909289,
+      "learning_rate": 7.922075231093628e-07,
+      "loss": 2.676,
+      "step": 4555
+    },
+    {
+      "epoch": 0.8765331152902698,
+      "grad_norm": 3.3498831309294936,
+      "learning_rate": 7.897782754020889e-07,
+      "loss": 2.776,
+      "step": 4556
+    },
+    {
+      "epoch": 0.8767255062286566,
+      "grad_norm": 3.3365585502448045,
+      "learning_rate": 7.873526048388025e-07,
+      "loss": 2.9438,
+      "step": 4557
+    },
+    {
+      "epoch": 0.8769178971670434,
+      "grad_norm": 3.4150411659327378,
+      "learning_rate": 7.849305123616091e-07,
+      "loss": 2.8371,
+      "step": 4558
+    },
+    {
+      "epoch": 0.8771102881054302,
+      "grad_norm": 3.6691464732918972,
+      "learning_rate": 7.825119989112173e-07,
+      "loss": 2.9998,
+      "step": 4559
+    },
+    {
+      "epoch": 0.8773026790438171,
+      "grad_norm": 3.5291661283103832,
+      "learning_rate": 7.800970654269513e-07,
+      "loss": 2.9149,
+      "step": 4560
+    },
+    {
+      "epoch": 0.8774950699822038,
+      "grad_norm": 3.868219519020724,
+      "learning_rate": 7.776857128467464e-07,
+      "loss": 2.9786,
+      "step": 4561
+    },
+    {
+      "epoch": 0.8776874609205907,
+      "grad_norm": 3.356558960282919,
+      "learning_rate": 7.75277942107141e-07,
+      "loss": 2.9695,
+      "step": 4562
+    },
+    {
+      "epoch": 0.8778798518589774,
+      "grad_norm": 3.5449886476970045,
+      "learning_rate": 7.728737541432862e-07,
+      "loss": 2.7058,
+      "step": 4563
+    },
+    {
+      "epoch": 0.8780722427973643,
+      "grad_norm": 3.189741116596015,
+      "learning_rate": 7.704731498889428e-07,
+      "loss": 2.8279,
+      "step": 4564
+    },
+    {
+      "epoch": 0.878264633735751,
+      "grad_norm": 3.3333102970664448,
+      "learning_rate": 7.680761302764727e-07,
+      "loss": 2.7732,
+      "step": 4565
+    },
+    {
+      "epoch": 0.8784570246741379,
+      "grad_norm": 3.3278795278001017,
+      "learning_rate": 7.65682696236858e-07,
+      "loss": 2.3931,
+      "step": 4566
+    },
+    {
+      "epoch": 0.8786494156125246,
+      "grad_norm": 3.4587910212429813,
+      "learning_rate": 7.632928486996749e-07,
+      "loss": 2.613,
+      "step": 4567
+    },
+    {
+      "epoch": 0.8788418065509115,
+      "grad_norm": 3.2617607889559515,
+      "learning_rate": 7.609065885931155e-07,
+      "loss": 2.8438,
+      "step": 4568
+    },
+    {
+      "epoch": 0.8790341974892982,
+      "grad_norm": 3.613158586549939,
+      "learning_rate": 7.585239168439762e-07,
+      "loss": 3.0675,
+      "step": 4569
+    },
+    {
+      "epoch": 0.8792265884276851,
+      "grad_norm": 3.667106550158173,
+      "learning_rate": 7.561448343776567e-07,
+      "loss": 2.9273,
+      "step": 4570
+    },
+    {
+      "epoch": 0.8794189793660718,
+      "grad_norm": 3.092253085888423,
+      "learning_rate": 7.537693421181658e-07,
+      "loss": 2.7305,
+      "step": 4571
+    },
+    {
+      "epoch": 0.8796113703044587,
+      "grad_norm": 3.210602837106173,
+      "learning_rate": 7.513974409881186e-07,
+      "loss": 2.8921,
+      "step": 4572
+    },
+    {
+      "epoch": 0.8798037612428454,
+      "grad_norm": 3.8025032107275263,
+      "learning_rate": 7.49029131908734e-07,
+      "loss": 3.2728,
+      "step": 4573
+    },
+    {
+      "epoch": 0.8799961521812323,
+      "grad_norm": 3.490795476963027,
+      "learning_rate": 7.466644157998371e-07,
+      "loss": 3.1601,
+      "step": 4574
+    },
+    {
+      "epoch": 0.880188543119619,
+      "grad_norm": 3.3544252039466835,
+      "learning_rate": 7.443032935798533e-07,
+      "loss": 3.0102,
+      "step": 4575
+    },
+    {
+      "epoch": 0.8803809340580059,
+      "grad_norm": 3.469488687285991,
+      "learning_rate": 7.419457661658169e-07,
+      "loss": 3.0946,
+      "step": 4576
+    },
+    {
+      "epoch": 0.8805733249963926,
+      "grad_norm": 3.1868336111062434,
+      "learning_rate": 7.395918344733644e-07,
+      "loss": 2.9363,
+      "step": 4577
+    },
+    {
+      "epoch": 0.8807657159347795,
+      "grad_norm": 3.857821701805082,
+      "learning_rate": 7.372414994167354e-07,
+      "loss": 2.938,
+      "step": 4578
+    },
+    {
+      "epoch": 0.8809581068731662,
+      "grad_norm": 3.4126963537499666,
+      "learning_rate": 7.348947619087754e-07,
+      "loss": 2.7825,
+      "step": 4579
+    },
+    {
+      "epoch": 0.8811504978115531,
+      "grad_norm": 3.2441640603740907,
+      "learning_rate": 7.325516228609264e-07,
+      "loss": 2.7462,
+      "step": 4580
+    },
+    {
+      "epoch": 0.8813428887499398,
+      "grad_norm": 3.5559656745654356,
+      "learning_rate": 7.302120831832382e-07,
+      "loss": 2.9726,
+      "step": 4581
+    },
+    {
+      "epoch": 0.8815352796883267,
+      "grad_norm": 3.528654438305884,
+      "learning_rate": 7.278761437843629e-07,
+      "loss": 2.9651,
+      "step": 4582
+    },
+    {
+      "epoch": 0.8817276706267135,
+      "grad_norm": 3.5718845234875105,
+      "learning_rate": 7.255438055715469e-07,
+      "loss": 3.1686,
+      "step": 4583
+    },
+    {
+      "epoch": 0.8819200615651003,
+      "grad_norm": 3.579986658549545,
+      "learning_rate": 7.232150694506512e-07,
+      "loss": 2.8641,
+      "step": 4584
+    },
+    {
+      "epoch": 0.8821124525034871,
+      "grad_norm": 3.430690652761111,
+      "learning_rate": 7.208899363261234e-07,
+      "loss": 2.5923,
+      "step": 4585
+    },
+    {
+      "epoch": 0.8823048434418739,
+      "grad_norm": 3.353039388001949,
+      "learning_rate": 7.185684071010224e-07,
+      "loss": 2.8062,
+      "step": 4586
+    },
+    {
+      "epoch": 0.8824972343802607,
+      "grad_norm": 3.576391307812584,
+      "learning_rate": 7.162504826770033e-07,
+      "loss": 2.8853,
+      "step": 4587
+    },
+    {
+      "epoch": 0.8826896253186475,
+      "grad_norm": 3.4746028894852614,
+      "learning_rate": 7.139361639543185e-07,
+      "loss": 2.7957,
+      "step": 4588
+    },
+    {
+      "epoch": 0.8828820162570343,
+      "grad_norm": 3.2563703686363263,
+      "learning_rate": 7.116254518318222e-07,
+      "loss": 3.0371,
+      "step": 4589
+    },
+    {
+      "epoch": 0.8830744071954211,
+      "grad_norm": 3.328468401151844,
+      "learning_rate": 7.093183472069753e-07,
+      "loss": 2.8995,
+      "step": 4590
+    },
+    {
+      "epoch": 0.8832667981338079,
+      "grad_norm": 3.4617344827677874,
+      "learning_rate": 7.070148509758223e-07,
+      "loss": 2.9092,
+      "step": 4591
+    },
+    {
+      "epoch": 0.8834591890721947,
+      "grad_norm": 3.5384695326833033,
+      "learning_rate": 7.047149640330197e-07,
+      "loss": 2.9159,
+      "step": 4592
+    },
+    {
+      "epoch": 0.8836515800105815,
+      "grad_norm": 3.2791540877016803,
+      "learning_rate": 7.024186872718164e-07,
+      "loss": 2.8672,
+      "step": 4593
+    },
+    {
+      "epoch": 0.8838439709489683,
+      "grad_norm": 3.5696549279750505,
+      "learning_rate": 7.001260215840567e-07,
+      "loss": 2.8823,
+      "step": 4594
+    },
+    {
+      "epoch": 0.8840363618873551,
+      "grad_norm": 3.8245381308448017,
+      "learning_rate": 6.978369678601892e-07,
+      "loss": 2.7119,
+      "step": 4595
+    },
+    {
+      "epoch": 0.8842287528257419,
+      "grad_norm": 3.3553791159398623,
+      "learning_rate": 6.955515269892533e-07,
+      "loss": 2.8133,
+      "step": 4596
+    },
+    {
+      "epoch": 0.8844211437641287,
+      "grad_norm": 3.341260366735051,
+      "learning_rate": 6.932696998588895e-07,
+      "loss": 2.9731,
+      "step": 4597
+    },
+    {
+      "epoch": 0.8846135347025155,
+      "grad_norm": 3.3415979825870674,
+      "learning_rate": 6.909914873553347e-07,
+      "loss": 2.9461,
+      "step": 4598
+    },
+    {
+      "epoch": 0.8848059256409023,
+      "grad_norm": 3.431742714466809,
+      "learning_rate": 6.887168903634178e-07,
+      "loss": 2.9079,
+      "step": 4599
+    },
+    {
+      "epoch": 0.8849983165792891,
+      "grad_norm": 3.574728352211099,
+      "learning_rate": 6.864459097665654e-07,
+      "loss": 2.9567,
+      "step": 4600
+    },
+    {
+      "epoch": 0.8851907075176759,
+      "grad_norm": 3.3507575127224283,
+      "learning_rate": 6.84178546446802e-07,
+      "loss": 3.0646,
+      "step": 4601
+    },
+    {
+      "epoch": 0.8853830984560627,
+      "grad_norm": 3.4056266151455583,
+      "learning_rate": 6.819148012847454e-07,
+      "loss": 2.9275,
+      "step": 4602
+    },
+    {
+      "epoch": 0.8855754893944495,
+      "grad_norm": 3.35668268448188,
+      "learning_rate": 6.796546751596089e-07,
+      "loss": 2.873,
+      "step": 4603
+    },
+    {
+      "epoch": 0.8857678803328364,
+      "grad_norm": 3.3337780448172216,
+      "learning_rate": 6.77398168949196e-07,
+      "loss": 2.8257,
+      "step": 4604
+    },
+    {
+      "epoch": 0.8859602712712231,
+      "grad_norm": 3.4586507767807237,
+      "learning_rate": 6.751452835299111e-07,
+      "loss": 2.9486,
+      "step": 4605
+    },
+    {
+      "epoch": 0.88615266220961,
+      "grad_norm": 3.552762717900001,
+      "learning_rate": 6.728960197767475e-07,
+      "loss": 3.092,
+      "step": 4606
+    },
+    {
+      "epoch": 0.8863450531479967,
+      "grad_norm": 3.5838874246256607,
+      "learning_rate": 6.706503785632934e-07,
+      "loss": 2.8542,
+      "step": 4607
+    },
+    {
+      "epoch": 0.8865374440863836,
+      "grad_norm": 3.2618887309866413,
+      "learning_rate": 6.68408360761732e-07,
+      "loss": 2.9956,
+      "step": 4608
+    },
+    {
+      "epoch": 0.8867298350247703,
+      "grad_norm": 2.245526379066973,
+      "learning_rate": 6.661699672428334e-07,
+      "loss": 2.5391,
+      "step": 4609
+    },
+    {
+      "epoch": 0.8869222259631572,
+      "grad_norm": 3.4269977085366876,
+      "learning_rate": 6.639351988759657e-07,
+      "loss": 2.8864,
+      "step": 4610
+    },
+    {
+      "epoch": 0.8871146169015439,
+      "grad_norm": 2.3449677529990716,
+      "learning_rate": 6.6170405652909e-07,
+      "loss": 2.5327,
+      "step": 4611
+    },
+    {
+      "epoch": 0.8873070078399308,
+      "grad_norm": 3.257258925607618,
+      "learning_rate": 6.594765410687487e-07,
+      "loss": 2.6637,
+      "step": 4612
+    },
+    {
+      "epoch": 0.8874993987783175,
+      "grad_norm": 3.6917552764427692,
+      "learning_rate": 6.57252653360092e-07,
+      "loss": 2.9617,
+      "step": 4613
+    },
+    {
+      "epoch": 0.8876917897167044,
+      "grad_norm": 3.7538032382865696,
+      "learning_rate": 6.550323942668469e-07,
+      "loss": 2.8921,
+      "step": 4614
+    },
+    {
+      "epoch": 0.8878841806550911,
+      "grad_norm": 3.501278768881063,
+      "learning_rate": 6.528157646513378e-07,
+      "loss": 2.9858,
+      "step": 4615
+    },
+    {
+      "epoch": 0.888076571593478,
+      "grad_norm": 4.032817589092647,
+      "learning_rate": 6.506027653744796e-07,
+      "loss": 2.9418,
+      "step": 4616
+    },
+    {
+      "epoch": 0.8882689625318647,
+      "grad_norm": 3.5269334885482007,
+      "learning_rate": 6.483933972957734e-07,
+      "loss": 2.8819,
+      "step": 4617
+    },
+    {
+      "epoch": 0.8884613534702516,
+      "grad_norm": 3.971253033165648,
+      "learning_rate": 6.461876612733109e-07,
+      "loss": 2.9411,
+      "step": 4618
+    },
+    {
+      "epoch": 0.8886537444086383,
+      "grad_norm": 2.102210846962537,
+      "learning_rate": 6.4398555816378e-07,
+      "loss": 2.5213,
+      "step": 4619
+    },
+    {
+      "epoch": 0.8888461353470252,
+      "grad_norm": 3.3941341361752166,
+      "learning_rate": 6.417870888224476e-07,
+      "loss": 2.9689,
+      "step": 4620
+    },
+    {
+      "epoch": 0.8890385262854119,
+      "grad_norm": 3.6334727547283006,
+      "learning_rate": 6.395922541031741e-07,
+      "loss": 2.9824,
+      "step": 4621
+    },
+    {
+      "epoch": 0.8892309172237988,
+      "grad_norm": 3.4898940697759695,
+      "learning_rate": 6.374010548584119e-07,
+      "loss": 2.9698,
+      "step": 4622
+    },
+    {
+      "epoch": 0.8894233081621855,
+      "grad_norm": 3.5364638422114356,
+      "learning_rate": 6.352134919391928e-07,
+      "loss": 2.9882,
+      "step": 4623
+    },
+    {
+      "epoch": 0.8896156991005724,
+      "grad_norm": 3.6357875648064204,
+      "learning_rate": 6.330295661951436e-07,
+      "loss": 2.9514,
+      "step": 4624
+    },
+    {
+      "epoch": 0.8898080900389591,
+      "grad_norm": 3.465705542442403,
+      "learning_rate": 6.308492784744746e-07,
+      "loss": 2.9295,
+      "step": 4625
+    },
+    {
+      "epoch": 0.890000480977346,
+      "grad_norm": 3.4031534921396864,
+      "learning_rate": 6.286726296239854e-07,
+      "loss": 2.6775,
+      "step": 4626
+    },
+    {
+      "epoch": 0.8901928719157328,
+      "grad_norm": 3.426430275160139,
+      "learning_rate": 6.264996204890628e-07,
+      "loss": 2.8701,
+      "step": 4627
+    },
+    {
+      "epoch": 0.8903852628541196,
+      "grad_norm": 3.7067251655947233,
+      "learning_rate": 6.24330251913674e-07,
+      "loss": 2.8152,
+      "step": 4628
+    },
+    {
+      "epoch": 0.8905776537925064,
+      "grad_norm": 3.3089598627430505,
+      "learning_rate": 6.221645247403807e-07,
+      "loss": 2.9435,
+      "step": 4629
+    },
+    {
+      "epoch": 0.8907700447308932,
+      "grad_norm": 3.311507320449331,
+      "learning_rate": 6.200024398103255e-07,
+      "loss": 2.864,
+      "step": 4630
+    },
+    {
+      "epoch": 0.89096243566928,
+      "grad_norm": 3.5773490784806565,
+      "learning_rate": 6.178439979632367e-07,
+      "loss": 2.7738,
+      "step": 4631
+    },
+    {
+      "epoch": 0.8911548266076668,
+      "grad_norm": 3.467878722523039,
+      "learning_rate": 6.156892000374293e-07,
+      "loss": 3.0565,
+      "step": 4632
+    },
+    {
+      "epoch": 0.8913472175460536,
+      "grad_norm": 3.7892939444824525,
+      "learning_rate": 6.135380468698004e-07,
+      "loss": 2.9375,
+      "step": 4633
+    },
+    {
+      "epoch": 0.8915396084844404,
+      "grad_norm": 3.387504781803891,
+      "learning_rate": 6.113905392958342e-07,
+      "loss": 3.1644,
+      "step": 4634
+    },
+    {
+      "epoch": 0.8917319994228272,
+      "grad_norm": 3.536514948682397,
+      "learning_rate": 6.092466781495976e-07,
+      "loss": 2.954,
+      "step": 4635
+    },
+    {
