diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,24591 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 3507,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00028514399771884804,
+      "grad_norm": 73.35986033232847,
+      "learning_rate": 9.433962264150944e-08,
+      "loss": 0.9276,
+      "step": 1
+    },
+    {
+      "epoch": 0.0005702879954376961,
+      "grad_norm": 258.36563461213944,
+      "learning_rate": 1.886792452830189e-07,
+      "loss": 0.9933,
+      "step": 2
+    },
+    {
+      "epoch": 0.000855431993156544,
+      "grad_norm": 166.05860991922756,
+      "learning_rate": 2.8301886792452833e-07,
+      "loss": 0.8841,
+      "step": 3
+    },
+    {
+      "epoch": 0.0011405759908753922,
+      "grad_norm": 80.04075689751123,
+      "learning_rate": 3.773584905660378e-07,
+      "loss": 0.8703,
+      "step": 4
+    },
+    {
+      "epoch": 0.00142571998859424,
+      "grad_norm": 94.13324251180957,
+      "learning_rate": 4.716981132075472e-07,
+      "loss": 0.7985,
+      "step": 5
+    },
+    {
+      "epoch": 0.001710863986313088,
+      "grad_norm": 95.32563753359732,
+      "learning_rate": 5.660377358490567e-07,
+      "loss": 0.8165,
+      "step": 6
+    },
+    {
+      "epoch": 0.001996007984031936,
+      "grad_norm": 107.85429854364156,
+      "learning_rate": 6.603773584905661e-07,
+      "loss": 0.7712,
+      "step": 7
+    },
+    {
+      "epoch": 0.0022811519817507843,
+      "grad_norm": 81.4209081472541,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.8074,
+      "step": 8
+    },
+    {
+      "epoch": 0.0025662959794696323,
+      "grad_norm": 102.25957565902327,
+      "learning_rate": 8.490566037735849e-07,
+      "loss": 0.71,
+      "step": 9
+    },
+    {
+      "epoch": 0.00285143997718848,
+      "grad_norm": 93.50952676307716,
+      "learning_rate": 9.433962264150944e-07,
+      "loss": 0.5691,
+      "step": 10
+    },
+    {
+      "epoch": 0.003136583974907328,
+      "grad_norm": 28.59917240226574,
+      "learning_rate": 1.037735849056604e-06,
+      "loss": 0.2667,
+      "step": 11
+    },
+    {
+      "epoch": 0.003421727972626176,
+      "grad_norm": 31.65592294166331,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.1884,
+      "step": 12
+    },
+    {
+      "epoch": 0.0037068719703450244,
+      "grad_norm": 19.767941232025557,
+      "learning_rate": 1.2264150943396227e-06,
+      "loss": 0.1949,
+      "step": 13
+    },
+    {
+      "epoch": 0.003992015968063872,
+      "grad_norm": 18.679288467520557,
+      "learning_rate": 1.3207547169811322e-06,
+      "loss": 0.1665,
+      "step": 14
+    },
+    {
+      "epoch": 0.00427715996578272,
+      "grad_norm": 11.019999790672726,
+      "learning_rate": 1.4150943396226415e-06,
+      "loss": 0.083,
+      "step": 15
+    },
+    {
+      "epoch": 0.004562303963501569,
+      "grad_norm": 10.59204980956432,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 0.1157,
+      "step": 16
+    },
+    {
+      "epoch": 0.004847447961220416,
+      "grad_norm": 3.1617167080707995,
+      "learning_rate": 1.6037735849056604e-06,
+      "loss": 0.0322,
+      "step": 17
+    },
+    {
+      "epoch": 0.0051325919589392645,
+      "grad_norm": 4.146835741088157,
+      "learning_rate": 1.6981132075471698e-06,
+      "loss": 0.0497,
+      "step": 18
+    },
+    {
+      "epoch": 0.005417735956658112,
+      "grad_norm": 7.254151558151055,
+      "learning_rate": 1.7924528301886793e-06,
+      "loss": 0.0879,
+      "step": 19
+    },
+    {
+      "epoch": 0.00570287995437696,
+      "grad_norm": 42.70760196636873,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.1596,
+      "step": 20
+    },
+    {
+      "epoch": 0.005988023952095809,
+      "grad_norm": 9.554484439254637,
+      "learning_rate": 1.981132075471698e-06,
+      "loss": 0.1138,
+      "step": 21
+    },
+    {
+      "epoch": 0.006273167949814656,
+      "grad_norm": 7.238913126784841,
+      "learning_rate": 2.075471698113208e-06,
+      "loss": 0.0812,
+      "step": 22
+    },
+    {
+      "epoch": 0.006558311947533505,
+      "grad_norm": 4.259308808989036,
+      "learning_rate": 2.1698113207547173e-06,
+      "loss": 0.0413,
+      "step": 23
+    },
+    {
+      "epoch": 0.006843455945252352,
+      "grad_norm": 6.566396501150489,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 0.0549,
+      "step": 24
+    },
+    {
+      "epoch": 0.0071285999429712005,
+      "grad_norm": 5.066416110741747,
+      "learning_rate": 2.358490566037736e-06,
+      "loss": 0.037,
+      "step": 25
+    },
+    {
+      "epoch": 0.007413743940690049,
+      "grad_norm": 4.723721752248351,
+      "learning_rate": 2.4528301886792453e-06,
+      "loss": 0.0585,
+      "step": 26
+    },
+    {
+      "epoch": 0.007698887938408896,
+      "grad_norm": 2.560432972171629,
+      "learning_rate": 2.547169811320755e-06,
+      "loss": 0.0371,
+      "step": 27
+    },
+    {
+      "epoch": 0.007984031936127744,
+      "grad_norm": 4.2665726201781276,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.0507,
+      "step": 28
+    },
+    {
+      "epoch": 0.008269175933846592,
+      "grad_norm": 4.63734681578259,
+      "learning_rate": 2.7358490566037738e-06,
+      "loss": 0.0511,
+      "step": 29
+    },
+    {
+      "epoch": 0.00855431993156544,
+      "grad_norm": 2.5768861044123095,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.0357,
+      "step": 30
+    },
+    {
+      "epoch": 0.008839463929284289,
+      "grad_norm": 3.634984231769664,
+      "learning_rate": 2.9245283018867924e-06,
+      "loss": 0.0625,
+      "step": 31
+    },
+    {
+      "epoch": 0.009124607927003137,
+      "grad_norm": 4.840413439326456,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.0414,
+      "step": 32
+    },
+    {
+      "epoch": 0.009409751924721984,
+      "grad_norm": 9.287836956870658,
+      "learning_rate": 3.1132075471698115e-06,
+      "loss": 0.0882,
+      "step": 33
+    },
+    {
+      "epoch": 0.009694895922440832,
+      "grad_norm": 6.973138828619164,
+      "learning_rate": 3.207547169811321e-06,
+      "loss": 0.0593,
+      "step": 34
+    },
+    {
+      "epoch": 0.00998003992015968,
+      "grad_norm": 3.6359531137785113,
+      "learning_rate": 3.30188679245283e-06,
+      "loss": 0.0335,
+      "step": 35
+    },
+    {
+      "epoch": 0.010265183917878529,
+      "grad_norm": 3.4832523392273527,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 0.0621,
+      "step": 36
+    },
+    {
+      "epoch": 0.010550327915597377,
+      "grad_norm": 6.152559798979658,
+      "learning_rate": 3.4905660377358493e-06,
+      "loss": 0.0768,
+      "step": 37
+    },
+    {
+      "epoch": 0.010835471913316224,
+      "grad_norm": 3.142333726843115,
+      "learning_rate": 3.5849056603773586e-06,
+      "loss": 0.0421,
+      "step": 38
+    },
+    {
+      "epoch": 0.011120615911035072,
+      "grad_norm": 4.4376556128819615,
+      "learning_rate": 3.679245283018868e-06,
+      "loss": 0.0618,
+      "step": 39
+    },
+    {
+      "epoch": 0.01140575990875392,
+      "grad_norm": 3.119036132535718,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.0395,