+      "epoch": 0.891924390361214,
+      "grad_norm": 3.240411656199628,
+      "learning_rate": 6.071064642637404e-07,
+      "loss": 2.7286,
+      "step": 4636
+    },
+    {
+      "epoch": 0.8921167812996008,
+      "grad_norm": 3.5719151431266765,
+      "learning_rate": 6.049698984695007e-07,
+      "loss": 2.9454,
+      "step": 4637
+    },
+    {
+      "epoch": 0.8923091722379876,
+      "grad_norm": 3.5823285250685237,
+      "learning_rate": 6.028369815966917e-07,
+      "loss": 3.0964,
+      "step": 4638
+    },
+    {
+      "epoch": 0.8925015631763744,
+      "grad_norm": 3.5554654695877863,
+      "learning_rate": 6.007077144737161e-07,
+      "loss": 2.8571,
+      "step": 4639
+    },
+    {
+      "epoch": 0.8926939541147612,
+      "grad_norm": 3.3625225599095057,
+      "learning_rate": 5.985820979275569e-07,
+      "loss": 2.9852,
+      "step": 4640
+    },
+    {
+      "epoch": 0.892886345053148,
+      "grad_norm": 3.5300068933931756,
+      "learning_rate": 5.96460132783776e-07,
+      "loss": 2.8696,
+      "step": 4641
+    },
+    {
+      "epoch": 0.8930787359915348,
+      "grad_norm": 2.5062416565385464,
+      "learning_rate": 5.943418198665251e-07,
+      "loss": 2.6462,
+      "step": 4642
+    },
+    {
+      "epoch": 0.8932711269299216,
+      "grad_norm": 3.285925977714323,
+      "learning_rate": 5.922271599985286e-07,
+      "loss": 2.8676,
+      "step": 4643
+    },
+    {
+      "epoch": 0.8934635178683084,
+      "grad_norm": 3.494763134347294,
+      "learning_rate": 5.901161540010969e-07,
+      "loss": 3.064,
+      "step": 4644
+    },
+    {
+      "epoch": 0.8936559088066952,
+      "grad_norm": 3.637889631693448,
+      "learning_rate": 5.880088026941233e-07,
+      "loss": 2.7702,
+      "step": 4645
+    },
+    {
+      "epoch": 0.893848299745082,
+      "grad_norm": 3.47278220669439,
+      "learning_rate": 5.859051068960741e-07,
+      "loss": 2.9508,
+      "step": 4646
+    },
+    {
+      "epoch": 0.8940406906834688,
+      "grad_norm": 3.591134205090606,
+      "learning_rate": 5.838050674240025e-07,
+      "loss": 3.0271,
+      "step": 4647
+    },
+    {
+      "epoch": 0.8942330816218557,
+      "grad_norm": 3.7377978557832163,
+      "learning_rate": 5.817086850935416e-07,
+      "loss": 2.9597,
+      "step": 4648
+    },
+    {
+      "epoch": 0.8944254725602424,
+      "grad_norm": 3.369936929357503,
+      "learning_rate": 5.796159607189001e-07,
+      "loss": 2.8485,
+      "step": 4649
+    },
+    {
+      "epoch": 0.8946178634986293,
+      "grad_norm": 3.538355962035486,
+      "learning_rate": 5.775268951128676e-07,
+      "loss": 3.0995,
+      "step": 4650
+    },
+    {
+      "epoch": 0.894810254437016,
+      "grad_norm": 3.524712448266645,
+      "learning_rate": 5.754414890868154e-07,
+      "loss": 2.9919,
+      "step": 4651
+    },
+    {
+      "epoch": 0.8950026453754029,
+      "grad_norm": 3.4872542005454648,
+      "learning_rate": 5.733597434506877e-07,
+      "loss": 2.9428,
+      "step": 4652
+    },
+    {
+      "epoch": 0.8951950363137896,
+      "grad_norm": 3.244982431393289,
+      "learning_rate": 5.712816590130133e-07,
+      "loss": 2.7476,
+      "step": 4653
+    },
+    {
+      "epoch": 0.8953874272521765,
+      "grad_norm": 3.6353156858066638,
+      "learning_rate": 5.692072365808954e-07,
+      "loss": 3.1217,
+      "step": 4654
+    },
+    {
+      "epoch": 0.8955798181905632,
+      "grad_norm": 3.3201237052401957,
+      "learning_rate": 5.671364769600162e-07,
+      "loss": 3.0207,
+      "step": 4655
+    },
+    {
+      "epoch": 0.8957722091289501,
+      "grad_norm": 3.5088906606752497,
+      "learning_rate": 5.650693809546348e-07,
+      "loss": 2.738,
+      "step": 4656
+    },
+    {
+      "epoch": 0.8959646000673368,
+      "grad_norm": 3.3084488349219416,
+      "learning_rate": 5.630059493675866e-07,
+      "loss": 2.7429,
+      "step": 4657
+    },
+    {
+      "epoch": 0.8961569910057237,
+      "grad_norm": 3.3844999841945294,
+      "learning_rate": 5.60946183000285e-07,
+      "loss": 2.8676,
+      "step": 4658
+    },
+    {
+      "epoch": 0.8963493819441104,
+      "grad_norm": 3.2253431937650006,
+      "learning_rate": 5.5889008265272e-07,
+      "loss": 3.0246,
+      "step": 4659
+    },
+    {
+      "epoch": 0.8965417728824973,
+      "grad_norm": 3.4369427749511945,
+      "learning_rate": 5.568376491234562e-07,
+      "loss": 2.9021,
+      "step": 4660
+    },
+    {
+      "epoch": 0.896734163820884,
+      "grad_norm": 3.7255720289120062,
+      "learning_rate": 5.547888832096382e-07,
+      "loss": 2.9128,
+      "step": 4661
+    },
+    {
+      "epoch": 0.8969265547592709,
+      "grad_norm": 3.5648084508916127,
+      "learning_rate": 5.527437857069784e-07,
+      "loss": 2.959,
+      "step": 4662
+    },
+    {
+      "epoch": 0.8971189456976576,
+      "grad_norm": 3.4208244515115624,
+      "learning_rate": 5.507023574097725e-07,
+      "loss": 3.0962,
+      "step": 4663
+    },
+    {
+      "epoch": 0.8973113366360445,
+      "grad_norm": 3.4754030611610034,
+      "learning_rate": 5.486645991108875e-07,
+      "loss": 2.6427,
+      "step": 4664
+    },
+    {
+      "epoch": 0.8975037275744312,
+      "grad_norm": 3.468395782845079,
+      "learning_rate": 5.466305116017623e-07,
+      "loss": 3.0246,
+      "step": 4665
+    },
+    {
+      "epoch": 0.897696118512818,
+      "grad_norm": 3.478700691368497,
+      "learning_rate": 5.446000956724174e-07,
+      "loss": 3.1207,
+      "step": 4666
+    },
+    {
+      "epoch": 0.8978885094512048,
+      "grad_norm": 3.4296696023820172,
+      "learning_rate": 5.425733521114396e-07,
+      "loss": 3.013,
+      "step": 4667
+    },
+    {
+      "epoch": 0.8980809003895917,
+      "grad_norm": 3.3709066040102242,
+      "learning_rate": 5.405502817059937e-07,
+      "loss": 2.9848,
+      "step": 4668
+    },
+    {
+      "epoch": 0.8982732913279785,
+      "grad_norm": 3.6440945437496506,
+      "learning_rate": 5.385308852418191e-07,
+      "loss": 2.9683,
+      "step": 4669
+    },
+    {
+      "epoch": 0.8984656822663653,
+      "grad_norm": 3.532262805715973,
+      "learning_rate": 5.365151635032218e-07,
+      "loss": 2.7223,
+      "step": 4670
+    },
+    {
+      "epoch": 0.8986580732047521,
+      "grad_norm": 3.365272240366215,
+      "learning_rate": 5.345031172730875e-07,
+      "loss": 2.7212,
+      "step": 4671
+    },
+    {
+      "epoch": 0.8988504641431388,
+      "grad_norm": 3.402886437187726,
+      "learning_rate": 5.324947473328735e-07,
+      "loss": 2.76,
+      "step": 4672
+    },
+    {
+      "epoch": 0.8990428550815257,
+      "grad_norm": 3.6231540515815617,
+      "learning_rate": 5.304900544626046e-07,
+      "loss": 3.1313,
+      "step": 4673
+    },
+    {
+      "epoch": 0.8992352460199124,
+      "grad_norm": 3.531316617223311,
+      "learning_rate": 5.284890394408826e-07,
+      "loss": 2.6915,
+      "step": 4674
+    },
+    {
+      "epoch": 0.8994276369582993,
+      "grad_norm": 3.488151539507205,
+      "learning_rate": 5.264917030448757e-07,
+      "loss": 2.8244,
+      "step": 4675
+    },
+    {
+      "epoch": 0.899620027896686,
+      "grad_norm": 3.367652343908863,
+      "learning_rate": 5.244980460503268e-07,
+      "loss": 2.9952,
+      "step": 4676
+    },
+    {
+      "epoch": 0.8998124188350729,
+      "grad_norm": 3.52094871392947,
+      "learning_rate": 5.225080692315532e-07,
+      "loss": 3.0476,
+      "step": 4677
+    },
+    {
+      "epoch": 0.9000048097734596,
+      "grad_norm": 3.5082045482277797,
+      "learning_rate": 5.205217733614353e-07,
+      "loss": 2.9493,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9001972007118465,
+      "grad_norm": 2.4421854269281176,
+      "learning_rate": 5.185391592114286e-07,
+      "loss": 2.8438,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9003895916502332,
+      "grad_norm": 3.597407454243557,
+      "learning_rate": 5.165602275515592e-07,
+      "loss": 2.8803,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9005819825886201,
+      "grad_norm": 3.2834668457003615,
+      "learning_rate": 5.145849791504187e-07,
+      "loss": 2.7701,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9007743735270068,
+      "grad_norm": 3.795326973950792,
+      "learning_rate": 5.126134147751716e-07,
+      "loss": 3.0877,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9009667644653937,
+      "grad_norm": 3.372706362608499,
+      "learning_rate": 5.106455351915507e-07,
+      "loss": 2.7616,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9011591554037804,
+      "grad_norm": 3.4717069160081295,
+      "learning_rate": 5.086813411638581e-07,
+      "loss": 3.0659,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9013515463421673,
+      "grad_norm": 3.5768087901657557,
+      "learning_rate": 5.067208334549656e-07,
+      "loss": 2.9411,
+      "step": 4685
+    },
+    {
+      "epoch": 0.901543937280554,
+      "grad_norm": 3.3163575238779566,
+      "learning_rate": 5.047640128263087e-07,
+      "loss": 2.8974,
+      "step": 4686
+    },
+    {
+      "epoch": 0.9017363282189409,
+      "grad_norm": 3.5540183921834423,
+      "learning_rate": 5.028108800378961e-07,
+      "loss": 2.9393,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9019287191573276,
+      "grad_norm": 3.4267922675961913,
+      "learning_rate": 5.008614358483021e-07,
+      "loss": 3.1885,
+      "step": 4688
+    },
+    {
+      "epoch": 0.9021211100957145,
+      "grad_norm": 3.6056820584946236,
+      "learning_rate": 4.989156810146667e-07,
+      "loss": 2.8749,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9023135010341012,
+      "grad_norm": 3.2428100869287784,
+      "learning_rate": 4.969736162927019e-07,
+      "loss": 2.9168,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9025058919724881,
+      "grad_norm": 3.3390129700610585,
+      "learning_rate": 4.9503524243668e-07,
+      "loss": 3.0433,
+      "step": 4691
+    },
+    {
+      "epoch": 0.902698282910875,
+      "grad_norm": 3.6773647103991194,
+      "learning_rate": 4.931005601994432e-07,
+      "loss": 3.1798,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9028906738492617,
+      "grad_norm": 3.713577049524262,
+      "learning_rate": 4.911695703324038e-07,
+      "loss": 2.6648,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9030830647876485,
+      "grad_norm": 3.4568151440112005,
+      "learning_rate": 4.892422735855284e-07,
+      "loss": 2.9047,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9032754557260353,
+      "grad_norm": 3.4617190480968825,
+      "learning_rate": 4.873186707073663e-07,
+      "loss": 2.9735,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9034678466644221,
+      "grad_norm": 3.518406860860172,
+      "learning_rate": 4.853987624450151e-07,
+      "loss": 2.627,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9036602376028089,
+      "grad_norm": 3.5004320679950935,
+      "learning_rate": 4.834825495441475e-07,
+      "loss": 2.9618,
+      "step": 4697
+    },
+    {
+      "epoch": 0.9038526285411957,
+      "grad_norm": 3.275530257501013,
+      "learning_rate": 4.815700327490014e-07,
+      "loss": 2.8352,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9040450194795825,
+      "grad_norm": 3.730911880176004,
+      "learning_rate": 4.796612128023726e-07,
+      "loss": 2.8431,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9042374104179693,
+      "grad_norm": 3.631299405623268,
+      "learning_rate": 4.777560904456236e-07,
+      "loss": 2.8467,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9044298013563561,
+      "grad_norm": 3.6997644628223942,
+      "learning_rate": 4.7585466641868696e-07,
+      "loss": 2.7928,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9046221922947429,
+      "grad_norm": 2.2033489953178074,
+      "learning_rate": 4.7395694146004976e-07,
+      "loss": 2.3831,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9048145832331297,
+      "grad_norm": 3.509679484897604,
+      "learning_rate": 4.7206291630677024e-07,
+      "loss": 2.9604,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9050069741715165,
+      "grad_norm": 3.414728222860063,
+      "learning_rate": 4.7017259169446104e-07,
+      "loss": 2.8278,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9051993651099033,
+      "grad_norm": 3.34849495191963,
+      "learning_rate": 4.6828596835730487e-07,
+      "loss": 2.5606,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9053917560482901,
+      "grad_norm": 3.5727275402110354,
+      "learning_rate": 4.664030470280467e-07,
+      "loss": 2.8555,
+      "step": 4706
+    },
+    {
+      "epoch": 0.9055841469866769,
+      "grad_norm": 3.4718251141833756,
+      "learning_rate": 4.645238284379883e-07,
+      "loss": 2.9205,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9057765379250637,
+      "grad_norm": 3.2167874587997742,
+      "learning_rate": 4.626483133169968e-07,
+      "loss": 3.1403,
+      "step": 4708
+    },
+    {
+      "epoch": 0.9059689288634505,
+      "grad_norm": 3.5029078518353085,
+      "learning_rate": 4.60776502393504e-07,
+      "loss": 2.9121,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9061613198018373,
+      "grad_norm": 3.7445366581172475,
+      "learning_rate": 4.5890839639449514e-07,
+      "loss": 2.7689,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9063537107402241,
+      "grad_norm": 3.63129191247655,
+      "learning_rate": 4.5704399604552417e-07,
+      "loss": 2.9831,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9065461016786109,
+      "grad_norm": 3.520911384557416,
+      "learning_rate": 4.551833020707008e-07,
+      "loss": 2.847,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9067384926169978,
+      "grad_norm": 3.8287766843579756,
+      "learning_rate": 4.533263151926981e-07,
+      "loss": 2.9104,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9069308835553845,
+      "grad_norm": 3.3810145301224415,
+      "learning_rate": 4.514730361327502e-07,
+      "loss": 3.0228,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9071232744937714,
+      "grad_norm": 3.432244325738811,
+      "learning_rate": 4.4962346561064574e-07,
+      "loss": 2.865,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9073156654321581,
+      "grad_norm": 3.3813617649762966,
+      "learning_rate": 4.4777760434473796e-07,
+      "loss": 2.6647,