+      "step": 40
+    },
+    {
+      "epoch": 0.011690903906472769,
+      "grad_norm": 2.8819748677454564,
+      "learning_rate": 3.8679245283018875e-06,
+      "loss": 0.036,
+      "step": 41
+    },
+    {
+      "epoch": 0.011976047904191617,
+      "grad_norm": 3.0285102030013147,
+      "learning_rate": 3.962264150943396e-06,
+      "loss": 0.0451,
+      "step": 42
+    },
+    {
+      "epoch": 0.012261191901910464,
+      "grad_norm": 2.4198586652875713,
+      "learning_rate": 4.056603773584906e-06,
+      "loss": 0.0359,
+      "step": 43
+    },
+    {
+      "epoch": 0.012546335899629312,
+      "grad_norm": 4.01556344965596,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.0654,
+      "step": 44
+    },
+    {
+      "epoch": 0.01283147989734816,
+      "grad_norm": 4.152755564375582,
+      "learning_rate": 4.245283018867925e-06,
+      "loss": 0.0688,
+      "step": 45
+    },
+    {
+      "epoch": 0.01311662389506701,
+      "grad_norm": 5.650506818749513,
+      "learning_rate": 4.339622641509435e-06,
+      "loss": 0.0621,
+      "step": 46
+    },
+    {
+      "epoch": 0.013401767892785858,
+      "grad_norm": 4.772858263739585,
+      "learning_rate": 4.4339622641509435e-06,
+      "loss": 0.043,
+      "step": 47
+    },
+    {
+      "epoch": 0.013686911890504704,
+      "grad_norm": 3.281056276083866,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 0.0409,
+      "step": 48
+    },
+    {
+      "epoch": 0.013972055888223553,
+      "grad_norm": 2.665340695119781,
+      "learning_rate": 4.622641509433963e-06,
+      "loss": 0.0261,
+      "step": 49
+    },
+    {
+      "epoch": 0.014257199885942401,
+      "grad_norm": 2.1051762190060757,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.046,
+      "step": 50
+    },
+    {
+      "epoch": 0.01454234388366125,
+      "grad_norm": 4.140757787048907,
+      "learning_rate": 4.811320754716982e-06,
+      "loss": 0.0399,
+      "step": 51
+    },
+    {
+      "epoch": 0.014827487881380098,
+      "grad_norm": 4.806224421900785,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.0733,
+      "step": 52
+    },
+    {
+      "epoch": 0.015112631879098944,
+      "grad_norm": 5.244037355061446,
+      "learning_rate": 5e-06,
+      "loss": 0.0829,
+      "step": 53
+    },
+    {
+      "epoch": 0.015397775876817793,
+      "grad_norm": 2.529956827555466,
+      "learning_rate": 5.09433962264151e-06,
+      "loss": 0.0414,
+      "step": 54
+    },
+    {
+      "epoch": 0.01568291987453664,
+      "grad_norm": 3.1547732507353037,
+      "learning_rate": 5.188679245283019e-06,
+      "loss": 0.0492,
+      "step": 55
+    },
+    {
+      "epoch": 0.015968063872255488,
+      "grad_norm": 3.6253749901832553,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 0.0409,
+      "step": 56
+    },
+    {
+      "epoch": 0.016253207869974338,
+      "grad_norm": 8.40170080651301,
+      "learning_rate": 5.377358490566038e-06,
+      "loss": 0.0788,
+      "step": 57
+    },
+    {
+      "epoch": 0.016538351867693184,
+      "grad_norm": 2.2891004276086444,
+      "learning_rate": 5.4716981132075475e-06,
+      "loss": 0.0452,
+      "step": 58
+    },
+    {
+      "epoch": 0.016823495865412035,
+      "grad_norm": 1.968563607614644,
+      "learning_rate": 5.566037735849057e-06,
+      "loss": 0.0374,
+      "step": 59
+    },
+    {
+      "epoch": 0.01710863986313088,
+      "grad_norm": 2.2826667099920424,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.0318,
+      "step": 60
+    },
+    {
+      "epoch": 0.017393783860849728,
+      "grad_norm": 3.075829515844115,
+      "learning_rate": 5.754716981132076e-06,
+      "loss": 0.0456,
+      "step": 61
+    },
+    {
+      "epoch": 0.017678927858568578,
+      "grad_norm": 2.641793301245682,
+      "learning_rate": 5.849056603773585e-06,
+      "loss": 0.0382,
+      "step": 62
+    },
+    {
+      "epoch": 0.017964071856287425,
+      "grad_norm": 3.0357867774553275,
+      "learning_rate": 5.943396226415095e-06,
+      "loss": 0.0391,
+      "step": 63
+    },
+    {
+      "epoch": 0.018249215854006275,
+      "grad_norm": 1.3459497638743567,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.0192,
+      "step": 64
+    },
+    {
+      "epoch": 0.01853435985172512,
+      "grad_norm": 0.8896089027285865,
+      "learning_rate": 6.132075471698113e-06,
+      "loss": 0.026,
+      "step": 65
+    },
+    {
+      "epoch": 0.018819503849443968,
+      "grad_norm": 17.332060998609155,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 0.0781,
+      "step": 66
+    },
+    {
+      "epoch": 0.019104647847162818,
+      "grad_norm": 2.071064848030776,
+      "learning_rate": 6.320754716981132e-06,
+      "loss": 0.0431,
+      "step": 67
+    },
+    {
+      "epoch": 0.019389791844881665,
+      "grad_norm": 4.062997719292194,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.0354,
+      "step": 68
+    },
+    {
+      "epoch": 0.019674935842600515,
+      "grad_norm": 2.350390848534792,
+      "learning_rate": 6.5094339622641515e-06,
+      "loss": 0.0274,
+      "step": 69
+    },
+    {
+      "epoch": 0.01996007984031936,
+      "grad_norm": 2.566245599377888,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.0699,
+      "step": 70
+    },
+    {
+      "epoch": 0.020245223838038208,
+      "grad_norm": 3.5006687012241122,
+      "learning_rate": 6.69811320754717e-06,
+      "loss": 0.0684,
+      "step": 71
+    },
+    {
+      "epoch": 0.020530367835757058,
+      "grad_norm": 4.243644946589001,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.0957,
+      "step": 72
+    },
+    {
+      "epoch": 0.020815511833475905,
+      "grad_norm": 1.5269476114061467,
+      "learning_rate": 6.886792452830189e-06,
+      "loss": 0.0372,
+      "step": 73
+    },
+    {
+      "epoch": 0.021100655831194755,
+      "grad_norm": 41.74516371814436,
+      "learning_rate": 6.981132075471699e-06,
+      "loss": 0.2785,
+      "step": 74
+    },
+    {
+      "epoch": 0.0213857998289136,
+      "grad_norm": 3.8783540833989076,
+      "learning_rate": 7.0754716981132075e-06,
+      "loss": 0.0375,
+      "step": 75
+    },
+    {
+      "epoch": 0.021670943826632448,
+      "grad_norm": 4.643543581746758,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.036,
+      "step": 76
+    },
+    {
+      "epoch": 0.021956087824351298,
+      "grad_norm": 4.600001137811726,
+      "learning_rate": 7.264150943396226e-06,
+      "loss": 0.0636,
+      "step": 77
+    },
+    {
+      "epoch": 0.022241231822070145,
+      "grad_norm": 3.128559332322622,
+      "learning_rate": 7.358490566037736e-06,
+      "loss": 0.0429,
+      "step": 78
+    },
+    {
+      "epoch": 0.022526375819788995,
+      "grad_norm": 2.8641659687703678,
+      "learning_rate": 7.452830188679246e-06,
+      "loss": 0.042,
+      "step": 79
+    },
+    {
+      "epoch": 0.02281151981750784,
+      "grad_norm": 12.271691655074866,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.0862,
+      "step": 80
+    },
+    {
+      "epoch": 0.023096663815226688,
+      "grad_norm": 5.5069227290477665,
+      "learning_rate": 7.641509433962266e-06,
+      "loss": 0.0403,
+      "step": 81
+    },
+    {
+      "epoch": 0.023381807812945538,
+      "grad_norm": 1.4595795029247436,
+      "learning_rate": 7.735849056603775e-06,
+      "loss": 0.0266,
+      "step": 82
+    },
+    {
+      "epoch": 0.023666951810664385,
+      "grad_norm": 8.876102538357202,
+      "learning_rate": 7.830188679245284e-06,
+      "loss": 0.0521,
+      "step": 83
+    },
+    {
+      "epoch": 0.023952095808383235,
+      "grad_norm": 1.6782963289907105,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.0326,
+      "step": 84
+    },
+    {
+      "epoch": 0.02423723980610208,
+      "grad_norm": 2.681755314762461,
+      "learning_rate": 8.018867924528303e-06,
+      "loss": 0.0839,
+      "step": 85
+    },
+    {
+      "epoch": 0.024522383803820928,
+      "grad_norm": 5.714032889612906,
+      "learning_rate": 8.113207547169812e-06,
+      "loss": 0.0316,
+      "step": 86
+    },
+    {
+      "epoch": 0.02480752780153978,
+      "grad_norm": 5.028130886749829,
+      "learning_rate": 8.207547169811321e-06,
+      "loss": 0.0731,
+      "step": 87
+    },
+    {
+      "epoch": 0.025092671799258625,
+      "grad_norm": 3.0570511261289917,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.0333,
+      "step": 88
+    },
+    {
+      "epoch": 0.025377815796977475,
+      "grad_norm": 5.146547952949438,
+      "learning_rate": 8.39622641509434e-06,
+      "loss": 0.0514,
+      "step": 89
+    },
+    {
+      "epoch": 0.02566295979469632,
+      "grad_norm": 4.7531766250387255,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.0527,
+      "step": 90
+    },
+    {
+      "epoch": 0.02594810379241517,
+      "grad_norm": 4.335096216573719,
+      "learning_rate": 8.58490566037736e-06,
+      "loss": 0.0268,
+      "step": 91
+    },
+    {
+      "epoch": 0.02623324779013402,
+      "grad_norm": 2.132428089649692,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 0.0168,
+      "step": 92
+    },
+    {
+      "epoch": 0.026518391787852865,
+      "grad_norm": 6.833117952820334,
+      "learning_rate": 8.773584905660378e-06,
+      "loss": 0.0868,
+      "step": 93
+    },
+    {
+      "epoch": 0.026803535785571715,
+      "grad_norm": 11.160832029176879,
+      "learning_rate": 8.867924528301887e-06,
+      "loss": 0.0925,
+      "step": 94
+    },
+    {
+      "epoch": 0.027088679783290562,
+      "grad_norm": 1.8890461785442043,
+      "learning_rate": 8.962264150943398e-06,
+      "loss": 0.0376,
+      "step": 95
+    },
+    {
+      "epoch": 0.02737382378100941,
+      "grad_norm": 2.0039805311692573,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 0.0534,
+      "step": 96
+    },
+    {
+      "epoch": 0.02765896777872826,
+      "grad_norm": 3.256805619531125,
+      "learning_rate": 9.150943396226416e-06,
+      "loss": 0.0476,
+      "step": 97
+    },
+    {
+      "epoch": 0.027944111776447105,
+      "grad_norm": 2.3004970717202595,
+      "learning_rate": 9.245283018867926e-06,
+      "loss": 0.061,
+      "step": 98
+    },
+    {
+      "epoch": 0.028229255774165955,
+      "grad_norm": 1.8448990035600177,
+      "learning_rate": 9.339622641509435e-06,
+      "loss": 0.0422,
+      "step": 99
+    },
+    {
+      "epoch": 0.028514399771884802,
+      "grad_norm": 4.002283614106026,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.0603,
+      "step": 100
+    },
+    {
+      "epoch": 0.02879954376960365,
+      "grad_norm": 8.180190475095822,
+      "learning_rate": 9.528301886792455e-06,
+      "loss": 0.1227,
+      "step": 101
+    },
+    {
+      "epoch": 0.0290846877673225,
+      "grad_norm": 6.4834435384859015,
+      "learning_rate": 9.622641509433963e-06,
+      "loss": 0.0903,
+      "step": 102
+    },
+    {
+      "epoch": 0.029369831765041345,
+      "grad_norm": 6.419140173971885,
+      "learning_rate": 9.716981132075472e-06,
+      "loss": 0.0973,
+      "step": 103
+    },
+    {
+      "epoch": 0.029654975762760195,
+      "grad_norm": 5.167609682986552,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.0464,
+      "step": 104
+    },
+    {
+      "epoch": 0.029940119760479042,
+      "grad_norm": 2.6286695618463733,
+      "learning_rate": 9.905660377358492e-06,
+      "loss": 0.047,
+      "step": 105
+    },
+    {
+      "epoch": 0.03022526375819789,
+      "grad_norm": 2.385355633734913,
+      "learning_rate": 1e-05,
+      "loss": 0.0366,
+      "step": 106
+    },
+    {
+      "epoch": 0.03051040775591674,
+      "grad_norm": 5.045824475523923,
+      "learning_rate": 9.999997866825128e-06,
+      "loss": 0.0928,
+      "step": 107
+    },
+    {
+      "epoch": 0.030795551753635585,
+      "grad_norm": 5.882161471784182,
+      "learning_rate": 9.999991467302332e-06,
+      "loss": 0.1048,
+      "step": 108
+    },
+    {
+      "epoch": 0.031080695751354435,
+      "grad_norm": 8.71717621944331,
+      "learning_rate": 9.99998080143707e-06,
+      "loss": 0.0885,
+      "step": 109
+    },
+    {
+      "epoch": 0.03136583974907328,
+      "grad_norm": 2.090876100875096,
+      "learning_rate": 9.999965869238445e-06,
+      "loss": 0.0592,
+      "step": 110
+    },
+    {
+      "epoch": 0.03165098374679213,
+      "grad_norm": 15.890424714577145,
+      "learning_rate": 9.999946670719197e-06,
+      "loss": 0.0438,
+      "step": 111
+    },
+    {
+      "epoch": 0.031936127744510975,
+      "grad_norm": 1.4259811755691145,
+      "learning_rate": 9.99992320589571e-06,
+      "loss": 0.0395,
+      "step": 112
+    },
+    {
+      "epoch": 0.03222127174222983,
+      "grad_norm": 1.9327015405417511,
+      "learning_rate": 9.999895474788003e-06,
+      "loss": 0.0371,
+      "step": 113
+    },
+    {
+      "epoch": 0.032506415739948676,
+      "grad_norm": 1.8632732706869015,
+      "learning_rate": 9.999863477419739e-06,
+      "loss": 0.0245,
+      "step": 114
+    },
+    {
+      "epoch": 0.03279155973766752,
+      "grad_norm": 1.8917211984814235,
+      "learning_rate": 9.99982721381822e-06,
+      "loss": 0.0407,
+      "step": 115
+    },
+    {
+      "epoch": 0.03307670373538637,
+      "grad_norm": 2.2247432119892956,
+      "learning_rate": 9.999786684014393e-06,
+      "loss": 0.0334,
+      "step": 116
+    },
+    {
+      "epoch": 0.033361847733105215,
+      "grad_norm": 5.678799374308072,
+      "learning_rate": 9.999741888042832e-06,
+      "loss": 0.0646,
+      "step": 117
+    },
+    {
+      "epoch": 0.03364699173082407,
+      "grad_norm": 2.2979438290144656,
+      "learning_rate": 9.99969282594177e-06,
+      "loss": 0.0352,
+      "step": 118
+    },
+    {
+      "epoch": 0.033932135728542916,
+      "grad_norm": 3.401247202535555,
+      "learning_rate": 9.999639497753062e-06,
+      "loss": 0.0457,
+      "step": 119
+    },
+    {
+      "epoch": 0.03421727972626176,
+      "grad_norm": 1.3220531908758983,
+      "learning_rate": 9.999581903522214e-06,
+      "loss": 0.0148,
+      "step": 120
+    },
+    {
+      "epoch": 0.03450242372398061,
+      "grad_norm": 1.0582272674397575,
+      "learning_rate": 9.999520043298374e-06,