+      "step": 4716
+    },
+    {
+      "epoch": 0.907508056370545,
+      "grad_norm": 3.445886521793834,
+      "learning_rate": 4.4593545305193774e-07,
+      "loss": 3.2322,
+      "step": 4717
+    },
+    {
+      "epoch": 0.9077004473089317,
+      "grad_norm": 4.123573584859031,
+      "learning_rate": 4.440970124477173e-07,
+      "loss": 3.0576,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9078928382473186,
+      "grad_norm": 3.3428020987845226,
+      "learning_rate": 4.4226228324610544e-07,
+      "loss": 3.0418,
+      "step": 4719
+    },
+    {
+      "epoch": 0.9080852291857053,
+      "grad_norm": 3.4677537936901595,
+      "learning_rate": 4.404312661596877e-07,
+      "loss": 3.1164,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9082776201240922,
+      "grad_norm": 3.463707000832662,
+      "learning_rate": 4.386039618996119e-07,
+      "loss": 2.8314,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9084700110624789,
+      "grad_norm": 3.557480120904109,
+      "learning_rate": 4.3678037117558135e-07,
+      "loss": 3.045,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9086624020008658,
+      "grad_norm": 3.770559901026438,
+      "learning_rate": 4.349604946958563e-07,
+      "loss": 2.9221,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9088547929392525,
+      "grad_norm": 3.4985384601273735,
+      "learning_rate": 4.331443331672591e-07,
+      "loss": 2.7804,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9090471838776394,
+      "grad_norm": 3.2178924311589276,
+      "learning_rate": 4.313318872951633e-07,
+      "loss": 2.7771,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9092395748160261,
+      "grad_norm": 3.3569933909272023,
+      "learning_rate": 4.295231577835024e-07,
+      "loss": 3.0306,
+      "step": 4726
+    },
+    {
+      "epoch": 0.909431965754413,
+      "grad_norm": 3.5351402650392396,
+      "learning_rate": 4.2771814533476787e-07,
+      "loss": 2.9429,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9096243566927997,
+      "grad_norm": 3.335510519613083,
+      "learning_rate": 4.2591685065000223e-07,
+      "loss": 2.8912,
+      "step": 4728
+    },
+    {
+      "epoch": 0.9098167476311866,
+      "grad_norm": 3.5283639000044213,
+      "learning_rate": 4.241192744288092e-07,
+      "loss": 3.0184,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9100091385695733,
+      "grad_norm": 3.29072904970253,
+      "learning_rate": 4.223254173693492e-07,
+      "loss": 2.782,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9102015295079602,
+      "grad_norm": 3.5093331285706952,
+      "learning_rate": 4.2053528016833267e-07,
+      "loss": 2.8288,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9103939204463469,
+      "grad_norm": 3.8231653051510115,
+      "learning_rate": 4.1874886352103015e-07,
+      "loss": 2.9322,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9105863113847338,
+      "grad_norm": 3.4814278341959928,
+      "learning_rate": 4.1696616812126333e-07,
+      "loss": 2.7941,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9107787023231205,
+      "grad_norm": 3.423516893903137,
+      "learning_rate": 4.1518719466141165e-07,
+      "loss": 3.026,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9109710932615074,
+      "grad_norm": 3.591302040266137,
+      "learning_rate": 4.13411943832408e-07,
+      "loss": 3.0641,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9111634841998942,
+      "grad_norm": 3.5183355779823495,
+      "learning_rate": 4.116404163237386e-07,
+      "loss": 3.0067,
+      "step": 4736
+    },
+    {
+      "epoch": 0.911355875138281,
+      "grad_norm": 3.504191142843561,
+      "learning_rate": 4.0987261282344425e-07,
+      "loss": 2.9703,
+      "step": 4737
+    },
+    {
+      "epoch": 0.9115482660766678,
+      "grad_norm": 3.3377543349620433,
+      "learning_rate": 4.081085340181223e-07,
+      "loss": 2.8547,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9117406570150546,
+      "grad_norm": 3.4654635115625005,
+      "learning_rate": 4.06348180592917e-07,
+      "loss": 3.036,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9119330479534414,
+      "grad_norm": 3.3088248436262893,
+      "learning_rate": 4.0459155323153034e-07,
+      "loss": 3.0203,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9121254388918282,
+      "grad_norm": 3.411021153591767,
+      "learning_rate": 4.028386526162176e-07,
+      "loss": 2.8269,
+      "step": 4741
+    },
+    {
+      "epoch": 0.912317829830215,
+      "grad_norm": 3.54518214399701,
+      "learning_rate": 4.010894794277831e-07,
+      "loss": 3.0162,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9125102207686018,
+      "grad_norm": 3.6323432399873172,
+      "learning_rate": 3.993440343455879e-07,
+      "loss": 3.0844,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9127026117069886,
+      "grad_norm": 3.538574039183639,
+      "learning_rate": 3.976023180475397e-07,
+      "loss": 2.9923,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9128950026453754,
+      "grad_norm": 3.5284416845658235,
+      "learning_rate": 3.95864331210104e-07,
+      "loss": 2.8389,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9130873935837622,
+      "grad_norm": 3.5780765579540024,
+      "learning_rate": 3.941300745082932e-07,
+      "loss": 2.9878,
+      "step": 4746
+    },
+    {
+      "epoch": 0.913279784522149,
+      "grad_norm": 3.3494112768940143,
+      "learning_rate": 3.9239954861567177e-07,
+      "loss": 3.0044,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9134721754605358,
+      "grad_norm": 3.4972091879038345,
+      "learning_rate": 3.906727542043598e-07,
+      "loss": 2.8701,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9136645663989226,
+      "grad_norm": 3.557583789951801,
+      "learning_rate": 3.8894969194502083e-07,
+      "loss": 3.002,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9138569573373094,
+      "grad_norm": 3.597408227488293,
+      "learning_rate": 3.87230362506873e-07,
+      "loss": 2.9554,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9140493482756962,
+      "grad_norm": 3.543058130071156,
+      "learning_rate": 3.8551476655768527e-07,
+      "loss": 3.3495,
+      "step": 4751
+    },
+    {
+      "epoch": 0.914241739214083,
+      "grad_norm": 3.5216526872493934,
+      "learning_rate": 3.8380290476377255e-07,
+      "loss": 3.0711,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9144341301524698,
+      "grad_norm": 3.331401899175473,
+      "learning_rate": 3.8209477779000637e-07,
+      "loss": 2.7435,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9146265210908566,
+      "grad_norm": 3.6685781243062427,
+      "learning_rate": 3.803903862998004e-07,
+      "loss": 2.878,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9148189120292434,
+      "grad_norm": 3.3260953952520844,
+      "learning_rate": 3.7868973095512185e-07,
+      "loss": 2.8135,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9150113029676302,
+      "grad_norm": 3.336838844958165,
+      "learning_rate": 3.7699281241648565e-07,
+      "loss": 3.1148,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9152036939060171,
+      "grad_norm": 3.4408877234150577,
+      "learning_rate": 3.7529963134295466e-07,
+      "loss": 2.9533,
+      "step": 4757
+    },
+    {
+      "epoch": 0.9153960848444038,
+      "grad_norm": 3.3724201204949242,
+      "learning_rate": 3.736101883921406e-07,
+      "loss": 2.9929,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9155884757827907,
+      "grad_norm": 3.6561092115817693,
+      "learning_rate": 3.719244842202074e-07,
+      "loss": 2.978,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9157808667211774,
+      "grad_norm": 2.302212096018329,
+      "learning_rate": 3.702425194818582e-07,
+      "loss": 2.5007,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9159732576595643,
+      "grad_norm": 2.5832090741008282,
+      "learning_rate": 3.685642948303503e-07,
+      "loss": 2.5623,
+      "step": 4761
+    },
+    {
+      "epoch": 0.916165648597951,
+      "grad_norm": 3.4383366649440212,
+      "learning_rate": 3.66889810917489e-07,
+      "loss": 2.9009,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9163580395363379,
+      "grad_norm": 3.5771168182098565,
+      "learning_rate": 3.6521906839362187e-07,
+      "loss": 2.926,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9165504304747246,
+      "grad_norm": 3.5871972041788247,
+      "learning_rate": 3.635520679076465e-07,
+      "loss": 3.1304,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9167428214131115,
+      "grad_norm": 3.3887139734177474,
+      "learning_rate": 3.6188881010700725e-07,
+      "loss": 2.9172,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9169352123514982,
+      "grad_norm": 3.249795788170536,
+      "learning_rate": 3.6022929563769513e-07,
+      "loss": 3.0882,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9171276032898851,
+      "grad_norm": 3.543012077988988,
+      "learning_rate": 3.5857352514424573e-07,
+      "loss": 3.1034,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9173199942282718,
+      "grad_norm": 3.4737025226518177,
+      "learning_rate": 3.5692149926974006e-07,
+      "loss": 2.8482,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9175123851666587,
+      "grad_norm": 2.1861717764689392,
+      "learning_rate": 3.552732186558072e-07,
+      "loss": 2.6192,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9177047761050454,
+      "grad_norm": 3.4037806824051886,
+      "learning_rate": 3.536286839426195e-07,
+      "loss": 3.0174,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9178971670434323,
+      "grad_norm": 3.6310277789091923,
+      "learning_rate": 3.51987895768896e-07,
+      "loss": 3.0786,
+      "step": 4771
+    },
+    {
+      "epoch": 0.918089557981819,
+      "grad_norm": 3.684430906356241,
+      "learning_rate": 3.5035085477190143e-07,
+      "loss": 2.9765,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9182819489202059,
+      "grad_norm": 3.4467265599092185,
+      "learning_rate": 3.4871756158744054e-07,
+      "loss": 3.149,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9184743398585926,
+      "grad_norm": 3.2293801913635845,
+      "learning_rate": 3.4708801684986693e-07,
+      "loss": 2.8961,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9186667307969795,
+      "grad_norm": 3.3031795030549453,
+      "learning_rate": 3.454622211920766e-07,
+      "loss": 3.0615,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9188591217353662,
+      "grad_norm": 3.3565493262530355,
+      "learning_rate": 3.4384017524551116e-07,
+      "loss": 2.6972,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9190515126737531,
+      "grad_norm": 3.4761519673708343,
+      "learning_rate": 3.422218796401544e-07,
+      "loss": 3.0417,
+      "step": 4777
+    },
+    {
+      "epoch": 0.9192439036121399,
+      "grad_norm": 3.6963947209430352,
+      "learning_rate": 3.4060733500453247e-07,
+      "loss": 2.8364,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9194362945505267,
+      "grad_norm": 3.2212469554509955,
+      "learning_rate": 3.38996541965716e-07,
+      "loss": 2.7768,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9196286854889135,
+      "grad_norm": 2.390555125524239,
+      "learning_rate": 3.3738950114932e-07,
+      "loss": 2.6094,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9198210764273003,
+      "grad_norm": 3.478057065506971,
+      "learning_rate": 3.3578621317949755e-07,
+      "loss": 2.7856,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9200134673656871,
+      "grad_norm": 3.5340396992042797,
+      "learning_rate": 3.341866786789505e-07,
+      "loss": 2.9481,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9202058583040739,
+      "grad_norm": 3.3530800683774817,
+      "learning_rate": 3.325908982689185e-07,
+      "loss": 2.7423,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9203982492424607,
+      "grad_norm": 3.7140258586808796,
+      "learning_rate": 3.309988725691837e-07,
+      "loss": 3.0742,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9205906401808475,
+      "grad_norm": 3.3745751139269204,
+      "learning_rate": 3.294106021980714e-07,
+      "loss": 2.9713,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9207830311192343,
+      "grad_norm": 3.2372613334080604,
+      "learning_rate": 3.278260877724471e-07,
+      "loss": 2.7421,
+      "step": 4786
+    },
+    {
+      "epoch": 0.9209754220576211,
+      "grad_norm": 3.6373658128892616,
+      "learning_rate": 3.2624532990771507e-07,
+      "loss": 2.854,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9211678129960079,
+      "grad_norm": 3.545614672020459,
+      "learning_rate": 3.2466832921782986e-07,
+      "loss": 2.9965,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9213602039343947,
+      "grad_norm": 3.320439900811736,
+      "learning_rate": 3.2309508631527486e-07,
+      "loss": 2.9536,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9215525948727815,
+      "grad_norm": 3.4299986805938647,
+      "learning_rate": 3.215256018110824e-07,
+      "loss": 2.7225,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9217449858111683,
+      "grad_norm": 2.1778144737944256,
+      "learning_rate": 3.199598763148215e-07,
+      "loss": 2.5293,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9219373767495551,
+      "grad_norm": 3.2082707148510963,
+      "learning_rate": 3.183979104346002e-07,
+      "loss": 2.7017,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9221297676879419,
+      "grad_norm": 3.558492337200569,
+      "learning_rate": 3.1683970477706994e-07,
+      "loss": 2.8222,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9223221586263287,
+      "grad_norm": 3.6097346944785627,
+      "learning_rate": 3.1528525994741876e-07,
+      "loss": 3.0472,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9225145495647155,
+      "grad_norm": 3.5144797470553986,
+      "learning_rate": 3.13734576549376e-07,
+      "loss": 2.9129,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9227069405031023,
+      "grad_norm": 3.320975674355928,
+      "learning_rate": 3.121876551852099e-07,
+      "loss": 2.7507,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9228993314414891,
+      "grad_norm": 3.5521099083927408,
+      "learning_rate": 3.1064449645572536e-07,