+      "loss": 0.0214,
+      "step": 121
+    },
+    {
+      "epoch": 0.034787567721699456,
+      "grad_norm": 1.5349811431314222,
+      "learning_rate": 9.99945391713432e-06,
+      "loss": 0.0237,
+      "step": 122
+    },
+    {
+      "epoch": 0.03507271171941831,
+      "grad_norm": 5.061670772936218,
+      "learning_rate": 9.999383525086478e-06,
+      "loss": 0.0393,
+      "step": 123
+    },
+    {
+      "epoch": 0.035357855717137156,
+      "grad_norm": 3.8169991926514313,
+      "learning_rate": 9.999308867214908e-06,
+      "loss": 0.0622,
+      "step": 124
+    },
+    {
+      "epoch": 0.035642999714856,
+      "grad_norm": 33.60130740332581,
+      "learning_rate": 9.999229943583318e-06,
+      "loss": 0.0982,
+      "step": 125
+    },
+    {
+      "epoch": 0.03592814371257485,
+      "grad_norm": 1.7047090809257088,
+      "learning_rate": 9.999146754259048e-06,
+      "loss": 0.0321,
+      "step": 126
+    },
+    {
+      "epoch": 0.036213287710293696,
+      "grad_norm": 2.9744010851446103,
+      "learning_rate": 9.999059299313082e-06,
+      "loss": 0.0207,
+      "step": 127
+    },
+    {
+      "epoch": 0.03649843170801255,
+      "grad_norm": 4.240730548981665,
+      "learning_rate": 9.998967578820042e-06,
+      "loss": 0.0289,
+      "step": 128
+    },
+    {
+      "epoch": 0.036783575705731396,
+      "grad_norm": 3.3812717287163463,
+      "learning_rate": 9.998871592858193e-06,
+      "loss": 0.0495,
+      "step": 129
+    },
+    {
+      "epoch": 0.03706871970345024,
+      "grad_norm": 1.8193189165651151,
+      "learning_rate": 9.998771341509434e-06,
+      "loss": 0.0667,
+      "step": 130
+    },
+    {
+      "epoch": 0.03735386370116909,
+      "grad_norm": 6.060011032912712,
+      "learning_rate": 9.998666824859307e-06,
+      "loss": 0.1307,
+      "step": 131
+    },
+    {
+      "epoch": 0.037639007698887936,
+      "grad_norm": 2.402353091095481,
+      "learning_rate": 9.998558042996993e-06,
+      "loss": 0.0736,
+      "step": 132
+    },
+    {
+      "epoch": 0.03792415169660679,
+      "grad_norm": 5.839543926949511,
+      "learning_rate": 9.998444996015314e-06,
+      "loss": 0.0945,
+      "step": 133
+    },
+    {
+      "epoch": 0.038209295694325636,
+      "grad_norm": 1.4473227758832712,
+      "learning_rate": 9.998327684010727e-06,
+      "loss": 0.0432,
+      "step": 134
+    },
+    {
+      "epoch": 0.03849443969204448,
+      "grad_norm": 2.396507073747093,
+      "learning_rate": 9.998206107083333e-06,
+      "loss": 0.0679,
+      "step": 135
+    },
+    {
+      "epoch": 0.03877958368976333,
+      "grad_norm": 2.6450829119213073,
+      "learning_rate": 9.998080265336867e-06,
+      "loss": 0.0206,
+      "step": 136
+    },
+    {
+      "epoch": 0.039064727687482176,
+      "grad_norm": 8.44539739688482,
+      "learning_rate": 9.997950158878712e-06,
+      "loss": 0.0362,
+      "step": 137
+    },
+    {
+      "epoch": 0.03934987168520103,
+      "grad_norm": 3.940799243397122,
+      "learning_rate": 9.997815787819876e-06,
+      "loss": 0.0452,
+      "step": 138
+    },
+    {
+      "epoch": 0.039635015682919876,
+      "grad_norm": 1.2577768949582078,
+      "learning_rate": 9.997677152275019e-06,
+      "loss": 0.029,
+      "step": 139
+    },
+    {
+      "epoch": 0.03992015968063872,
+      "grad_norm": 2.172531523049211,
+      "learning_rate": 9.997534252362432e-06,
+      "loss": 0.0289,
+      "step": 140
+    },
+    {
+      "epoch": 0.04020530367835757,
+      "grad_norm": 1.7163630652895192,
+      "learning_rate": 9.99738708820405e-06,
+      "loss": 0.0374,
+      "step": 141
+    },
+    {
+      "epoch": 0.040490447676076416,
+      "grad_norm": 1.9849481929945414,
+      "learning_rate": 9.99723565992544e-06,
+      "loss": 0.0385,
+      "step": 142
+    },
+    {
+      "epoch": 0.04077559167379527,
+      "grad_norm": 1.0957767923455073,
+      "learning_rate": 9.997079967655816e-06,
+      "loss": 0.0345,
+      "step": 143
+    },
+    {
+      "epoch": 0.041060735671514116,
+      "grad_norm": 1.845401941137247,
+      "learning_rate": 9.996920011528022e-06,
+      "loss": 0.0298,
+      "step": 144
+    },
+    {
+      "epoch": 0.04134587966923296,
+      "grad_norm": 1.709458777909131,
+      "learning_rate": 9.996755791678544e-06,
+      "loss": 0.023,
+      "step": 145
+    },
+    {
+      "epoch": 0.04163102366695181,
+      "grad_norm": 2.416877485324215,
+      "learning_rate": 9.996587308247507e-06,
+      "loss": 0.041,
+      "step": 146
+    },
+    {
+      "epoch": 0.041916167664670656,
+      "grad_norm": 1.8466041607808619,
+      "learning_rate": 9.996414561378671e-06,
+      "loss": 0.0308,
+      "step": 147
+    },
+    {
+      "epoch": 0.04220131166238951,
+      "grad_norm": 2.7453406578055746,
+      "learning_rate": 9.996237551219439e-06,
+      "loss": 0.0805,
+      "step": 148
+    },
+    {
+      "epoch": 0.042486455660108356,
+      "grad_norm": 1.3012783290944823,
+      "learning_rate": 9.996056277920845e-06,
+      "loss": 0.0459,
+      "step": 149
+    },
+    {
+      "epoch": 0.0427715996578272,
+      "grad_norm": 1.6845595049974698,
+      "learning_rate": 9.995870741637566e-06,
+      "loss": 0.0206,
+      "step": 150
+    },
+    {
+      "epoch": 0.04305674365554605,
+      "grad_norm": 1.9673407080566452,
+      "learning_rate": 9.995680942527915e-06,
+      "loss": 0.0277,
+      "step": 151
+    },
+    {
+      "epoch": 0.043341887653264896,
+      "grad_norm": 1.096649599808759,
+      "learning_rate": 9.99548688075384e-06,
+      "loss": 0.0104,
+      "step": 152
+    },
+    {
+      "epoch": 0.04362703165098375,
+      "grad_norm": 4.340215678141815,
+      "learning_rate": 9.99528855648093e-06,
+      "loss": 0.0782,
+      "step": 153
+    },
+    {
+      "epoch": 0.043912175648702596,
+      "grad_norm": 0.962700831894613,
+      "learning_rate": 9.995085969878408e-06,
+      "loss": 0.0084,
+      "step": 154
+    },
+    {
+      "epoch": 0.04419731964642144,
+      "grad_norm": 1.3431951362099261,
+      "learning_rate": 9.994879121119134e-06,
+      "loss": 0.0347,
+      "step": 155
+    },
+    {
+      "epoch": 0.04448246364414029,
+      "grad_norm": 2.396590839420942,
+      "learning_rate": 9.99466801037961e-06,
+      "loss": 0.0518,
+      "step": 156
+    },
+    {
+      "epoch": 0.044767607641859136,
+      "grad_norm": 1.87825274755581,
+      "learning_rate": 9.994452637839964e-06,
+      "loss": 0.0316,
+      "step": 157
+    },
+    {
+      "epoch": 0.04505275163957799,
+      "grad_norm": 2.6008704806074707,
+      "learning_rate": 9.994233003683972e-06,
+      "loss": 0.0338,
+      "step": 158
+    },
+    {
+      "epoch": 0.045337895637296836,
+      "grad_norm": 3.2329437736534476,
+      "learning_rate": 9.994009108099038e-06,
+      "loss": 0.0821,
+      "step": 159
+    },
+    {
+      "epoch": 0.04562303963501568,
+      "grad_norm": 1.59183218675036,
+      "learning_rate": 9.99378095127621e-06,
+      "loss": 0.0245,
+      "step": 160
+    },
+    {
+      "epoch": 0.04590818363273453,
+      "grad_norm": 1.6404471636929772,
+      "learning_rate": 9.993548533410162e-06,
+      "loss": 0.0188,
+      "step": 161
+    },
+    {
+      "epoch": 0.046193327630453376,
+      "grad_norm": 5.826108372985289,
+      "learning_rate": 9.993311854699214e-06,
+      "loss": 0.0963,
+      "step": 162
+    },
+    {
+      "epoch": 0.04647847162817223,
+      "grad_norm": 3.8052953300535393,
+      "learning_rate": 9.993070915345313e-06,
+      "loss": 0.063,
+      "step": 163
+    },
+    {
+      "epoch": 0.046763615625891077,
+      "grad_norm": 3.166036914377814,
+      "learning_rate": 9.992825715554047e-06,
+      "loss": 0.0391,
+      "step": 164
+    },
+    {
+      "epoch": 0.04704875962360992,
+      "grad_norm": 0.9851882069270612,
+      "learning_rate": 9.992576255534637e-06,
+      "loss": 0.0412,
+      "step": 165
+    },
+    {
+      "epoch": 0.04733390362132877,
+      "grad_norm": 1.9312831934515782,
+      "learning_rate": 9.99232253549994e-06,
+      "loss": 0.0201,
+      "step": 166
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.902848957710233,
+      "learning_rate": 9.992064555666448e-06,
+      "loss": 0.0235,
+      "step": 167
+    },
+    {
+      "epoch": 0.04790419161676647,
+      "grad_norm": 2.4508039029457254,
+      "learning_rate": 9.991802316254286e-06,
+      "loss": 0.0236,
+      "step": 168
+    },
+    {
+      "epoch": 0.04818933561448532,
+      "grad_norm": 2.052929331756409,
+      "learning_rate": 9.991535817487218e-06,
+      "loss": 0.038,
+      "step": 169
+    },
+    {
+      "epoch": 0.04847447961220416,
+      "grad_norm": 1.480706200852706,
+      "learning_rate": 9.991265059592638e-06,
+      "loss": 0.0556,
+      "step": 170
+    },
+    {
+      "epoch": 0.04875962360992301,
+      "grad_norm": 4.557244503271385,
+      "learning_rate": 9.990990042801573e-06,
+      "loss": 0.0669,
+      "step": 171
+    },
+    {
+      "epoch": 0.049044767607641856,
+      "grad_norm": 2.087362034480544,
+      "learning_rate": 9.990710767348692e-06,
+      "loss": 0.0531,
+      "step": 172
+    },
+    {
+      "epoch": 0.04932991160536071,
+      "grad_norm": 1.290096043611282,
+      "learning_rate": 9.990427233472286e-06,
+      "loss": 0.0347,
+      "step": 173
+    },
+    {
+      "epoch": 0.04961505560307956,
+      "grad_norm": 4.300260420077906,
+      "learning_rate": 9.990139441414291e-06,
+      "loss": 0.0686,
+      "step": 174
+    },
+    {
+      "epoch": 0.0499001996007984,
+      "grad_norm": 2.3277629405503455,
+      "learning_rate": 9.989847391420268e-06,
+      "loss": 0.0355,
+      "step": 175
+    },
+    {
+      "epoch": 0.05018534359851725,
+      "grad_norm": 1.0758597876794864,
+      "learning_rate": 9.989551083739416e-06,
+      "loss": 0.0379,
+      "step": 176
+    },
+    {
+      "epoch": 0.0504704875962361,
+      "grad_norm": 2.296501614131635,
+      "learning_rate": 9.989250518624566e-06,
+      "loss": 0.0463,
+      "step": 177
+    },
+    {
+      "epoch": 0.05075563159395495,
+      "grad_norm": 1.3724194734047541,
+      "learning_rate": 9.98894569633218e-06,
+      "loss": 0.0314,
+      "step": 178
+    },
+    {
+      "epoch": 0.0510407755916738,
+      "grad_norm": 1.7362884773572984,
+      "learning_rate": 9.988636617122354e-06,
+      "loss": 0.0585,
+      "step": 179
+    },
+    {
+      "epoch": 0.05132591958939264,
+      "grad_norm": 0.7782751167457802,
+      "learning_rate": 9.988323281258817e-06,
+      "loss": 0.0309,
+      "step": 180
+    },
+    {
+      "epoch": 0.05161106358711149,
+      "grad_norm": 2.6395856764774583,
+      "learning_rate": 9.988005689008926e-06,
+      "loss": 0.0502,
+      "step": 181
+    },
+    {
+      "epoch": 0.05189620758483034,
+      "grad_norm": 2.978324666472321,
+      "learning_rate": 9.987683840643679e-06,
+      "loss": 0.0429,
+      "step": 182
+    },
+    {
+      "epoch": 0.05218135158254919,
+      "grad_norm": 3.4011230134034167,
+      "learning_rate": 9.987357736437691e-06,
+      "loss": 0.0637,
+      "step": 183
+    },
+    {
+      "epoch": 0.05246649558026804,
+      "grad_norm": 1.941822329126047,
+      "learning_rate": 9.987027376669224e-06,
+      "loss": 0.0315,
+      "step": 184
+    },
+    {
+      "epoch": 0.052751639577986884,
+      "grad_norm": 2.8549897117814567,
+      "learning_rate": 9.986692761620163e-06,
+      "loss": 0.0382,
+      "step": 185
+    },
+    {
+      "epoch": 0.05303678357570573,
+      "grad_norm": 2.582295889648147,
+      "learning_rate": 9.986353891576021e-06,
+      "loss": 0.0385,
+      "step": 186
+    },
+    {
+      "epoch": 0.05332192757342458,
+      "grad_norm": 2.471870150657678,
+      "learning_rate": 9.98601076682595e-06,
+      "loss": 0.0322,
+      "step": 187
+    },
+    {
+      "epoch": 0.05360707157114343,
+      "grad_norm": 2.6685644561553006,
+      "learning_rate": 9.985663387662726e-06,
+      "loss": 0.0603,
+      "step": 188
+    },
+    {
+      "epoch": 0.05389221556886228,
+      "grad_norm": 2.4073947494866186,
+      "learning_rate": 9.985311754382758e-06,
+      "loss": 0.0362,
+      "step": 189
+    },
+    {
+      "epoch": 0.054177359566581124,
+      "grad_norm": 2.4836029774438666,
+      "learning_rate": 9.984955867286083e-06,
+      "loss": 0.0384,
+      "step": 190
+    },
+    {
+      "epoch": 0.05446250356429997,
+      "grad_norm": 2.5496037219151018,
+      "learning_rate": 9.98459572667637e-06,
+      "loss": 0.055,
+      "step": 191
+    },
+    {
+      "epoch": 0.05474764756201882,
+      "grad_norm": 2.2290862127290585,
+      "learning_rate": 9.984231332860914e-06,
+      "loss": 0.03,
+      "step": 192
+    },
+    {
+      "epoch": 0.05503279155973767,
+      "grad_norm": 2.0539664937082893,
+      "learning_rate": 9.983862686150644e-06,
+      "loss": 0.0486,
+      "step": 193
+    },
+    {
+      "epoch": 0.05531793555745652,
+      "grad_norm": 1.347135991088387,
+      "learning_rate": 9.983489786860115e-06,
+      "loss": 0.0335,
+      "step": 194
+    },
+    {
+      "epoch": 0.055603079555175364,
+      "grad_norm": 1.152598221757865,
+      "learning_rate": 9.983112635307508e-06,
+      "loss": 0.0417,
+      "step": 195
+    },
+    {
+      "epoch": 0.05588822355289421,
+      "grad_norm": 1.1394323676423384,
+      "learning_rate": 9.982731231814637e-06,
+      "loss": 0.04,
+      "step": 196
+    },
+    {
+      "epoch": 0.05617336755061306,
+      "grad_norm": 0.5751858615186785,
+      "learning_rate": 9.982345576706942e-06,
+      "loss": 0.0257,
+      "step": 197
+    },
+    {
+      "epoch": 0.05645851154833191,
+      "grad_norm": 2.228495954241724,
+      "learning_rate": 9.981955670313491e-06,
+      "loss": 0.0244,
+      "step": 198
+    },
+    {
+      "epoch": 0.05674365554605076,
+      "grad_norm": 2.1167156058410566,
+      "learning_rate": 9.98156151296698e-06,
+      "loss": 0.0255,
+      "step": 199
+    },
+    {
+      "epoch": 0.057028799543769604,
+      "grad_norm": 1.1499351454121258,
+      "learning_rate": 9.981163105003731e-06,
+      "loss": 0.0399,
+      "step": 200
+    },
+    {
+      "epoch": 0.05731394354148845,
+      "grad_norm": 1.0863620605603235,