+      "loss": 3.1376,
+      "step": 4797
+    },
+    {
+      "epoch": 0.9230917223798759,
+      "grad_norm": 3.4291976942362346,
+      "learning_rate": 3.091051009602675e-07,
+      "loss": 3.1056,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9232841133182627,
+      "grad_norm": 3.433111991922318,
+      "learning_rate": 3.0756946929672017e-07,
+      "loss": 2.7965,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9234765042566495,
+      "grad_norm": 3.3387201156646236,
+      "learning_rate": 3.060376020615052e-07,
+      "loss": 3.0304,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9236688951950364,
+      "grad_norm": 3.270382309006731,
+      "learning_rate": 3.0450949984958347e-07,
+      "loss": 2.6392,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9238612861334231,
+      "grad_norm": 3.5345969813613025,
+      "learning_rate": 3.0298516325444893e-07,
+      "loss": 3.0737,
+      "step": 4802
+    },
+    {
+      "epoch": 0.92405367707181,
+      "grad_norm": 3.53071667733963,
+      "learning_rate": 3.0146459286813924e-07,
+      "loss": 2.9262,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9242460680101967,
+      "grad_norm": 3.30380765264328,
+      "learning_rate": 2.999477892812264e-07,
+      "loss": 2.5171,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9244384589485836,
+      "grad_norm": 3.5030214636824337,
+      "learning_rate": 2.984347530828158e-07,
+      "loss": 2.8315,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9246308498869703,
+      "grad_norm": 3.5620907260448194,
+      "learning_rate": 2.969254848605585e-07,
+      "loss": 3.3349,
+      "step": 4806
+    },
+    {
+      "epoch": 0.9248232408253572,
+      "grad_norm": 3.2223717653421033,
+      "learning_rate": 2.9541998520063344e-07,
+      "loss": 2.7463,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9250156317637439,
+      "grad_norm": 3.203802873136298,
+      "learning_rate": 2.9391825468775946e-07,
+      "loss": 3.0877,
+      "step": 4808
+    },
+    {
+      "epoch": 0.9252080227021308,
+      "grad_norm": 3.7669694831317346,
+      "learning_rate": 2.9242029390519454e-07,
+      "loss": 3.0329,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9254004136405175,
+      "grad_norm": 3.576733198928374,
+      "learning_rate": 2.909261034347255e-07,
+      "loss": 2.8232,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9255928045789044,
+      "grad_norm": 2.3493531001718417,
+      "learning_rate": 2.894356838566792e-07,
+      "loss": 2.5123,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9257851955172911,
+      "grad_norm": 3.646337804835925,
+      "learning_rate": 2.879490357499204e-07,
+      "loss": 2.9142,
+      "step": 4812
+    },
+    {
+      "epoch": 0.925977586455678,
+      "grad_norm": 3.7448098904304445,
+      "learning_rate": 2.864661596918428e-07,
+      "loss": 3.014,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9261699773940647,
+      "grad_norm": 3.362402976888832,
+      "learning_rate": 2.849870562583812e-07,
+      "loss": 2.8174,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9263623683324516,
+      "grad_norm": 3.5121448901725607,
+      "learning_rate": 2.8351172602399945e-07,
+      "loss": 3.0797,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9265547592708383,
+      "grad_norm": 3.4313519896379123,
+      "learning_rate": 2.8204016956169924e-07,
+      "loss": 3.016,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9267471502092252,
+      "grad_norm": 3.8296312895406404,
+      "learning_rate": 2.8057238744301994e-07,
+      "loss": 3.0332,
+      "step": 4817
+    },
+    {
+      "epoch": 0.9269395411476119,
+      "grad_norm": 3.4614290492294173,
+      "learning_rate": 2.7910838023802676e-07,
+      "loss": 2.9548,
+      "step": 4818
+    },
+    {
+      "epoch": 0.9271319320859988,
+      "grad_norm": 3.6520033629797752,
+      "learning_rate": 2.7764814851532485e-07,
+      "loss": 3.0253,
+      "step": 4819
+    },
+    {
+      "epoch": 0.9273243230243855,
+      "grad_norm": 3.7575686260523176,
+      "learning_rate": 2.761916928420527e-07,
+      "loss": 3.0951,
+      "step": 4820
+    },
+    {
+      "epoch": 0.9275167139627724,
+      "grad_norm": 3.503863530496036,
+      "learning_rate": 2.74739013783879e-07,
+      "loss": 2.7708,
+      "step": 4821
+    },
+    {
+      "epoch": 0.9277091049011592,
+      "grad_norm": 2.3917887982838772,
+      "learning_rate": 2.7329011190500797e-07,
+      "loss": 2.7409,
+      "step": 4822
+    },
+    {
+      "epoch": 0.927901495839546,
+      "grad_norm": 3.424487724381096,
+      "learning_rate": 2.7184498776817615e-07,
+      "loss": 2.8986,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9280938867779328,
+      "grad_norm": 3.5278677236468416,
+      "learning_rate": 2.704036419346534e-07,
+      "loss": 2.6408,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9282862777163196,
+      "grad_norm": 3.4577834897100477,
+      "learning_rate": 2.689660749642442e-07,
+      "loss": 2.9666,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9284786686547064,
+      "grad_norm": 3.3729441144338366,
+      "learning_rate": 2.675322874152786e-07,
+      "loss": 2.9341,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9286710595930932,
+      "grad_norm": 3.5007074010756165,
+      "learning_rate": 2.6610227984462556e-07,
+      "loss": 3.0358,
+      "step": 4827
+    },
+    {
+      "epoch": 0.92886345053148,
+      "grad_norm": 3.329634619538399,
+      "learning_rate": 2.646760528076842e-07,
+      "loss": 2.8452,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9290558414698667,
+      "grad_norm": 3.49473110798476,
+      "learning_rate": 2.6325360685838243e-07,
+      "loss": 3.0805,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9292482324082536,
+      "grad_norm": 2.176800713082535,
+      "learning_rate": 2.618349425491851e-07,
+      "loss": 2.42,
+      "step": 4830
+    },
+    {
+      "epoch": 0.9294406233466403,
+      "grad_norm": 2.3958090215307304,
+      "learning_rate": 2.604200604310825e-07,
+      "loss": 2.552,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9296330142850272,
+      "grad_norm": 3.8501904546023633,
+      "learning_rate": 2.590089610535984e-07,
+      "loss": 3.0696,
+      "step": 4832
+    },
+    {
+      "epoch": 0.929825405223414,
+      "grad_norm": 3.6463947876137346,
+      "learning_rate": 2.5760164496479e-07,
+      "loss": 2.8306,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9300177961618008,
+      "grad_norm": 3.196537679702906,
+      "learning_rate": 2.5619811271123897e-07,
+      "loss": 2.8461,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9302101871001875,
+      "grad_norm": 3.3788929542641273,
+      "learning_rate": 2.5479836483806586e-07,
+      "loss": 2.9748,
+      "step": 4835
+    },
+    {
+      "epoch": 0.9304025780385744,
+      "grad_norm": 2.2356868718589302,
+      "learning_rate": 2.5340240188891143e-07,
+      "loss": 2.6042,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9305949689769611,
+      "grad_norm": 3.3981814798618477,
+      "learning_rate": 2.520102244059552e-07,
+      "loss": 2.9663,
+      "step": 4837
+    },
+    {
+      "epoch": 0.930787359915348,
+      "grad_norm": 3.95290725247484,
+      "learning_rate": 2.506218329299026e-07,
+      "loss": 2.8867,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9309797508537347,
+      "grad_norm": 3.541407912482514,
+      "learning_rate": 2.4923722799998664e-07,
+      "loss": 3.091,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9311721417921216,
+      "grad_norm": 3.1793101280613776,
+      "learning_rate": 2.4785641015397375e-07,
+      "loss": 2.7504,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9313645327305083,
+      "grad_norm": 3.5643033655278287,
+      "learning_rate": 2.464793799281573e-07,
+      "loss": 2.9898,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9315569236688952,
+      "grad_norm": 3.878223036411535,
+      "learning_rate": 2.4510613785735936e-07,
+      "loss": 2.8888,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9317493146072819,
+      "grad_norm": 3.730797920247435,
+      "learning_rate": 2.4373668447493225e-07,
+      "loss": 2.791,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9319417055456688,
+      "grad_norm": 3.4544775554022813,
+      "learning_rate": 2.423710203127561e-07,
+      "loss": 2.7455,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9321340964840557,
+      "grad_norm": 3.3877048612515823,
+      "learning_rate": 2.410091459012376e-07,
+      "loss": 2.9376,
+      "step": 4845
+    },
+    {
+      "epoch": 0.9323264874224424,
+      "grad_norm": 3.2771371546858927,
+      "learning_rate": 2.3965106176931375e-07,
+      "loss": 2.9045,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9325188783608293,
+      "grad_norm": 3.739810621471825,
+      "learning_rate": 2.3829676844444926e-07,
+      "loss": 2.9683,
+      "step": 4847
+    },
+    {
+      "epoch": 0.932711269299216,
+      "grad_norm": 3.4877697244674315,
+      "learning_rate": 2.3694626645263675e-07,
+      "loss": 3.0045,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9329036602376028,
+      "grad_norm": 3.4100346796884966,
+      "learning_rate": 2.3559955631839436e-07,
+      "loss": 2.8646,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9330960511759896,
+      "grad_norm": 3.545947275409003,
+      "learning_rate": 2.3425663856476932e-07,
+      "loss": 2.9682,
+      "step": 4850
+    },
+    {
+      "epoch": 0.9332884421143764,
+      "grad_norm": 3.5382512175563448,
+      "learning_rate": 2.3291751371333438e-07,
+      "loss": 3.1222,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9334808330527632,
+      "grad_norm": 3.4956113446202184,
+      "learning_rate": 2.3158218228419127e-07,
+      "loss": 2.881,
+      "step": 4852
+    },
+    {
+      "epoch": 0.93367322399115,
+      "grad_norm": 3.3279313569314763,
+      "learning_rate": 2.3025064479596625e-07,
+      "loss": 2.9831,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9338656149295368,
+      "grad_norm": 2.787180649728798,
+      "learning_rate": 2.2892290176581678e-07,
+      "loss": 2.6755,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9340580058679236,
+      "grad_norm": 3.153667751005726,
+      "learning_rate": 2.2759895370941809e-07,
+      "loss": 2.9567,
+      "step": 4855
+    },
+    {
+      "epoch": 0.9342503968063104,
+      "grad_norm": 3.516066203955112,
+      "learning_rate": 2.2627880114097779e-07,
+      "loss": 2.7327,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9344427877446972,
+      "grad_norm": 3.302155074725721,
+      "learning_rate": 2.2496244457323013e-07,
+      "loss": 3.0899,
+      "step": 4857
+    },
+    {
+      "epoch": 0.934635178683084,
+      "grad_norm": 3.4203120741282453,
+      "learning_rate": 2.2364988451742953e-07,
+      "loss": 2.905,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9348275696214708,
+      "grad_norm": 2.2835692566348915,
+      "learning_rate": 2.2234112148336373e-07,
+      "loss": 2.58,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9350199605598576,
+      "grad_norm": 3.30247072837686,
+      "learning_rate": 2.2103615597933613e-07,
+      "loss": 2.9321,
+      "step": 4860
+    },
+    {
+      "epoch": 0.9352123514982444,
+      "grad_norm": 3.8021005471245,
+      "learning_rate": 2.1973498851218244e-07,
+      "loss": 2.6802,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9354047424366312,
+      "grad_norm": 3.5888327383283354,
+      "learning_rate": 2.1843761958726283e-07,
+      "loss": 3.1137,
+      "step": 4862
+    },
+    {
+      "epoch": 0.935597133375018,
+      "grad_norm": 3.364653008919371,
+      "learning_rate": 2.1714404970845647e-07,
+      "loss": 3.0211,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9357895243134048,
+      "grad_norm": 3.4793942434529055,
+      "learning_rate": 2.1585427937817594e-07,
+      "loss": 2.8275,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9359819152517916,
+      "grad_norm": 3.934884136685751,
+      "learning_rate": 2.145683090973494e-07,
+      "loss": 2.8106,
+      "step": 4865
+    },
+    {
+      "epoch": 0.9361743061901785,
+      "grad_norm": 3.7310271954409657,
+      "learning_rate": 2.1328613936543396e-07,
+      "loss": 3.1029,
+      "step": 4866
+    },
+    {
+      "epoch": 0.9363666971285652,
+      "grad_norm": 3.390827450990479,
+      "learning_rate": 2.1200777068041134e-07,
+      "loss": 3.0046,
+      "step": 4867
+    },
+    {
+      "epoch": 0.9365590880669521,
+      "grad_norm": 2.408207185207818,
+      "learning_rate": 2.1073320353878102e-07,
+      "loss": 2.5504,
+      "step": 4868
+    },
+    {
+      "epoch": 0.9367514790053388,
+      "grad_norm": 3.542832380774945,
+      "learning_rate": 2.0946243843557367e-07,
+      "loss": 2.8627,
+      "step": 4869
+    },
+    {
+      "epoch": 0.9369438699437257,
+      "grad_norm": 3.562084240122408,
+      "learning_rate": 2.0819547586434008e-07,
+      "loss": 2.9576,
+      "step": 4870
+    },
+    {
+      "epoch": 0.9371362608821124,
+      "grad_norm": 3.246272807282012,
+      "learning_rate": 2.0693231631715105e-07,
+      "loss": 2.7666,
+      "step": 4871
+    },
+    {
+      "epoch": 0.9373286518204993,
+      "grad_norm": 3.3207845213943545,
+      "learning_rate": 2.0567296028460638e-07,
+      "loss": 2.8263,
+      "step": 4872
+    },
+    {
+      "epoch": 0.937521042758886,
+      "grad_norm": 3.434087656930109,
+      "learning_rate": 2.0441740825582258e-07,
+      "loss": 2.9627,
+      "step": 4873
+    },
+    {
+      "epoch": 0.9377134336972729,
+      "grad_norm": 3.551417486600615,
+      "learning_rate": 2.0316566071844402e-07,
+      "loss": 2.9265,
+      "step": 4874
+    },
+    {
+      "epoch": 0.9379058246356596,
+      "grad_norm": 3.8198008369127834,
+      "learning_rate": 2.0191771815863292e-07,
+      "loss": 2.9683,
+      "step": 4875
+    },
+    {
+      "epoch": 0.9380982155740465,
+      "grad_norm": 3.416417677072006,
+      "learning_rate": 2.0067358106107714e-07,
+      "loss": 2.8529,
+      "step": 4876
+    },
+    {
+      "epoch": 0.9382906065124332,
+      "grad_norm": 3.567308758957792,
+      "learning_rate": 1.994332499089846e-07,
+      "loss": 3.0851,
+      "step": 4877
+    },
+    {
+      "epoch": 0.9384829974508201,
+      "grad_norm": 3.6294885377784194,