+      "learning_rate": 9.980760446763693e-06,
+      "loss": 0.0336,
+      "step": 201
+    },
+    {
+      "epoch": 0.0575990875392073,
+      "grad_norm": 1.8021997995153731,
+      "learning_rate": 9.980353538590441e-06,
+      "loss": 0.049,
+      "step": 202
+    },
+    {
+      "epoch": 0.05788423153692615,
+      "grad_norm": 0.8060673496091526,
+      "learning_rate": 9.97994238083118e-06,
+      "loss": 0.02,
+      "step": 203
+    },
+    {
+      "epoch": 0.058169375534645,
+      "grad_norm": 2.8431892970960226,
+      "learning_rate": 9.97952697383674e-06,
+      "loss": 0.0459,
+      "step": 204
+    },
+    {
+      "epoch": 0.058454519532363844,
+      "grad_norm": 2.2266671109715475,
+      "learning_rate": 9.979107317961572e-06,
+      "loss": 0.0379,
+      "step": 205
+    },
+    {
+      "epoch": 0.05873966353008269,
+      "grad_norm": 1.4259804628148744,
+      "learning_rate": 9.978683413563755e-06,
+      "loss": 0.0346,
+      "step": 206
+    },
+    {
+      "epoch": 0.05902480752780154,
+      "grad_norm": 1.9019710115897548,
+      "learning_rate": 9.978255261004996e-06,
+      "loss": 0.0321,
+      "step": 207
+    },
+    {
+      "epoch": 0.05930995152552039,
+      "grad_norm": 3.2763885427824433,
+      "learning_rate": 9.977822860650626e-06,
+      "loss": 0.0327,
+      "step": 208
+    },
+    {
+      "epoch": 0.05959509552323924,
+      "grad_norm": 4.085746329437377,
+      "learning_rate": 9.977386212869597e-06,
+      "loss": 0.0785,
+      "step": 209
+    },
+    {
+      "epoch": 0.059880239520958084,
+      "grad_norm": 1.4461803192099967,
+      "learning_rate": 9.976945318034487e-06,
+      "loss": 0.0372,
+      "step": 210
+    },
+    {
+      "epoch": 0.06016538351867693,
+      "grad_norm": 1.6159101224346348,
+      "learning_rate": 9.9765001765215e-06,
+      "loss": 0.0432,
+      "step": 211
+    },
+    {
+      "epoch": 0.06045052751639578,
+      "grad_norm": 2.109127926712865,
+      "learning_rate": 9.976050788710462e-06,
+      "loss": 0.0474,
+      "step": 212
+    },
+    {
+      "epoch": 0.06073567151411463,
+      "grad_norm": 1.6172712792447292,
+      "learning_rate": 9.97559715498482e-06,
+      "loss": 0.0299,
+      "step": 213
+    },
+    {
+      "epoch": 0.06102081551183348,
+      "grad_norm": 1.1823346555015843,
+      "learning_rate": 9.975139275731649e-06,
+      "loss": 0.0224,
+      "step": 214
+    },
+    {
+      "epoch": 0.061305959509552324,
+      "grad_norm": 1.2000963098670507,
+      "learning_rate": 9.97467715134164e-06,
+      "loss": 0.0362,
+      "step": 215
+    },
+    {
+      "epoch": 0.06159110350727117,
+      "grad_norm": 1.6613781060710882,
+      "learning_rate": 9.974210782209113e-06,
+      "loss": 0.0382,
+      "step": 216
+    },
+    {
+      "epoch": 0.06187624750499002,
+      "grad_norm": 2.2297827161797406,
+      "learning_rate": 9.973740168732006e-06,
+      "loss": 0.0407,
+      "step": 217
+    },
+    {
+      "epoch": 0.06216139150270887,
+      "grad_norm": 3.0544112277949713,
+      "learning_rate": 9.973265311311877e-06,
+      "loss": 0.0534,
+      "step": 218
+    },
+    {
+      "epoch": 0.06244653550042772,
+      "grad_norm": 1.8606402993779663,
+      "learning_rate": 9.972786210353913e-06,
+      "loss": 0.031,
+      "step": 219
+    },
+    {
+      "epoch": 0.06273167949814656,
+      "grad_norm": 1.9835976433808742,
+      "learning_rate": 9.97230286626691e-06,
+      "loss": 0.0512,
+      "step": 220
+    },
+    {
+      "epoch": 0.06301682349586542,
+      "grad_norm": 2.4671374293137474,
+      "learning_rate": 9.971815279463294e-06,
+      "loss": 0.0261,
+      "step": 221
+    },
+    {
+      "epoch": 0.06330196749358426,
+      "grad_norm": 1.9426433678441557,
+      "learning_rate": 9.971323450359109e-06,
+      "loss": 0.0354,
+      "step": 222
+    },
+    {
+      "epoch": 0.06358711149130311,
+      "grad_norm": 2.833304865793606,
+      "learning_rate": 9.970827379374016e-06,
+      "loss": 0.0658,
+      "step": 223
+    },
+    {
+      "epoch": 0.06387225548902195,
+      "grad_norm": 1.364065742400369,
+      "learning_rate": 9.9703270669313e-06,
+      "loss": 0.023,
+      "step": 224
+    },
+    {
+      "epoch": 0.0641573994867408,
+      "grad_norm": 2.0598784617586783,
+      "learning_rate": 9.96982251345786e-06,
+      "loss": 0.0303,
+      "step": 225
+    },
+    {
+      "epoch": 0.06444254348445966,
+      "grad_norm": 1.6800528252103155,
+      "learning_rate": 9.969313719384217e-06,
+      "loss": 0.0232,
+      "step": 226
+    },
+    {
+      "epoch": 0.0647276874821785,
+      "grad_norm": 2.5347983635463067,
+      "learning_rate": 9.96880068514451e-06,
+      "loss": 0.051,
+      "step": 227
+    },
+    {
+      "epoch": 0.06501283147989735,
+      "grad_norm": 4.198478034633089,
+      "learning_rate": 9.968283411176499e-06,
+      "loss": 0.0637,
+      "step": 228
+    },
+    {
+      "epoch": 0.06529797547761619,
+      "grad_norm": 2.4608303176324013,
+      "learning_rate": 9.967761897921553e-06,
+      "loss": 0.0356,
+      "step": 229
+    },
+    {
+      "epoch": 0.06558311947533504,
+      "grad_norm": 1.1830386086249287,
+      "learning_rate": 9.967236145824666e-06,
+      "loss": 0.0225,
+      "step": 230
+    },
+    {
+      "epoch": 0.0658682634730539,
+      "grad_norm": 2.335905650803093,
+      "learning_rate": 9.966706155334445e-06,
+      "loss": 0.0484,
+      "step": 231
+    },
+    {
+      "epoch": 0.06615340747077274,
+      "grad_norm": 0.7365538318490851,
+      "learning_rate": 9.966171926903116e-06,
+      "loss": 0.0242,
+      "step": 232
+    },
+    {
+      "epoch": 0.06643855146849159,
+      "grad_norm": 0.8332673431312235,
+      "learning_rate": 9.965633460986521e-06,
+      "loss": 0.0143,
+      "step": 233
+    },
+    {
+      "epoch": 0.06672369546621043,
+      "grad_norm": 1.1115342044179635,
+      "learning_rate": 9.965090758044116e-06,
+      "loss": 0.0315,
+      "step": 234
+    },
+    {
+      "epoch": 0.06700883946392928,
+      "grad_norm": 2.828412875905856,
+      "learning_rate": 9.964543818538974e-06,
+      "loss": 0.0724,
+      "step": 235
+    },
+    {
+      "epoch": 0.06729398346164814,
+      "grad_norm": 1.0745851388736904,
+      "learning_rate": 9.963992642937782e-06,
+      "loss": 0.0155,
+      "step": 236
+    },
+    {
+      "epoch": 0.06757912745936698,
+      "grad_norm": 1.9642520660921505,
+      "learning_rate": 9.963437231710838e-06,
+      "loss": 0.0296,
+      "step": 237
+    },
+    {
+      "epoch": 0.06786427145708583,
+      "grad_norm": 0.6051981185401907,
+      "learning_rate": 9.962877585332062e-06,
+      "loss": 0.0177,
+      "step": 238
+    },
+    {
+      "epoch": 0.06814941545480467,
+      "grad_norm": 4.122674586469668,
+      "learning_rate": 9.962313704278981e-06,
+      "loss": 0.0523,
+      "step": 239
+    },
+    {
+      "epoch": 0.06843455945252352,
+      "grad_norm": 1.6897654144243972,
+      "learning_rate": 9.96174558903274e-06,
+      "loss": 0.0132,
+      "step": 240
+    },
+    {
+      "epoch": 0.06871970345024238,