+      "learning_rate": 1.9819672518408662e-07,
+      "loss": 2.8823,
+      "step": 4878
+    },
+    {
+      "epoch": 0.9386753883892068,
+      "grad_norm": 3.8417298494413044,
+      "learning_rate": 1.9696400736663457e-07,
+      "loss": 2.9016,
+      "step": 4879
+    },
+    {
+      "epoch": 0.9388677793275937,
+      "grad_norm": 3.7374589667363662,
+      "learning_rate": 1.9573509693540104e-07,
+      "loss": 2.8655,
+      "step": 4880
+    },
+    {
+      "epoch": 0.9390601702659804,
+      "grad_norm": 3.4729576675138123,
+      "learning_rate": 1.9450999436768093e-07,
+      "loss": 2.9345,
+      "step": 4881
+    },
+    {
+      "epoch": 0.9392525612043673,
+      "grad_norm": 3.3444618548698553,
+      "learning_rate": 1.9328870013929134e-07,
+      "loss": 2.776,
+      "step": 4882
+    },
+    {
+      "epoch": 0.939444952142754,
+      "grad_norm": 3.6037681153520555,
+      "learning_rate": 1.9207121472456846e-07,
+      "loss": 2.8539,
+      "step": 4883
+    },
+    {
+      "epoch": 0.9396373430811409,
+      "grad_norm": 2.5095244641670935,
+      "learning_rate": 1.9085753859636736e-07,
+      "loss": 2.6048,
+      "step": 4884
+    },
+    {
+      "epoch": 0.9398297340195276,
+      "grad_norm": 3.4459011367581884,
+      "learning_rate": 1.8964767222606873e-07,
+      "loss": 2.737,
+      "step": 4885
+    },
+    {
+      "epoch": 0.9400221249579145,
+      "grad_norm": 3.4771212106006004,
+      "learning_rate": 1.8844161608356782e-07,
+      "loss": 3.0184,
+      "step": 4886
+    },
+    {
+      "epoch": 0.9402145158963013,
+      "grad_norm": 2.2918424841955436,
+      "learning_rate": 1.872393706372866e-07,
+      "loss": 2.5843,
+      "step": 4887
+    },
+    {
+      "epoch": 0.9404069068346881,
+      "grad_norm": 3.2442298002298613,
+      "learning_rate": 1.8604093635416155e-07,
+      "loss": 3.1574,
+      "step": 4888
+    },
+    {
+      "epoch": 0.9405992977730749,
+      "grad_norm": 2.264158967293115,
+      "learning_rate": 1.8484631369964922e-07,
+      "loss": 2.4639,
+      "step": 4889
+    },
+    {
+      "epoch": 0.9407916887114617,
+      "grad_norm": 3.735937508826978,
+      "learning_rate": 1.8365550313772852e-07,
+      "loss": 3.0224,
+      "step": 4890
+    },
+    {
+      "epoch": 0.9409840796498485,
+      "grad_norm": 3.2004687653039317,
+      "learning_rate": 1.8246850513089832e-07,
+      "loss": 2.7497,
+      "step": 4891
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 2.287341712784248,
+      "learning_rate": 1.8128532014017098e-07,
+      "loss": 2.569,
+      "step": 4892
+    },
+    {
+      "epoch": 0.9413688615266221,
+      "grad_norm": 3.5734581932230185,
+      "learning_rate": 1.8010594862508669e-07,
+      "loss": 2.7612,
+      "step": 4893
+    },
+    {
+      "epoch": 0.9415612524650089,
+      "grad_norm": 3.440124191615203,
+      "learning_rate": 1.789303910436968e-07,
+      "loss": 2.8682,
+      "step": 4894
+    },
+    {
+      "epoch": 0.9417536434033957,
+      "grad_norm": 3.2203357472654455,
+      "learning_rate": 1.777586478525739e-07,
+      "loss": 2.8931,
+      "step": 4895
+    },
+    {
+      "epoch": 0.9419460343417825,
+      "grad_norm": 3.6645513276253032,
+      "learning_rate": 1.7659071950681172e-07,
+      "loss": 3.1465,
+      "step": 4896
+    },
+    {
+      "epoch": 0.9421384252801693,
+      "grad_norm": 3.5181861150829388,
+      "learning_rate": 1.754266064600174e-07,
+      "loss": 2.8889,
+      "step": 4897
+    },
+    {
+      "epoch": 0.9423308162185561,
+      "grad_norm": 3.761112551150385,
+      "learning_rate": 1.742663091643204e-07,
+      "loss": 3.1139,
+      "step": 4898
+    },
+    {
+      "epoch": 0.9425232071569429,
+      "grad_norm": 3.3221997434027464,
+      "learning_rate": 1.7310982807036915e-07,
+      "loss": 3.0416,
+      "step": 4899
+    },
+    {
+      "epoch": 0.9427155980953297,
+      "grad_norm": 3.588546230242903,
+      "learning_rate": 1.719571636273243e-07,
+      "loss": 2.86,
+      "step": 4900
+    },
+    {
+      "epoch": 0.9429079890337165,
+      "grad_norm": 3.518761492747777,
+      "learning_rate": 1.7080831628286886e-07,
+      "loss": 2.6574,
+      "step": 4901
+    },
+    {
+      "epoch": 0.9431003799721033,
+      "grad_norm": 3.6615549802422125,
+      "learning_rate": 1.6966328648320152e-07,
+      "loss": 2.9445,
+      "step": 4902
+    },
+    {
+      "epoch": 0.9432927709104901,
+      "grad_norm": 3.60248768263274,
+      "learning_rate": 1.685220746730387e-07,
+      "loss": 2.8453,
+      "step": 4903
+    },
+    {
+      "epoch": 0.9434851618488769,
+      "grad_norm": 3.643688282836156,
+      "learning_rate": 1.673846812956137e-07,
+      "loss": 2.8599,
+      "step": 4904
+    },
+    {
+      "epoch": 0.9436775527872637,
+      "grad_norm": 3.5298334045085458,
+      "learning_rate": 1.6625110679267642e-07,
+      "loss": 2.9753,
+      "step": 4905
+    },
+    {
+      "epoch": 0.9438699437256505,
+      "grad_norm": 3.293256905872579,
+      "learning_rate": 1.6512135160449583e-07,
+      "loss": 3.0828,
+      "step": 4906
+    },
+    {
+      "epoch": 0.9440623346640373,
+      "grad_norm": 3.528921323802585,
+      "learning_rate": 1.6399541616985648e-07,
+      "loss": 2.9227,
+      "step": 4907
+    },
+    {
+      "epoch": 0.9442547256024241,
+      "grad_norm": 3.214710382049467,
+      "learning_rate": 1.6287330092605525e-07,
+      "loss": 2.7445,
+      "step": 4908
+    },
+    {
+      "epoch": 0.9444471165408109,
+      "grad_norm": 3.587031320069889,
+      "learning_rate": 1.6175500630891128e-07,
+      "loss": 2.9272,
+      "step": 4909
+    },
+    {
+      "epoch": 0.9446395074791978,
+      "grad_norm": 3.4702287677914914,
+      "learning_rate": 1.6064053275275716e-07,
+      "loss": 3.028,
+      "step": 4910
+    },
+    {
+      "epoch": 0.9448318984175845,
+      "grad_norm": 3.5068910174454255,
+      "learning_rate": 1.5952988069044105e-07,
+      "loss": 2.8788,
+      "step": 4911
+    },
+    {
+      "epoch": 0.9450242893559714,
+      "grad_norm": 3.5565841528372566,
+      "learning_rate": 1.5842305055332796e-07,
+      "loss": 2.813,
+      "step": 4912
+    },
+    {
+      "epoch": 0.9452166802943581,
+      "grad_norm": 2.15388798971836,
+      "learning_rate": 1.573200427712973e-07,
+      "loss": 2.5702,
+      "step": 4913
+    },
+    {
+      "epoch": 0.945409071232745,
+      "grad_norm": 3.4076736956199163,
+      "learning_rate": 1.562208577727442e-07,
+      "loss": 3.2593,
+      "step": 4914
+    },
+    {
+      "epoch": 0.9456014621711317,
+      "grad_norm": 2.3364228597761074,
+      "learning_rate": 1.551254959845805e-07,
+      "loss": 2.7164,
+      "step": 4915
+    },
+    {
+      "epoch": 0.9457938531095186,
+      "grad_norm": 3.585548919088363,
+      "learning_rate": 1.540339578322314e-07,
+      "loss": 2.984,
+      "step": 4916
+    },
+    {
+      "epoch": 0.9459862440479053,
+      "grad_norm": 3.5181388519113503,
+      "learning_rate": 1.5294624373963894e-07,
+      "loss": 3.0138,
+      "step": 4917
+    },
+    {
+      "epoch": 0.9461786349862922,
+      "grad_norm": 3.3818303451437832,
+      "learning_rate": 1.5186235412925744e-07,
+      "loss": 2.9034,
+      "step": 4918
+    },
+    {
+      "epoch": 0.9463710259246789,
+      "grad_norm": 3.4952752292118574,
+      "learning_rate": 1.5078228942205674e-07,
+      "loss": 2.7695,
+      "step": 4919
+    },
+    {
+      "epoch": 0.9465634168630658,
+      "grad_norm": 4.028556884327146,
+      "learning_rate": 1.4970605003752359e-07,
+      "loss": 2.9443,
+      "step": 4920
+    },
+    {
+      "epoch": 0.9467558078014525,
+      "grad_norm": 3.904565563768553,
+      "learning_rate": 1.4863363639365357e-07,
+      "loss": 3.1104,
+      "step": 4921
+    },
+    {
+      "epoch": 0.9469481987398394,
+      "grad_norm": 3.60366447878719,
+      "learning_rate": 1.4756504890696466e-07,
+      "loss": 3.2523,
+      "step": 4922
+    },
+    {
+      "epoch": 0.9471405896782261,
+      "grad_norm": 3.508426793980492,
+      "learning_rate": 1.4650028799247928e-07,
+      "loss": 2.8159,
+      "step": 4923
+    },
+    {
+      "epoch": 0.947332980616613,
+      "grad_norm": 3.236886143188479,
+      "learning_rate": 1.454393540637411e-07,
+      "loss": 2.9366,
+      "step": 4924
+    },
+    {
+      "epoch": 0.9475253715549997,
+      "grad_norm": 3.464744547613178,
+      "learning_rate": 1.4438224753280384e-07,
+      "loss": 2.8814,
+      "step": 4925
+    },
+    {
+      "epoch": 0.9477177624933866,
+      "grad_norm": 3.543200615819879,
+      "learning_rate": 1.4332896881023462e-07,
+      "loss": 3.0557,
+      "step": 4926
+    },
+    {
+      "epoch": 0.9479101534317733,
+      "grad_norm": 3.6945194156255816,
+      "learning_rate": 1.422795183051151e-07,
+      "loss": 2.8408,
+      "step": 4927
+    },
+    {
+      "epoch": 0.9481025443701602,
+      "grad_norm": 3.548412134940444,
+      "learning_rate": 1.4123389642504148e-07,
+      "loss": 2.7828,
+      "step": 4928
+    },
+    {
+      "epoch": 0.9482949353085469,
+      "grad_norm": 3.8328316316628155,
+      "learning_rate": 1.401921035761189e-07,
+      "loss": 3.0292,
+      "step": 4929
+    },
+    {
+      "epoch": 0.9484873262469338,
+      "grad_norm": 3.614691041708688,
+      "learning_rate": 1.3915414016296925e-07,
+      "loss": 2.8161,
+      "step": 4930
+    },
+    {
+      "epoch": 0.9486797171853206,
+      "grad_norm": 3.35295142896762,
+      "learning_rate": 1.381200065887256e-07,
+      "loss": 2.8447,
+      "step": 4931
+    },
+    {
+      "epoch": 0.9488721081237074,
+      "grad_norm": 3.703605677682909,
+      "learning_rate": 1.3708970325503222e-07,
+      "loss": 3.0498,
+      "step": 4932
+    },
+    {
+      "epoch": 0.9490644990620942,
+      "grad_norm": 3.731109247120978,
+      "learning_rate": 1.3606323056204795e-07,
+      "loss": 3.1982,
+      "step": 4933
+    },
+    {
+      "epoch": 0.949256890000481,
+      "grad_norm": 3.207079531543004,
+      "learning_rate": 1.3504058890844274e-07,
+      "loss": 2.8865,
+      "step": 4934
+    },
+    {
+      "epoch": 0.9494492809388678,
+      "grad_norm": 3.4931960203585697,
+      "learning_rate": 1.3402177869139887e-07,
+      "loss": 3.0085,
+      "step": 4935
+    },
+    {
+      "epoch": 0.9496416718772546,
+      "grad_norm": 3.34142528183289,
+      "learning_rate": 1.3300680030661096e-07,
+      "loss": 2.9807,
+      "step": 4936
+    },
+    {
+      "epoch": 0.9498340628156414,
+      "grad_norm": 3.513160024739883,
+      "learning_rate": 1.3199565414828363e-07,
+      "loss": 2.9746,
+      "step": 4937
+    },
+    {
+      "epoch": 0.9500264537540282,
+      "grad_norm": 3.470810129197769,
+      "learning_rate": 1.3098834060913612e-07,
+      "loss": 3.1528,
+      "step": 4938
+    },
+    {
+      "epoch": 0.950218844692415,
+      "grad_norm": 3.411749669604239,
+      "learning_rate": 1.2998486008039545e-07,
+      "loss": 3.0004,
+      "step": 4939
+    },
+    {
+      "epoch": 0.9504112356308018,
+      "grad_norm": 3.577354077243171,
+      "learning_rate": 1.289852129518032e-07,
+      "loss": 2.8371,
+      "step": 4940
+    },
+    {
+      "epoch": 0.9506036265691886,
+      "grad_norm": 3.339758499926422,
+      "learning_rate": 1.2798939961161217e-07,
+      "loss": 3.043,
+      "step": 4941
+    },
+    {
+      "epoch": 0.9507960175075754,
+      "grad_norm": 3.512688956006123,
+      "learning_rate": 1.269974204465818e-07,
+      "loss": 2.8921,
+      "step": 4942
+    },
+    {
+      "epoch": 0.9509884084459622,
+      "grad_norm": 3.4189454011728504,
+      "learning_rate": 1.2600927584198618e-07,
+      "loss": 2.8672,
+      "step": 4943
+    },
+    {
+      "epoch": 0.951180799384349,
+      "grad_norm": 3.523346941688603,
+      "learning_rate": 1.2502496618161165e-07,
+      "loss": 3.0119,
+      "step": 4944
+    },
+    {
+      "epoch": 0.9513731903227358,
+      "grad_norm": 3.343726318179,
+      "learning_rate": 1.2404449184774904e-07,
+      "loss": 2.9413,
+      "step": 4945
+    },
+    {
+      "epoch": 0.9515655812611226,
+      "grad_norm": 3.6707574891666868,
+      "learning_rate": 1.2306785322120596e-07,
+      "loss": 2.9235,
+      "step": 4946
+    },
+    {
+      "epoch": 0.9517579721995094,
+      "grad_norm": 3.377382018331764,
+      "learning_rate": 1.220950506812968e-07,
+      "loss": 2.817,
+      "step": 4947
+    },
+    {
+      "epoch": 0.9519503631378962,
+      "grad_norm": 3.6097725979223663,
+      "learning_rate": 1.2112608460584707e-07,
+      "loss": 2.7026,
+      "step": 4948
+    },
+    {
+      "epoch": 0.952142754076283,
+      "grad_norm": 3.233877195188728,
+      "learning_rate": 1.2016095537119242e-07,
+      "loss": 2.8691,
+      "step": 4949
+    },
+    {
+      "epoch": 0.9523351450146698,
+      "grad_norm": 3.334519829128023,
+      "learning_rate": 1.1919966335217636e-07,
+      "loss": 3.0878,
+      "step": 4950
+    },
+    {
+      "epoch": 0.9525275359530566,
+      "grad_norm": 3.414162977277624,
+      "learning_rate": 1.1824220892215465e-07,
+      "loss": 2.8477,
+      "step": 4951
+    },
+    {
+      "epoch": 0.9527199268914434,
+      "grad_norm": 3.736144120442636,
+      "learning_rate": 1.1728859245299207e-07,
+      "loss": 2.9669,
+      "step": 4952
+    },
+    {
+      "epoch": 0.9529123178298302,
+      "grad_norm": 3.6609097670443607,
+      "learning_rate": 1.1633881431506122e-07,
+      "loss": 2.9061,
+      "step": 4953
+    },
+    {
+      "epoch": 0.9531047087682171,
+      "grad_norm": 3.6084898133349945,
+      "learning_rate": 1.1539287487724594e-07,
+      "loss": 3.1314,
+      "step": 4954
+    },
+    {
+      "epoch": 0.9532970997066038,
+      "grad_norm": 3.408562301688697,
+      "learning_rate": 1.1445077450693786e-07,
+      "loss": 2.9594,
+      "step": 4955
+    },
+    {
+      "epoch": 0.9534894906449907,
+      "grad_norm": 3.4297051417944306,
+      "learning_rate": 1.1351251357003656e-07,
+      "loss": 2.9917,
+      "step": 4956
+    },
+    {
+      "epoch": 0.9536818815833774,
+      "grad_norm": 3.3630717131814034,
+      "learning_rate": 1.1257809243095385e-07,
+      "loss": 2.9181,
+      "step": 4957
+    },
+    {
+      "epoch": 0.9538742725217643,
+      "grad_norm": 3.320452848974101,
+      "learning_rate": 1.1164751145260722e-07,
+      "loss": 3.0001,