+      "grad_norm": 3.1253283435389196,
+      "learning_rate": 9.961173240078092e-06,
+      "loss": 0.0321,
+      "step": 241
+    },
+    {
+      "epoch": 0.06900484744796122,
+      "grad_norm": 1.3965076576538442,
+      "learning_rate": 9.960596657903407e-06,
+      "loss": 0.029,
+      "step": 242
+    },
+    {
+      "epoch": 0.06928999144568007,
+      "grad_norm": 1.9660002863377228,
+      "learning_rate": 9.960015843000666e-06,
+      "loss": 0.0522,
+      "step": 243
+    },
+    {
+      "epoch": 0.06957513544339891,
+      "grad_norm": 1.3233952783403555,
+      "learning_rate": 9.959430795865457e-06,
+      "loss": 0.0353,
+      "step": 244
+    },
+    {
+      "epoch": 0.06986027944111776,
+      "grad_norm": 1.7033322453474304,
+      "learning_rate": 9.958841516996989e-06,
+      "loss": 0.0223,
+      "step": 245
+    },
+    {
+      "epoch": 0.07014542343883662,
+      "grad_norm": 2.5317243818879764,
+      "learning_rate": 9.95824800689807e-06,
+      "loss": 0.0598,
+      "step": 246
+    },
+    {
+      "epoch": 0.07043056743655546,
+      "grad_norm": 2.69732116281974,
+      "learning_rate": 9.957650266075129e-06,
+      "loss": 0.0419,
+      "step": 247
+    },
+    {
+      "epoch": 0.07071571143427431,
+      "grad_norm": 1.4150311093853096,
+      "learning_rate": 9.957048295038197e-06,
+      "loss": 0.0472,
+      "step": 248
+    },
+    {
+      "epoch": 0.07100085543199315,
+      "grad_norm": 1.2224281181933392,
+      "learning_rate": 9.95644209430092e-06,
+      "loss": 0.0389,
+      "step": 249
+    },
+    {
+      "epoch": 0.071285999429712,
+      "grad_norm": 51.194853358315605,
+      "learning_rate": 9.955831664380548e-06,
+      "loss": 0.0451,
+      "step": 250
+    },
+    {
+      "epoch": 0.07157114342743086,
+      "grad_norm": 1.46254711704071,
+      "learning_rate": 9.955217005797946e-06,
+      "loss": 0.0276,
+      "step": 251
+    },
+    {
+      "epoch": 0.0718562874251497,
+      "grad_norm": 3.4636231015416974,
+      "learning_rate": 9.954598119077583e-06,
+      "loss": 0.0396,
+      "step": 252
+    },
+    {
+      "epoch": 0.07214143142286855,
+      "grad_norm": 0.8518861435588755,
+      "learning_rate": 9.953975004747535e-06,
+      "loss": 0.0313,
+      "step": 253
+    },
+    {
+      "epoch": 0.07242657542058739,
+      "grad_norm": 1.4742031582432285,
+      "learning_rate": 9.953347663339487e-06,
+      "loss": 0.0397,
+      "step": 254
+    },
+    {
+      "epoch": 0.07271171941830624,
+      "grad_norm": 0.8505976022407974,
+      "learning_rate": 9.95271609538873e-06,
+      "loss": 0.0259,
+      "step": 255
+    },
+    {
+      "epoch": 0.0729968634160251,
+      "grad_norm": 1.3056007814387238,
+      "learning_rate": 9.952080301434165e-06,
+      "loss": 0.0109,
+      "step": 256
+    },
+    {
+      "epoch": 0.07328200741374394,
+      "grad_norm": 2.734719524535765,
+      "learning_rate": 9.951440282018294e-06,
+      "loss": 0.0424,
+      "step": 257
+    },
+    {
+      "epoch": 0.07356715141146279,
+      "grad_norm": 0.915900286106093,
+      "learning_rate": 9.950796037687224e-06,
+      "loss": 0.0279,
+      "step": 258
+    },
+    {
+      "epoch": 0.07385229540918163,
+      "grad_norm": 1.1576885381186377,
+      "learning_rate": 9.950147568990672e-06,
+      "loss": 0.0291,
+      "step": 259
+    },
+    {
+      "epoch": 0.07413743940690048,
+      "grad_norm": 1.1396907116761987,
+      "learning_rate": 9.949494876481957e-06,
+      "loss": 0.0573,
+      "step": 260
+    },
+    {
+      "epoch": 0.07442258340461934,
+      "grad_norm": 0.8954392948824229,
+      "learning_rate": 9.948837960718001e-06,
+      "loss": 0.0109,
+      "step": 261
+    },
+    {
+      "epoch": 0.07470772740233818,
+      "grad_norm": 2.2187692519372955,
+      "learning_rate": 9.94817682225933e-06,
+      "loss": 0.0405,
+      "step": 262
+    },
+    {
+      "epoch": 0.07499287140005703,
+      "grad_norm": 1.1502737187174694,
+      "learning_rate": 9.947511461670076e-06,
+      "loss": 0.0194,
+      "step": 263
+    },
+    {
+      "epoch": 0.07527801539777587,
+      "grad_norm": 1.061028206812728,
+      "learning_rate": 9.946841879517968e-06,
+      "loss": 0.0419,
+      "step": 264
+    },
+    {
+      "epoch": 0.07556315939549473,
+      "grad_norm": 1.3832069086922913,
+      "learning_rate": 9.94616807637434e-06,
+      "loss": 0.015,
+      "step": 265
+    },
+    {
+      "epoch": 0.07584830339321358,
+      "grad_norm": 0.6459735906834944,
+      "learning_rate": 9.945490052814133e-06,
+      "loss": 0.0255,
+      "step": 266
+    },
+    {
+      "epoch": 0.07613344739093242,
+      "grad_norm": 1.1494098773455543,
+      "learning_rate": 9.94480780941588e-06,
+      "loss": 0.0346,
+      "step": 267
+    },
+    {
+      "epoch": 0.07641859138865127,
+      "grad_norm": 0.9289181620653825,
+      "learning_rate": 9.944121346761718e-06,
+      "loss": 0.014,
+      "step": 268
+    },
+    {
+      "epoch": 0.07670373538637011,
+      "grad_norm": 1.2203907189231082,
+      "learning_rate": 9.943430665437388e-06,
+      "loss": 0.0167,
+      "step": 269
+    },
+    {
+      "epoch": 0.07698887938408897,
+      "grad_norm": 2.620259167299579,
+      "learning_rate": 9.942735766032228e-06,
+      "loss": 0.044,
+      "step": 270
+    },
+    {
+      "epoch": 0.07727402338180782,
+      "grad_norm": 1.7206246319986442,
+      "learning_rate": 9.94203664913917e-06,
+      "loss": 0.0235,
+      "step": 271
+    },
+    {
+      "epoch": 0.07755916737952666,
+      "grad_norm": 1.4642402716762304,
+      "learning_rate": 9.941333315354755e-06,
+      "loss": 0.0258,
+      "step": 272
+    },
+    {
+      "epoch": 0.07784431137724551,
+      "grad_norm": 1.7553631989445817,
+      "learning_rate": 9.940625765279112e-06,
+      "loss": 0.0345,
+      "step": 273
+    },
+    {
+      "epoch": 0.07812945537496435,
+      "grad_norm": 1.6716431070541198,
+      "learning_rate": 9.939913999515976e-06,
+      "loss": 0.0454,
+      "step": 274
+    },
+    {
+      "epoch": 0.0784145993726832,
+      "grad_norm": 1.9526398951956696,
+      "learning_rate": 9.939198018672671e-06,
+      "loss": 0.0382,
+      "step": 275
+    },
+    {
+      "epoch": 0.07869974337040206,
+      "grad_norm": 1.2954788594239341,
+      "learning_rate": 9.938477823360127e-06,
+      "loss": 0.0309,
+      "step": 276
+    },
+    {
+      "epoch": 0.0789848873681209,
+      "grad_norm": 4.4338813137709385,
+      "learning_rate": 9.937753414192862e-06,
+      "loss": 0.072,
+      "step": 277
+    },
+    {
+      "epoch": 0.07927003136583975,
+      "grad_norm": 1.56094475712123,
+      "learning_rate": 9.937024791788991e-06,
+      "loss": 0.0517,
+      "step": 278
+    },
+    {
+      "epoch": 0.07955517536355859,
+      "grad_norm": 1.7375848109236927,
+      "learning_rate": 9.93629195677023e-06,
+      "loss": 0.0345,
+      "step": 279
+    },
+    {
+      "epoch": 0.07984031936127745,
+      "grad_norm": 2.649452584682016,
+      "learning_rate": 9.935554909761882e-06,
+      "loss": 0.0594,
+      "step": 280
+    },
+    {