+      "step": 4958
+    },
+    {
+      "epoch": 0.954066663460151,
+      "grad_norm": 3.194403321520338,
+      "learning_rate": 1.107207709964242e-07,
+      "loss": 2.8951,
+      "step": 4959
+    },
+    {
+      "epoch": 0.9542590543985379,
+      "grad_norm": 3.463226869044062,
+      "learning_rate": 1.0979787142233911e-07,
+      "loss": 3.0262,
+      "step": 4960
+    },
+    {
+      "epoch": 0.9544514453369246,
+      "grad_norm": 3.490542843706729,
+      "learning_rate": 1.0887881308879633e-07,
+      "loss": 2.9154,
+      "step": 4961
+    },
+    {
+      "epoch": 0.9546438362753115,
+      "grad_norm": 3.564114306344288,
+      "learning_rate": 1.0796359635274701e-07,
+      "loss": 2.9365,
+      "step": 4962
+    },
+    {
+      "epoch": 0.9548362272136982,
+      "grad_norm": 3.400375912372053,
+      "learning_rate": 1.0705222156965011e-07,
+      "loss": 2.7084,
+      "step": 4963
+    },
+    {
+      "epoch": 0.9550286181520851,
+      "grad_norm": 3.609599567779918,
+      "learning_rate": 1.0614468909347476e-07,
+      "loss": 2.9299,
+      "step": 4964
+    },
+    {
+      "epoch": 0.9552210090904718,
+      "grad_norm": 3.3496159197957,
+      "learning_rate": 1.0524099927669563e-07,
+      "loss": 2.9531,
+      "step": 4965
+    },
+    {
+      "epoch": 0.9554134000288587,
+      "grad_norm": 3.3261400409865116,
+      "learning_rate": 1.0434115247029419e-07,
+      "loss": 2.9893,
+      "step": 4966
+    },
+    {
+      "epoch": 0.9556057909672454,
+      "grad_norm": 3.614633988950162,
+      "learning_rate": 1.0344514902376201e-07,
+      "loss": 2.962,
+      "step": 4967
+    },
+    {
+      "epoch": 0.9557981819056323,
+      "grad_norm": 3.77499946671362,
+      "learning_rate": 1.0255298928509627e-07,
+      "loss": 3.0122,
+      "step": 4968
+    },
+    {
+      "epoch": 0.955990572844019,
+      "grad_norm": 3.2223320643269413,
+      "learning_rate": 1.0166467360079979e-07,
+      "loss": 3.0173,
+      "step": 4969
+    },
+    {
+      "epoch": 0.9561829637824059,
+      "grad_norm": 3.550302781723397,
+      "learning_rate": 1.007802023158877e-07,
+      "loss": 2.8569,
+      "step": 4970
+    },
+    {
+      "epoch": 0.9563753547207926,
+      "grad_norm": 3.4675613173971045,
+      "learning_rate": 9.989957577387521e-08,
+      "loss": 2.7832,
+      "step": 4971
+    },
+    {
+      "epoch": 0.9565677456591795,
+      "grad_norm": 2.457682395084421,
+      "learning_rate": 9.902279431678874e-08,
+      "loss": 2.6055,
+      "step": 4972
+    },
+    {
+      "epoch": 0.9567601365975662,
+      "grad_norm": 3.133947936167308,
+      "learning_rate": 9.814985828516033e-08,
+      "loss": 2.8164,
+      "step": 4973
+    },
+    {
+      "epoch": 0.9569525275359531,
+      "grad_norm": 3.715175712837281,
+      "learning_rate": 9.728076801802656e-08,
+      "loss": 2.8477,
+      "step": 4974
+    },
+    {
+      "epoch": 0.9571449184743399,
+      "grad_norm": 3.373786964113746,
+      "learning_rate": 9.641552385293518e-08,
+      "loss": 2.7394,
+      "step": 4975
+    },
+    {
+      "epoch": 0.9573373094127267,
+      "grad_norm": 3.712811437596953,
+      "learning_rate": 9.555412612593518e-08,
+      "loss": 2.6772,
+      "step": 4976
+    },
+    {
+      "epoch": 0.9575297003511135,
+      "grad_norm": 3.6911269372123905,
+      "learning_rate": 9.469657517158226e-08,
+      "loss": 2.948,
+      "step": 4977
+    },
+    {
+      "epoch": 0.9577220912895003,
+      "grad_norm": 2.3598291334520662,
+      "learning_rate": 9.384287132294223e-08,
+      "loss": 2.6202,
+      "step": 4978
+    },
+    {
+      "epoch": 0.9579144822278871,
+      "grad_norm": 3.2801200308383374,
+      "learning_rate": 9.299301491158207e-08,
+      "loss": 3.1582,
+      "step": 4979
+    },
+    {
+      "epoch": 0.9581068731662739,
+      "grad_norm": 3.6793339854258833,
+      "learning_rate": 9.214700626757667e-08,
+      "loss": 3.0163,
+      "step": 4980
+    },
+    {
+      "epoch": 0.9582992641046607,
+      "grad_norm": 3.4758320501389695,
+      "learning_rate": 9.130484571950538e-08,
+      "loss": 2.8923,
+      "step": 4981
+    },
+    {
+      "epoch": 0.9584916550430475,
+      "grad_norm": 3.234826725314603,
+      "learning_rate": 9.046653359445323e-08,
+      "loss": 2.8658,
+      "step": 4982
+    },
+    {
+      "epoch": 0.9586840459814343,
+      "grad_norm": 3.4278757850864885,
+      "learning_rate": 8.963207021801423e-08,
+      "loss": 2.9394,
+      "step": 4983
+    },
+    {
+      "epoch": 0.958876436919821,
+      "grad_norm": 3.607453336281342,
+      "learning_rate": 8.880145591428024e-08,
+      "loss": 3.0159,
+      "step": 4984
+    },
+    {
+      "epoch": 0.9590688278582079,
+      "grad_norm": 3.321028817440024,
+      "learning_rate": 8.797469100585432e-08,
+      "loss": 2.8166,
+      "step": 4985
+    },
+    {
+      "epoch": 0.9592612187965947,
+      "grad_norm": 3.602621152415298,
+      "learning_rate": 8.715177581384182e-08,
+      "loss": 2.8792,
+      "step": 4986
+    },
+    {
+      "epoch": 0.9594536097349815,
+      "grad_norm": 3.314838899657632,
+      "learning_rate": 8.633271065785486e-08,
+      "loss": 2.6383,
+      "step": 4987
+    },
+    {
+      "epoch": 0.9596460006733682,
+      "grad_norm": 3.4065468283364577,
+      "learning_rate": 8.551749585600678e-08,
+      "loss": 2.7899,
+      "step": 4988
+    },
+    {
+      "epoch": 0.9598383916117551,
+      "grad_norm": 3.638240498836958,
+      "learning_rate": 8.470613172491981e-08,
+      "loss": 2.7352,
+      "step": 4989
+    },
+    {
+      "epoch": 0.9600307825501418,
+      "grad_norm": 3.8962782764746215,
+      "learning_rate": 8.389861857971748e-08,
+      "loss": 3.1338,
+      "step": 4990
+    },
+    {
+      "epoch": 0.9602231734885287,
+      "grad_norm": 3.6569469105087404,
+      "learning_rate": 8.309495673402778e-08,
+      "loss": 2.8797,
+      "step": 4991
+    },
+    {
+      "epoch": 0.9604155644269154,
+      "grad_norm": 3.22556300118932,
+      "learning_rate": 8.229514649998438e-08,
+      "loss": 2.8806,
+      "step": 4992
+    },
+    {
+      "epoch": 0.9606079553653023,
+      "grad_norm": 3.4628113472377327,
+      "learning_rate": 8.149918818822433e-08,
+      "loss": 3.0257,
+      "step": 4993
+    },
+    {
+      "epoch": 0.960800346303689,
+      "grad_norm": 3.3245252895261057,
+      "learning_rate": 8.070708210788925e-08,
+      "loss": 3.1105,
+      "step": 4994
+    },
+    {
+      "epoch": 0.9609927372420759,
+      "grad_norm": 3.2586413847620146,
+      "learning_rate": 7.991882856662303e-08,
+      "loss": 2.74,
+      "step": 4995
+    },
+    {
+      "epoch": 0.9611851281804628,
+      "grad_norm": 3.6983997146911554,
+      "learning_rate": 7.913442787057523e-08,
+      "loss": 2.997,
+      "step": 4996
+    },
+    {
+      "epoch": 0.9613775191188495,
+      "grad_norm": 3.728865878872687,
+      "learning_rate": 7.835388032439661e-08,
+      "loss": 3.0044,
+      "step": 4997
+    },
+    {
+      "epoch": 0.9615699100572364,
+      "grad_norm": 3.3396713107001457,
+      "learning_rate": 7.757718623124466e-08,
+      "loss": 2.7976,
+      "step": 4998
+    },
+    {
+      "epoch": 0.9617623009956231,
+      "grad_norm": 3.2806041655788705,
+      "learning_rate": 7.680434589277696e-08,
+      "loss": 2.6767,
+      "step": 4999
+    },
+    {
+      "epoch": 0.96195469193401,
+      "grad_norm": 3.5087879772907513,
+      "learning_rate": 7.603535960915675e-08,
+      "loss": 2.8417,
+      "step": 5000
+    },
+    {
+      "epoch": 0.9621470828723967,
+      "grad_norm": 3.2699514615240957,
+      "learning_rate": 7.527022767904957e-08,
+      "loss": 2.6364,
+      "step": 5001
+    },
+    {
+      "epoch": 0.9623394738107836,
+      "grad_norm": 3.917448984057674,
+      "learning_rate": 7.450895039962214e-08,
+      "loss": 2.9609,
+      "step": 5002
+    },
+    {
+      "epoch": 0.9625318647491703,
+      "grad_norm": 3.446083063914894,
+      "learning_rate": 7.375152806654685e-08,
+      "loss": 2.8442,
+      "step": 5003
+    },
+    {
+      "epoch": 0.9627242556875572,
+      "grad_norm": 3.8821187383451097,
+      "learning_rate": 7.299796097399947e-08,
+      "loss": 2.9802,
+      "step": 5004
+    },
+    {
+      "epoch": 0.9629166466259439,
+      "grad_norm": 3.587561181921725,
+      "learning_rate": 7.224824941465369e-08,
+      "loss": 2.9666,
+      "step": 5005
+    },
+    {
+      "epoch": 0.9631090375643307,
+      "grad_norm": 3.3314002139915067,
+      "learning_rate": 7.150239367969102e-08,
+      "loss": 2.746,
+      "step": 5006
+    },
+    {
+      "epoch": 0.9633014285027175,
+      "grad_norm": 3.4150347960644543,
+      "learning_rate": 7.076039405879309e-08,
+      "loss": 2.874,
+      "step": 5007
+    },
+    {
+      "epoch": 0.9634938194411043,
+      "grad_norm": 3.3571370115597974,
+      "learning_rate": 7.002225084014269e-08,
+      "loss": 2.8876,
+      "step": 5008
+    },
+    {
+      "epoch": 0.9636862103794911,
+      "grad_norm": 3.8293119933823054,
+      "learning_rate": 6.928796431042717e-08,
+      "loss": 3.127,
+      "step": 5009
+    },
+    {
+      "epoch": 0.963878601317878,
+      "grad_norm": 3.5663012287383076,
+      "learning_rate": 6.855753475483507e-08,
+      "loss": 2.7217,
+      "step": 5010
+    },
+    {
+      "epoch": 0.9640709922562647,
+      "grad_norm": 3.5751988457608426,
+      "learning_rate": 6.783096245705612e-08,
+      "loss": 3.0021,
+      "step": 5011
+    },
+    {
+      "epoch": 0.9642633831946515,
+      "grad_norm": 3.6308492006059208,
+      "learning_rate": 6.710824769928349e-08,
+      "loss": 3.1751,
+      "step": 5012
+    },
+    {
+      "epoch": 0.9644557741330383,
+      "grad_norm": 3.3938805133052674,
+      "learning_rate": 6.63893907622104e-08,
+      "loss": 3.0545,
+      "step": 5013
+    },
+    {
+      "epoch": 0.9646481650714251,
+      "grad_norm": 3.252625800450952,
+      "learning_rate": 6.567439192503244e-08,
+      "loss": 2.7425,
+      "step": 5014
+    },
+    {
+      "epoch": 0.9648405560098119,
+      "grad_norm": 3.647871041799464,
+      "learning_rate": 6.496325146544746e-08,
+      "loss": 3.0538,
+      "step": 5015
+    },
+    {
+      "epoch": 0.9650329469481987,
+      "grad_norm": 3.4970424201027375,
+      "learning_rate": 6.425596965965453e-08,
+      "loss": 3.0678,
+      "step": 5016
+    },
+    {
+      "epoch": 0.9652253378865855,
+      "grad_norm": 3.4996208484298577,
+      "learning_rate": 6.35525467823539e-08,
+      "loss": 2.9301,
+      "step": 5017
+    },
+    {
+      "epoch": 0.9654177288249723,
+      "grad_norm": 3.371373131719671,
+      "learning_rate": 6.285298310674703e-08,
+      "loss": 2.8659,
+      "step": 5018
+    },
+    {
+      "epoch": 0.9656101197633592,
+      "grad_norm": 3.4174262206719006,
+      "learning_rate": 6.215727890453438e-08,
+      "loss": 3.0598,
+      "step": 5019
+    },
+    {
+      "epoch": 0.9658025107017459,
+      "grad_norm": 3.292708233057196,
+      "learning_rate": 6.14654344459209e-08,
+      "loss": 2.9556,
+      "step": 5020
+    },
+    {
+      "epoch": 0.9659949016401328,
+      "grad_norm": 3.709563564972764,
+      "learning_rate": 6.077744999961166e-08,
+      "loss": 2.9719,
+      "step": 5021
+    },
+    {
+      "epoch": 0.9661872925785195,
+      "grad_norm": 2.4824906543404075,
+      "learning_rate": 6.009332583281069e-08,
+      "loss": 2.6038,
+      "step": 5022
+    },
+    {
+      "epoch": 0.9663796835169064,
+      "grad_norm": 3.643344320532489,
+      "learning_rate": 5.941306221122545e-08,
+      "loss": 2.8494,
+      "step": 5023
+    },
+    {
+      "epoch": 0.9665720744552931,
+      "grad_norm": 3.6630800378241997,
+      "learning_rate": 5.873665939906015e-08,
+      "loss": 3.0733,
+      "step": 5024
+    },
+    {
+      "epoch": 0.96676446539368,
+      "grad_norm": 3.7865959207202025,
+      "learning_rate": 5.806411765902353e-08,
+      "loss": 2.9821,
+      "step": 5025
+    },
+    {
+      "epoch": 0.9669568563320667,
+      "grad_norm": 3.2776751260809855,
+      "learning_rate": 5.739543725232333e-08,
+      "loss": 3.121,
+      "step": 5026
+    },
+    {
+      "epoch": 0.9671492472704536,
+      "grad_norm": 3.686803879775525,
+      "learning_rate": 5.673061843866623e-08,
+      "loss": 2.6,
+      "step": 5027
+    },
+    {
+      "epoch": 0.9673416382088403,
+      "grad_norm": 3.746208705945964,
+      "learning_rate": 5.606966147626125e-08,
+      "loss": 2.9449,
+      "step": 5028
+    },
+    {
+      "epoch": 0.9675340291472272,
+      "grad_norm": 3.4871017116612446,
+      "learning_rate": 5.5412566621815266e-08,
+      "loss": 2.9422,
+      "step": 5029
+    },
+    {
+      "epoch": 0.9677264200856139,
+      "grad_norm": 3.1902671580860384,
+      "learning_rate": 5.4759334130536355e-08,
+      "loss": 2.9667,
+      "step": 5030
+    },
+    {
+      "epoch": 0.9679188110240008,
+      "grad_norm": 2.253434411741728,
+      "learning_rate": 5.41099642561338e-08,
+      "loss": 2.5024,
+      "step": 5031
+    },
+    {
+      "epoch": 0.9681112019623875,
+      "grad_norm": 3.519208137881807,
+      "learning_rate": 5.3464457250814726e-08,
+      "loss": 2.7015,
+      "step": 5032
+    },
+    {
+      "epoch": 0.9683035929007744,
+      "grad_norm": 2.2583910342783677,
+      "learning_rate": 5.2822813365286386e-08,
+      "loss": 2.4964,
+      "step": 5033
+    },
+    {
+      "epoch": 0.9684959838391611,
+      "grad_norm": 3.51344015583975,
+      "learning_rate": 5.218503284875609e-08,
+      "loss": 2.8746,
+      "step": 5034
+    },
+    {
+      "epoch": 0.968688374777548,
+      "grad_norm": 3.322347997539002,
+      "learning_rate": 5.155111594893014e-08,
+      "loss": 3.0046,
+      "step": 5035
+    },
+    {
+      "epoch": 0.9688807657159347,
+      "grad_norm": 3.726037041650056,
+      "learning_rate": 5.092106291201604e-08,
+      "loss": 3.0166,
+      "step": 5036
+    },
+    {
+      "epoch": 0.9690731566543216,
+      "grad_norm": 3.632535303698385,
+      "learning_rate": 5.029487398271693e-08,
+      "loss": 2.8763,
+      "step": 5037
+    },
+    {
+      "epoch": 0.9692655475927083,
+      "grad_norm": 3.5716838284350754,
+      "learning_rate": 4.967254940423716e-08,
+      "loss": 2.7405,
+      "step": 5038
+    },
+    {
+      "epoch": 0.9694579385310952,
+      "grad_norm": 3.7647217350241498,
+      "learning_rate": 4.905408941828338e-08,
+      "loss": 2.7308,
+      "step": 5039
+    },
+    {
+      "epoch": 0.969650329469482,
+      "grad_norm": 3.5792286800352735,
+      "learning_rate": 4.8439494265055674e-08,
+      "loss": 2.9915,
+      "step": 5040
+    },
+    {
+      "epoch": 0.9698427204078688,
+      "grad_norm": 3.8555174473883147,
+      "learning_rate": 4.7828764183257545e-08,
+      "loss": 3.1554,
+      "step": 5041
+    },
+    {
+      "epoch": 0.9700351113462556,
+      "grad_norm": 3.6786233011846656,
+      "learning_rate": 4.722189941008703e-08,
+      "loss": 3.1179,
+      "step": 5042
+    },
+    {
+      "epoch": 0.9702275022846424,
+      "grad_norm": 3.423046572273997,
+      "learning_rate": 4.66189001812456e-08,
+      "loss": 2.8633,
+      "step": 5043
+    },
+    {
+      "epoch": 0.9704198932230292,
+      "grad_norm": 3.414206563108038,
+      "learning_rate": 4.6019766730930336e-08,
+      "loss": 2.6996,
+      "step": 5044
+    },
+    {
+      "epoch": 0.970612284161416,
+      "grad_norm": 3.255155407549468,
+      "learning_rate": 4.542449929183845e-08,
+      "loss": 2.8769,
+      "step": 5045
+    },
+    {
+      "epoch": 0.9708046750998028,
+      "grad_norm": 3.2819089258442005,
+      "learning_rate": 4.4833098095165005e-08,
+      "loss": 2.9245,
+      "step": 5046
+    },
+    {
+      "epoch": 0.9709970660381896,
+      "grad_norm": 3.5050309108633124,
+      "learning_rate": 4.424556337060182e-08,
+      "loss": 2.8467,
+      "step": 5047
+    },
+    {
+      "epoch": 0.9711894569765764,
+      "grad_norm": 3.847263852578572,
+      "learning_rate": 4.366189534634191e-08,
+      "loss": 2.849,
+      "step": 5048
+    },
+    {
+      "epoch": 0.9713818479149632,
+      "grad_norm": 3.419043895484552,
+      "learning_rate": 4.308209424907506e-08,
+      "loss": 3.0381,
+      "step": 5049
+    },
+    {
+      "epoch": 0.97157423885335,
+      "grad_norm": 3.4543873616066856,
+      "learning_rate": 4.250616030399002e-08,
+      "loss": 2.9384,
+      "step": 5050
+    },
+    {
+      "epoch": 0.9717666297917368,
+      "grad_norm": 3.332655762256418,
+      "learning_rate": 4.1934093734771194e-08,
+      "loss": 2.7837,
+      "step": 5051
+    },
+    {
+      "epoch": 0.9719590207301236,
+      "grad_norm": 3.1744642032479735,
+      "learning_rate": 4.13658947636042e-08,
+      "loss": 2.8148,
+      "step": 5052
+    },
+    {
+      "epoch": 0.9721514116685104,
+      "grad_norm": 3.3530036718886294,
+      "learning_rate": 4.080156361117027e-08,
+      "loss": 2.8659,
+      "step": 5053
+    },
+    {
+      "epoch": 0.9723438026068972,
+      "grad_norm": 3.5694019074289023,
+      "learning_rate": 4.024110049664853e-08,
+      "loss": 2.8021,
+      "step": 5054
+    },
+    {
+      "epoch": 0.972536193545284,
+      "grad_norm": 3.4512419391854245,
+      "learning_rate": 3.9684505637718194e-08,
+      "loss": 2.8027,
+      "step": 5055
+    },
+    {
+      "epoch": 0.9727285844836708,
+      "grad_norm": 3.7452484031706357,
+      "learning_rate": 3.913177925055189e-08,
+      "loss": 2.9363,
+      "step": 5056
+    },
+    {
+      "epoch": 0.9729209754220576,
+      "grad_norm": 3.6245553050288466,
+      "learning_rate": 3.858292154982457e-08,
+      "loss": 3.0772,
+      "step": 5057
+    },
+    {
+      "epoch": 0.9731133663604444,
+      "grad_norm": 3.236572530037318,
+      "learning_rate": 3.8037932748704596e-08,
+      "loss": 3.0049,
+      "step": 5058
+    },
+    {
+      "epoch": 0.9733057572988312,
+      "grad_norm": 3.2956646736087283,
+      "learning_rate": 3.7496813058859325e-08,
+      "loss": 2.7385,
+      "step": 5059
+    },
+    {
+      "epoch": 0.973498148237218,
+      "grad_norm": 3.3368301379965204,
+      "learning_rate": 3.695956269045509e-08,
+      "loss": 2.841,
+      "step": 5060
+    },
+    {
+      "epoch": 0.9736905391756048,
+      "grad_norm": 3.329505276786651,
+      "learning_rate": 3.642618185215163e-08,
+      "loss": 2.7925,
+      "step": 5061
+    },
+    {
+      "epoch": 0.9738829301139916,
+      "grad_norm": 3.8765733579905053,
+      "learning_rate": 3.589667075110992e-08,
+      "loss": 2.9972,
+      "step": 5062
+    },
+    {
+      "epoch": 0.9740753210523785,
+      "grad_norm": 2.2541971389774726,
+      "learning_rate": 3.537102959298322e-08,
+      "loss": 2.5059,
+      "step": 5063
+    },
+    {
+      "epoch": 0.9742677119907652,
+      "grad_norm": 2.324958266259294,
+      "learning_rate": 3.4849258581928224e-08,
+      "loss": 2.7636,
+      "step": 5064
+    },
+    {
+      "epoch": 0.9744601029291521,
+      "grad_norm": 3.2341711156082646,
+      "learning_rate": 3.4331357920591726e-08,
+      "loss": 2.7685,
+      "step": 5065
+    },
+    {
+      "epoch": 0.9746524938675388,
+      "grad_norm": 3.2477453303544745,
+      "learning_rate": 3.3817327810121706e-08,
+      "loss": 2.7318,
+      "step": 5066
+    },
+    {
+      "epoch": 0.9748448848059257,
+      "grad_norm": 3.4230373468392843,
+      "learning_rate": 3.3307168450160685e-08,
+      "loss": 2.9418,
+      "step": 5067
+    },
+    {
+      "epoch": 0.9750372757443124,
+      "grad_norm": 3.715791620019362,
+      "learning_rate": 3.280088003885018e-08,
+      "loss": 2.6694,
+      "step": 5068
+    },
+    {
+      "epoch": 0.9752296666826993,
+      "grad_norm": 3.6365348441956136,
+      "learning_rate": 3.229846277282511e-08,
+      "loss": 3.0416,
+      "step": 5069
+    },
+    {
+      "epoch": 0.975422057621086,
+      "grad_norm": 3.5653679945581023,
+      "learning_rate": 3.1799916847220505e-08,
+      "loss": 3.018,
+      "step": 5070
+    },
+    {
+      "epoch": 0.9756144485594729,
+      "grad_norm": 3.2813060952983304,
+      "learning_rate": 3.130524245566369e-08,
+      "loss": 3.0174,
+      "step": 5071
+    },
+    {
+      "epoch": 0.9758068394978596,
+      "grad_norm": 3.6698047371123206,
+      "learning_rate": 3.0814439790280984e-08,
+      "loss": 2.791,
+      "step": 5072
+    },
+    {
+      "epoch": 0.9759992304362465,
+      "grad_norm": 3.4919205933678557,
+      "learning_rate": 3.032750904169546e-08,
+      "loss": 2.9297,
+      "step": 5073
+    },
+    {
+      "epoch": 0.9761916213746332,
+      "grad_norm": 3.798959895846842,
+      "learning_rate": 2.9844450399024725e-08,
+      "loss": 2.9761,
+      "step": 5074
+    },
+    {
+      "epoch": 0.9763840123130201,
+      "grad_norm": 3.631520579077704,
+      "learning_rate": 2.936526404988427e-08,
+      "loss": 2.8516,
+      "step": 5075
+    },
+    {
+      "epoch": 0.9765764032514068,
+      "grad_norm": 3.323457883404211,
+      "learning_rate": 2.8889950180382985e-08,
+      "loss": 2.839,
+      "step": 5076
+    },
+    {
+      "epoch": 0.9767687941897937,
+      "grad_norm": 3.8511880113343637,
+      "learning_rate": 2.8418508975127655e-08,
+      "loss": 3.1393,
+      "step": 5077
+    },
+    {
+      "epoch": 0.9769611851281804,
+      "grad_norm": 3.361164315579236,
+      "learning_rate": 2.7950940617221812e-08,
+      "loss": 3.0264,
+      "step": 5078
+    },
+    {
+      "epoch": 0.9771535760665673,
+      "grad_norm": 3.6611241187187478,
+      "learning_rate": 2.7487245288261298e-08,
+      "loss": 2.8541,
+      "step": 5079
+    },
+    {
+      "epoch": 0.977345967004954,
+      "grad_norm": 3.4978540735844432,
+      "learning_rate": 2.702742316834206e-08,
+      "loss": 2.9232,
+      "step": 5080
+    },
+    {
+      "epoch": 0.9775383579433409,
+      "grad_norm": 3.693839907895843,
+      "learning_rate": 2.657147443605457e-08,
+      "loss": 3.103,
+      "step": 5081
+    },
+    {
+      "epoch": 0.9777307488817276,
+      "grad_norm": 2.1829254228274904,
+      "learning_rate": 2.6119399268480505e-08,
+      "loss": 2.5308,
+      "step": 5082
+    },
+    {
+      "epoch": 0.9779231398201145,
+      "grad_norm": 3.898861241853256,
+      "learning_rate": 2.5671197841203867e-08,
+      "loss": 2.9797,
+      "step": 5083
+    },
+    {
+      "epoch": 0.9781155307585013,
+      "grad_norm": 3.453216574326729,
+      "learning_rate": 2.5226870328299845e-08,
+      "loss": 2.9276,
+      "step": 5084
+    },
+    {
+      "epoch": 0.9783079216968881,
+      "grad_norm": 3.700588428780622,
+      "learning_rate": 2.4786416902339293e-08,
+      "loss": 2.9724,
+      "step": 5085
+    },
+    {
+      "epoch": 0.9785003126352749,
+      "grad_norm": 3.461331995830382,
+      "learning_rate": 2.4349837734390923e-08,
+      "loss": 2.9238,
+      "step": 5086
+    },
+    {
+      "epoch": 0.9786927035736617,
+      "grad_norm": 3.2802268762308606,
+      "learning_rate": 2.3917132994016878e-08,
+      "loss": 2.8489,
+      "step": 5087
+    },
+    {
+      "epoch": 0.9788850945120485,
+      "grad_norm": 3.4163856006732707,
+      "learning_rate": 2.3488302849272726e-08,
+      "loss": 2.8574,
+      "step": 5088
+    },
+    {
+      "epoch": 0.9790774854504353,
+      "grad_norm": 3.6151907420822535,
+      "learning_rate": 2.3063347466713014e-08,
+      "loss": 3.0354,
+      "step": 5089
+    },
+    {
+      "epoch": 0.9792698763888221,
+      "grad_norm": 3.8190827325976837,
+      "learning_rate": 2.264226701138461e-08,
+      "loss": 3.1112,
+      "step": 5090
+    },
+    {
+      "epoch": 0.9794622673272089,
+      "grad_norm": 3.6397353658104254,
+      "learning_rate": 2.2225061646830025e-08,
+      "loss": 2.9884,
+      "step": 5091
+    },
+    {
+      "epoch": 0.9796546582655957,
+      "grad_norm": 3.632280566037303,
+      "learning_rate": 2.181173153508853e-08,
+      "loss": 3.0774,
+      "step": 5092
+    },
+    {
+      "epoch": 0.9798470492039825,
+      "grad_norm": 3.6396873012958397,
+      "learning_rate": 2.1402276836691715e-08,
+      "loss": 3.0176,
+      "step": 5093
+    },
+    {
+      "epoch": 0.9800394401423693,
+      "grad_norm": 3.3712636939631717,
+      "learning_rate": 2.0996697710666812e-08,
+      "loss": 2.9784,
+      "step": 5094
+    },
+    {
+      "epoch": 0.9802318310807561,
+      "grad_norm": 3.3984939552182185,
+      "learning_rate": 2.059499431453671e-08,
+      "loss": 3.0224,
+      "step": 5095
+    },
+    {
+      "epoch": 0.9804242220191429,
+      "grad_norm": 3.3103472409591443,
+      "learning_rate": 2.0197166804317713e-08,
+      "loss": 2.8637,
+      "step": 5096
+    },
+    {
+      "epoch": 0.9806166129575297,
+      "grad_norm": 3.242625704747527,
+      "learning_rate": 1.9803215334522895e-08,
+      "loss": 2.8416,
+      "step": 5097
+    },
+    {
+      "epoch": 0.9808090038959165,
+      "grad_norm": 3.1941404595435445,
+      "learning_rate": 1.9413140058156533e-08,
+      "loss": 2.5209,
+      "step": 5098
+    },
+    {
+      "epoch": 0.9810013948343033,
+      "grad_norm": 3.7031798595500174,
+      "learning_rate": 1.9026941126721876e-08,
+      "loss": 2.7641,
+      "step": 5099
+    },
+    {
+      "epoch": 0.9811937857726901,
+      "grad_norm": 3.530454332635899,
+      "learning_rate": 1.864461869021117e-08,
+      "loss": 2.8539,
+      "step": 5100
+    },
+    {
+      "epoch": 0.9813861767110769,
+      "grad_norm": 3.4974422783301877,
+      "learning_rate": 1.826617289711563e-08,
+      "loss": 2.8804,
+      "step": 5101
+    },
+    {
+      "epoch": 0.9815785676494637,
+      "grad_norm": 3.5394025877092017,
+      "learning_rate": 1.7891603894418797e-08,
+      "loss": 3.0427,
+      "step": 5102
+    },
+    {
+      "epoch": 0.9817709585878505,
+      "grad_norm": 3.451521865034916,
+      "learning_rate": 1.7520911827598742e-08,
+      "loss": 2.945,
+      "step": 5103
+    },
+    {
+      "epoch": 0.9819633495262373,
+      "grad_norm": 3.5046046131511184,
+      "learning_rate": 1.7154096840629186e-08,
+      "loss": 2.9249,
+      "step": 5104
+    },
+    {
+      "epoch": 0.9821557404646241,
+      "grad_norm": 3.3908794965837967,
+      "learning_rate": 1.679115907597617e-08,
+      "loss": 3.0861,
+      "step": 5105
+    },
+    {
+      "epoch": 0.9823481314030109,
+      "grad_norm": 3.602796861867629,
+      "learning_rate": 1.6432098674600272e-08,
+      "loss": 2.9085,
+      "step": 5106
+    },
+    {
+      "epoch": 0.9825405223413978,
+      "grad_norm": 3.676907447915402,
+      "learning_rate": 1.6076915775956604e-08,
+      "loss": 2.6108,
+      "step": 5107
+    },
+    {
+      "epoch": 0.9827329132797845,
+      "grad_norm": 3.2510366207189483,
+      "learning_rate": 1.5725610517994816e-08,
+      "loss": 2.9875,
+      "step": 5108
+    },
+    {
+      "epoch": 0.9829253042181714,
+      "grad_norm": 3.3936615198321127,
+      "learning_rate": 1.537818303715688e-08,
+      "loss": 3.0913,
+      "step": 5109
+    },
+    {
+      "epoch": 0.9831176951565581,
+      "grad_norm": 3.3658118800068513,
+      "learning_rate": 1.50346334683793e-08,
+      "loss": 2.9985,
+      "step": 5110
+    },
+    {
+      "epoch": 0.983310086094945,
+      "grad_norm": 3.634433596846686,
+      "learning_rate": 1.4694961945093122e-08,
+      "loss": 3.0384,
+      "step": 5111
+    },
+    {
+      "epoch": 0.9835024770333317,
+      "grad_norm": 3.519520946419951,
+      "learning_rate": 1.4359168599223926e-08,
+      "loss": 2.8823,
+      "step": 5112
+    },
+    {
+      "epoch": 0.9836948679717186,
+      "grad_norm": 3.429973630486305,
+      "learning_rate": 1.4027253561188502e-08,
+      "loss": 2.8,
+      "step": 5113
+    },
+    {
+      "epoch": 0.9838872589101053,
+      "grad_norm": 3.630995675698039,
+      "learning_rate": 1.3699216959899287e-08,
+      "loss": 2.6409,
+      "step": 5114
+    },
+    {
+      "epoch": 0.9840796498484922,
+      "grad_norm": 3.391084824987237,
+      "learning_rate": 1.337505892276325e-08,
+      "loss": 2.8174,
+      "step": 5115
+    },
+    {
+      "epoch": 0.9842720407868789,
+      "grad_norm": 3.6186972396635517,
+      "learning_rate": 1.3054779575677468e-08,
+      "loss": 3.1336,
+      "step": 5116
+    },
+    {
+      "epoch": 0.9844644317252658,
+      "grad_norm": 3.487383792299187,
+      "learning_rate": 1.2738379043035765e-08,
+      "loss": 3.0221,
+      "step": 5117
+    },
+    {
+      "epoch": 0.9846568226636525,
+      "grad_norm": 2.426547785656183,
+      "learning_rate": 1.2425857447725397e-08,
+      "loss": 2.576,
+      "step": 5118
+    },
+    {
+      "epoch": 0.9848492136020394,
+      "grad_norm": 3.84263172875224,
+      "learning_rate": 1.211721491112372e-08,
+      "loss": 2.9527,
+      "step": 5119
+    },
+    {
+      "epoch": 0.9850416045404261,
+      "grad_norm": 3.506614685012352,
+      "learning_rate": 1.1812451553107063e-08,
+      "loss": 2.9463,
+      "step": 5120
+    },
+    {
+      "epoch": 0.985233995478813,
+      "grad_norm": 3.4827238017496747,
+      "learning_rate": 1.1511567492038522e-08,
+      "loss": 2.8849,
+      "step": 5121
+    },
+    {
+      "epoch": 0.9854263864171997,
+      "grad_norm": 3.364592451852956,
+      "learning_rate": 1.1214562844781285e-08,
+      "loss": 2.7324,
+      "step": 5122
+    },
+    {
+      "epoch": 0.9856187773555866,
+      "grad_norm": 3.238494778344058,
+      "learning_rate": 1.0921437726686413e-08,
+      "loss": 2.6571,
+      "step": 5123
+    },
+    {
+      "epoch": 0.9858111682939733,
+      "grad_norm": 3.6241737909613487,
+      "learning_rate": 1.0632192251601726e-08,
+      "loss": 2.982,
+      "step": 5124
+    },
+    {
+      "epoch": 0.9860035592323602,
+      "grad_norm": 3.4256042304642587,
+      "learning_rate": 1.0346826531865139e-08,
+      "loss": 2.9874,
+      "step": 5125
+    },
+    {
+      "epoch": 0.9861959501707469,
+      "grad_norm": 3.424191843332104,
+      "learning_rate": 1.006534067831022e-08,
+      "loss": 2.8869,
+      "step": 5126
+    },
+    {
+      "epoch": 0.9863883411091338,
+      "grad_norm": 3.47586997233516,
+      "learning_rate": 9.78773480026396e-09,
+      "loss": 3.05,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9865807320475206,
+      "grad_norm": 2.2292260798665144,
+      "learning_rate": 9.514009005543445e-09,
+      "loss": 2.5601,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9867731229859074,
+      "grad_norm": 3.390205300271474,
+      "learning_rate": 9.244163400462525e-09,
+      "loss": 2.8041,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9869655139242942,
+      "grad_norm": 3.6518165727930887,
+      "learning_rate": 8.97819808982403e-09,
+      "loss": 3.1096,
+      "step": 5130
+    },
+    {
+      "epoch": 0.987157904862681,
+      "grad_norm": 3.5647296257821415,
+      "learning_rate": 8.716113176927554e-09,
+      "loss": 2.9341,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9873502958010678,
+      "grad_norm": 3.5768676638006127,
+      "learning_rate": 8.457908763562783e-09,
+      "loss": 2.7749,
+      "step": 5132
+    },
+    {
+      "epoch": 0.9875426867394546,
+      "grad_norm": 3.6419924760763083,
+      "learning_rate": 8.203584950013942e-09,
+      "loss": 3.1078,
+      "step": 5133
+    },
+    {
+      "epoch": 0.9877350776778414,
+      "grad_norm": 3.428471248528095,
+      "learning_rate": 7.953141835057576e-09,
+      "loss": 2.9383,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9879274686162282,
+      "grad_norm": 2.313908686405991,
+      "learning_rate": 7.706579515962542e-09,
+      "loss": 2.5861,
+      "step": 5135
+    },
+    {
+      "epoch": 0.988119859554615,
+      "grad_norm": 3.413987952622437,
+      "learning_rate": 7.463898088490019e-09,
+      "loss": 2.9651,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9883122504930018,
+      "grad_norm": 3.2873937508499727,
+      "learning_rate": 7.225097646895718e-09,
+      "loss": 3.1213,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9885046414313886,
+      "grad_norm": 3.4647015878787024,
+      "learning_rate": 6.990178283927673e-09,
+      "loss": 2.9148,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9886970323697754,
+      "grad_norm": 3.7582028750117775,
+      "learning_rate": 6.759140090824012e-09,
+      "loss": 3.0838,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9888894233081622,
+      "grad_norm": 3.67433901728189,
+      "learning_rate": 6.531983157318511e-09,
+      "loss": 3.2672,
+      "step": 5140
+    },
+    {
+      "epoch": 0.989081814246549,
+      "grad_norm": 3.473399560782481,
+      "learning_rate": 6.308707571636152e-09,
+      "loss": 2.8971,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9892742051849358,
+      "grad_norm": 3.4944348307350617,
+      "learning_rate": 6.089313420494236e-09,
+      "loss": 2.9057,
+      "step": 5142
+    },
+    {
+      "epoch": 0.9894665961233226,
+      "grad_norm": 3.359863785740937,
+      "learning_rate": 5.87380078910349e-09,
+      "loss": 3.0416,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9896589870617094,
+      "grad_norm": 3.7449183109236572,
+      "learning_rate": 5.662169761165848e-09,
+      "loss": 2.9772,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9898513780000961,
+      "grad_norm": 3.6235937716988276,
+      "learning_rate": 5.4544204188777815e-09,
+      "loss": 2.931,
+      "step": 5145
+    },
+    {
+      "epoch": 0.990043768938483,
+      "grad_norm": 2.372606682057148,
+      "learning_rate": 5.25055284292475e-09,
+      "loss": 2.54,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9902361598768697,
+      "grad_norm": 3.5883414272274177,
+      "learning_rate": 5.050567112487858e-09,
+      "loss": 3.1431,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9904285508152566,
+      "grad_norm": 3.698605185309862,
+      "learning_rate": 4.85446330523831e-09,
+      "loss": 2.9528,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9906209417536435,
+      "grad_norm": 3.3602467073320783,
+      "learning_rate": 4.662241497341846e-09,
+      "loss": 2.9123,
+      "step": 5149
+    },
+    {
+      "epoch": 0.9908133326920302,
+      "grad_norm": 3.767161694083474,
+      "learning_rate": 4.473901763454302e-09,
+      "loss": 2.8736,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9910057236304171,
+      "grad_norm": 3.7840629015045883,
+      "learning_rate": 4.289444176724944e-09,
+      "loss": 2.8853,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9911981145688038,
+      "grad_norm": 3.3024908067751837,
+      "learning_rate": 4.108868808794242e-09,
+      "loss": 2.8441,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9913905055071907,
+      "grad_norm": 3.520408767818994,
+      "learning_rate": 3.932175729797205e-09,
+      "loss": 2.9767,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9915828964455774,
+      "grad_norm": 3.3928729661898775,
+      "learning_rate": 3.759365008357829e-09,
+      "loss": 3.1331,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9917752873839643,
+      "grad_norm": 3.571729739434865,
+      "learning_rate": 3.590436711594647e-09,
+      "loss": 3.0284,
+      "step": 5155
+    },
+    {
+      "epoch": 0.991967678322351,
+      "grad_norm": 3.306810220560557,
+      "learning_rate": 3.4253909051173985e-09,
+      "loss": 2.7705,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9921600692607379,
+      "grad_norm": 3.3488685177932767,
+      "learning_rate": 3.2642276530281404e-09,
+      "loss": 2.9735,
+      "step": 5157
+    },
+    {
+      "epoch": 0.9923524601991246,
+      "grad_norm": 3.569891611954561,
+      "learning_rate": 3.1069470179201365e-09,
+      "loss": 2.7811,
+      "step": 5158
+    },
+    {
+      "epoch": 0.9925448511375115,
+      "grad_norm": 3.437004877060182,
+      "learning_rate": 2.9535490608789687e-09,
+      "loss": 2.9165,
+      "step": 5159
+    },
+    {
+      "epoch": 0.9927372420758982,
+      "grad_norm": 3.536159015305853,
+      "learning_rate": 2.8040338414847545e-09,
+      "loss": 2.9626,
+      "step": 5160
+    },
+    {
+      "epoch": 0.992929633014285,
+      "grad_norm": 3.6651423076596155,
+      "learning_rate": 2.6584014178054894e-09,
+      "loss": 2.9526,
+      "step": 5161
+    },
+    {
+      "epoch": 0.9931220239526718,
+      "grad_norm": 3.563997087476329,
+      "learning_rate": 2.516651846403706e-09,
+      "loss": 3.0343,
+      "step": 5162
+    },
+    {
+      "epoch": 0.9933144148910587,
+      "grad_norm": 3.60641847601551,
+      "learning_rate": 2.378785182333143e-09,
+      "loss": 3.1772,
+      "step": 5163
+    },
+    {
+      "epoch": 0.9935068058294454,
+      "grad_norm": 3.4028080496800563,
+      "learning_rate": 2.2448014791398574e-09,
+      "loss": 3.0418,
+      "step": 5164
+    },
+    {
+      "epoch": 0.9936991967678322,
+      "grad_norm": 3.6070764368745984,
+      "learning_rate": 2.1147007888622227e-09,
+      "loss": 3.1816,
+      "step": 5165
+    },
+    {
+      "epoch": 0.993891587706219,
+      "grad_norm": 3.2175198451479834,
+      "learning_rate": 1.9884831620287094e-09,
+      "loss": 2.8865,
+      "step": 5166
+    },
+    {
+      "epoch": 0.9940839786446058,
+      "grad_norm": 3.478283021550019,
+      "learning_rate": 1.8661486476612144e-09,
+      "loss": 2.8712,
+      "step": 5167
+    },
+    {
+      "epoch": 0.9942763695829926,
+      "grad_norm": 3.5628464702262166,
+      "learning_rate": 1.747697293272843e-09,
+      "loss": 2.8557,
+      "step": 5168
+    },
+    {
+      "epoch": 0.9944687605213794,
+      "grad_norm": 3.2335017763397103,
+      "learning_rate": 1.633129144870127e-09,
+      "loss": 2.9712,
+      "step": 5169
+    },
+    {
+      "epoch": 0.9946611514597662,
+      "grad_norm": 3.3591864573285326,
+      "learning_rate": 1.5224442469474743e-09,
+      "loss": 2.976,
+      "step": 5170
+    },
+    {
+      "epoch": 0.994853542398153,
+      "grad_norm": 3.706058341630448,
+      "learning_rate": 1.4156426424960513e-09,
+      "loss": 2.9391,
+      "step": 5171
+    },
+    {
+      "epoch": 0.9950459333365399,
+      "grad_norm": 3.6202319504327773,
+      "learning_rate": 1.3127243729949003e-09,
+      "loss": 3.1098,
+      "step": 5172
+    },
+    {
+      "epoch": 0.9952383242749266,
+      "grad_norm": 3.413383031344289,
+      "learning_rate": 1.2136894784176011e-09,
+      "loss": 3.1228,
+      "step": 5173
+    },
+    {
+      "epoch": 0.9954307152133135,
+      "grad_norm": 3.6989521758796973,
+      "learning_rate": 1.1185379972256105e-09,
+      "loss": 2.9544,
+      "step": 5174
+    },
+    {
+      "epoch": 0.9956231061517002,
+      "grad_norm": 3.654354326080439,
+      "learning_rate": 1.0272699663782525e-09,
+      "loss": 2.9917,
+      "step": 5175
+    },
+    {
+      "epoch": 0.9958154970900871,
+      "grad_norm": 3.5257003365896917,
+      "learning_rate": 9.398854213193976e-10,
+      "loss": 2.7562,
+      "step": 5176
+    },
+    {
+      "epoch": 0.9960078880284738,
+      "grad_norm": 3.608460964231656,
+      "learning_rate": 8.563843959907836e-10,
+      "loss": 2.8731,
+      "step": 5177
+    },
+    {
+      "epoch": 0.9962002789668607,
+      "grad_norm": 3.313004541084495,
+      "learning_rate": 7.767669228231356e-10,
+      "loss": 2.9131,
+      "step": 5178
+    },
+    {
+      "epoch": 0.9963926699052474,
+      "grad_norm": 3.152464939975927,
+      "learning_rate": 7.010330327372749e-10,
+      "loss": 2.6153,
+      "step": 5179
+    },
+    {
+      "epoch": 0.9965850608436343,
+      "grad_norm": 3.882215378853476,
+      "learning_rate": 6.291827551474505e-10,
+      "loss": 3.1738,
+      "step": 5180
+    },
+    {
+      "epoch": 0.996777451782021,
+      "grad_norm": 3.2745025738396856,
+      "learning_rate": 5.612161179613385e-10,
+      "loss": 2.9891,
+      "step": 5181
+    },
+    {
+      "epoch": 0.9969698427204079,
+      "grad_norm": 3.574201400460626,
+      "learning_rate": 4.971331475756013e-10,
+      "loss": 2.7513,
+      "step": 5182
+    },
+    {
+      "epoch": 0.9971622336587946,
+      "grad_norm": 3.42646380076646,
+      "learning_rate": 4.369338688781088e-10,
+      "loss": 2.9006,
+      "step": 5183
+    },
+    {
+      "epoch": 0.9973546245971815,
+      "grad_norm": 3.28931870598415,
+      "learning_rate": 3.8061830525126797e-10,
+      "loss": 3.0113,
+      "step": 5184
+    },
+    {
+      "epoch": 0.9975470155355682,
+      "grad_norm": 3.184276539862303,
+      "learning_rate": 3.2818647856647234e-10,
+      "loss": 2.9302,
+      "step": 5185
+    },
+    {
+      "epoch": 0.9977394064739551,
+      "grad_norm": 3.045616513547693,
+      "learning_rate": 2.796384091885429e-10,
+      "loss": 2.726,
+      "step": 5186
+    },
+    {
+      "epoch": 0.9979317974123418,
+      "grad_norm": 3.3768690262319425,
+      "learning_rate": 2.3497411597128705e-10,
+      "loss": 2.9143,
+      "step": 5187
+    },
+    {
+      "epoch": 0.9981241883507287,
+      "grad_norm": 3.3818925973699088,
+      "learning_rate": 1.9419361626416e-10,
+      "loss": 3.0964,
+      "step": 5188
+    },
+    {
+      "epoch": 0.9983165792891154,
+      "grad_norm": 3.427853866560431,
+      "learning_rate": 1.5729692590338296e-10,
+      "loss": 3.2577,
+      "step": 5189
+    },
+    {
+      "epoch": 0.9985089702275023,
+      "grad_norm": 3.4011268789135007,
+      "learning_rate": 1.2428405921971476e-10,
+      "loss": 2.7463,
+      "step": 5190
+    },
+    {
+      "epoch": 0.998701361165889,
+      "grad_norm": 3.5129041262990497,
+      "learning_rate": 9.515502903734153e-11,
+      "loss": 2.9993,
+      "step": 5191
+    },
+    {
+      "epoch": 0.9988937521042759,
+      "grad_norm": 3.512362467026774,
+      "learning_rate": 6.990984666610523e-11,
+      "loss": 3.0921,
+      "step": 5192
+    },
+    {
+      "epoch": 0.9990861430426627,
+      "grad_norm": 3.8103974385069055,
+      "learning_rate": 4.854852191371606e-11,
+      "loss": 3.1893,
+      "step": 5193
+    },
+    {
+      "epoch": 0.9992785339810495,
+      "grad_norm": 3.5757166870484696,
+      "learning_rate": 3.1071063075760466e-11,
+      "loss": 3.1116,
+      "step": 5194
+    },
+    {
+      "epoch": 0.9994709249194363,
+      "grad_norm": 3.6975799613521656,
+      "learning_rate": 1.7477476940142013e-11,
+      "loss": 2.7586,
+      "step": 5195
+    },
+    {
+      "epoch": 0.9996633158578231,
+      "grad_norm": 3.481634744874384,
+      "learning_rate": 7.767768785971185e-12,
+      "loss": 2.8938,
+      "step": 5196
+    },
+    {
+      "epoch": 0.9998557067962099,
+      "grad_norm": 4.199112067779305,
+      "learning_rate": 1.9419423846755993e-12,
+      "loss": 2.2793,
+      "step": 5197
+    },
+    {
+      "epoch": 0.9998557067962099,
+      "step": 5197,
+      "total_flos": 1.1782245622218752e+16,
+      "train_loss": 3.131748151292887,
+      "train_runtime": 76857.5708,
+      "train_samples_per_second": 8.656,
+      "train_steps_per_second": 0.068
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 5197,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.1782245622218752e+16,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}