+      "epoch": 0.0801254633589963,
+      "grad_norm": 1.6499457117477292,
+      "learning_rate": 9.93481365139285e-06,
+      "loss": 0.0376,
+      "step": 281
+    },
+    {
+      "epoch": 0.08041060735671514,
+      "grad_norm": 1.1619489575147839,
+      "learning_rate": 9.934068182295622e-06,
+      "loss": 0.04,
+      "step": 282
+    },
+    {
+      "epoch": 0.08069575135443399,
+      "grad_norm": 1.7247725696145613,
+      "learning_rate": 9.933318503106291e-06,
+      "loss": 0.0323,
+      "step": 283
+    },
+    {
+      "epoch": 0.08098089535215283,
+      "grad_norm": 2.1577985956712666,
+      "learning_rate": 9.93256461446453e-06,
+      "loss": 0.0526,
+      "step": 284
+    },
+    {
+      "epoch": 0.08126603934987169,
+      "grad_norm": 0.8867124373033844,
+      "learning_rate": 9.931806517013612e-06,
+      "loss": 0.0144,
+      "step": 285
+    },
+    {
+      "epoch": 0.08155118334759054,
+      "grad_norm": 0.9443505140091544,
+      "learning_rate": 9.9310442114004e-06,
+      "loss": 0.0233,
+      "step": 286
+    },
+    {
+      "epoch": 0.08183632734530938,
+      "grad_norm": 0.8757124141344964,
+      "learning_rate": 9.930277698275347e-06,
+      "loss": 0.0235,
+      "step": 287
+    },
+    {
+      "epoch": 0.08212147134302823,
+      "grad_norm": 2.2248864075955805,
+      "learning_rate": 9.92950697829249e-06,
+      "loss": 0.0495,
+      "step": 288
+    },
+    {
+      "epoch": 0.08240661534074707,
+      "grad_norm": 1.794320095937349,
+      "learning_rate": 9.928732052109466e-06,
+      "loss": 0.0575,
+      "step": 289
+    },
+    {
+      "epoch": 0.08269175933846593,
+      "grad_norm": 0.5271974530133274,
+      "learning_rate": 9.927952920387497e-06,
+      "loss": 0.0174,
+      "step": 290
+    },
+    {
+      "epoch": 0.08297690333618478,
+      "grad_norm": 2.130600355467149,
+      "learning_rate": 9.92716958379139e-06,
+      "loss": 0.0269,
+      "step": 291
+    },
+    {
+      "epoch": 0.08326204733390362,
+      "grad_norm": 1.6078717603842572,
+      "learning_rate": 9.926382042989544e-06,
+      "loss": 0.0381,
+      "step": 292
+    },
+    {
+      "epoch": 0.08354719133162247,
+      "grad_norm": 1.067101418411665,
+      "learning_rate": 9.925590298653942e-06,
+      "loss": 0.0416,
+      "step": 293
+    },
+    {
+      "epoch": 0.08383233532934131,
+      "grad_norm": 1.8888059772289436,
+      "learning_rate": 9.924794351460159e-06,
+      "loss": 0.0431,
+      "step": 294
+    },
+    {
+      "epoch": 0.08411747932706017,
+      "grad_norm": 1.1366438944954262,
+      "learning_rate": 9.92399420208735e-06,
+      "loss": 0.034,
+      "step": 295
+    },
+    {
+      "epoch": 0.08440262332477902,
+      "grad_norm": 66.83013849628288,
+      "learning_rate": 9.923189851218259e-06,
+      "loss": 0.303,
+      "step": 296
+    },
+    {
+      "epoch": 0.08468776732249786,
+      "grad_norm": 2.2777335388392643,
+      "learning_rate": 9.922381299539214e-06,
+      "loss": 0.0426,
+      "step": 297
+    },
+    {
+      "epoch": 0.08497291132021671,
+      "grad_norm": 1.3777833351571647,
+      "learning_rate": 9.921568547740131e-06,
+      "loss": 0.0551,
+      "step": 298
+    },
+    {
+      "epoch": 0.08525805531793555,
+      "grad_norm": 0.6388835255348255,
+      "learning_rate": 9.920751596514502e-06,
+      "loss": 0.0228,
+      "step": 299
+    },
+    {
+      "epoch": 0.0855431993156544,
+      "grad_norm": 2.278298167977118,
+      "learning_rate": 9.919930446559412e-06,
+      "loss": 0.0349,
+      "step": 300
+    },
+    {
+      "epoch": 0.08582834331337326,
+      "grad_norm": 0.873059407793711,
+      "learning_rate": 9.91910509857552e-06,
+      "loss": 0.0212,
+      "step": 301
+    },
+    {
+      "epoch": 0.0861134873110921,
+      "grad_norm": 2.2132859296620038,
+      "learning_rate": 9.918275553267069e-06,
+      "loss": 0.0439,
+      "step": 302
+    },
+    {
+      "epoch": 0.08639863130881095,
+      "grad_norm": 2.615477056500074,
+      "learning_rate": 9.917441811341887e-06,
+      "loss": 0.0299,
+      "step": 303
+    },
+    {
+      "epoch": 0.08668377530652979,
+      "grad_norm": 1.0557781544658622,
+      "learning_rate": 9.916603873511386e-06,
+      "loss": 0.0158,
+      "step": 304
+    },
+    {
+      "epoch": 0.08696891930424865,
+      "grad_norm": 3.9400306558554754,
+      "learning_rate": 9.915761740490545e-06,
+      "loss": 0.0444,
+      "step": 305
+    },
+    {
+      "epoch": 0.0872540633019675,
+      "grad_norm": 3.1606798637531544,
+      "learning_rate": 9.914915412997937e-06,
+      "loss": 0.035,
+      "step": 306
+    },
+    {
+      "epoch": 0.08753920729968634,
+      "grad_norm": 2.2379796936077754,
+      "learning_rate": 9.914064891755703e-06,
+      "loss": 0.0353,
+      "step": 307
+    },
+    {
+      "epoch": 0.08782435129740519,
+      "grad_norm": 2.0924722936161007,
+      "learning_rate": 9.91321017748957e-06,
+      "loss": 0.0466,
+      "step": 308
+    },
+    {
+      "epoch": 0.08810949529512403,
+      "grad_norm": 1.1933158739643803,
+      "learning_rate": 9.91235127092884e-06,
+      "loss": 0.0237,
+      "step": 309
+    },
+    {
+      "epoch": 0.08839463929284289,
+      "grad_norm": 1.450838486064019,
+      "learning_rate": 9.911488172806392e-06,
+      "loss": 0.0296,
+      "step": 310
+    },
+    {
+      "epoch": 0.08867978329056174,
+      "grad_norm": 2.8659368821250095,
+      "learning_rate": 9.91062088385868e-06,
+      "loss": 0.0644,
+      "step": 311
+    },
+    {
+      "epoch": 0.08896492728828058,
+      "grad_norm": 2.3883067154612823,
+      "learning_rate": 9.909749404825736e-06,
+      "loss": 0.0558,
+      "step": 312
+    },
+    {
+      "epoch": 0.08925007128599943,
+      "grad_norm": 1.338846186227719,
+      "learning_rate": 9.90887373645117e-06,
+      "loss": 0.0142,
+      "step": 313
+    },
+    {
+      "epoch": 0.08953521528371827,
+      "grad_norm": 2.2418774559520105,
+      "learning_rate": 9.907993879482161e-06,
+      "loss": 0.0362,
+      "step": 314
+    },
+    {
+      "epoch": 0.08982035928143713,
+      "grad_norm": 1.5944639604254414,
+      "learning_rate": 9.907109834669465e-06,
+      "loss": 0.0277,
+      "step": 315
+    },
+    {
+      "epoch": 0.09010550327915598,
+      "grad_norm": 1.8710437857203388,
+      "learning_rate": 9.90622160276741e-06,
+      "loss": 0.0235,
+      "step": 316
+    },
+    {
+      "epoch": 0.09039064727687482,
+      "grad_norm": 1.4117522494101777,
+      "learning_rate": 9.905329184533897e-06,
+      "loss": 0.0328,
+      "step": 317
+    },
+    {
+      "epoch": 0.09067579127459367,
+      "grad_norm": 1.2111494335027815,
+      "learning_rate": 9.904432580730404e-06,
+      "loss": 0.0445,
+      "step": 318
+    },
+    {
+      "epoch": 0.09096093527231251,
+      "grad_norm": 1.0923531873837928,
+      "learning_rate": 9.90353179212197e-06,
+      "loss": 0.0363,
+      "step": 319
+    },
+    {
+      "epoch": 0.09124607927003137,
+      "grad_norm": 0.9108263761049934,
+      "learning_rate": 9.902626819477214e-06,
+      "loss": 0.0233,
+      "step": 320
+    },
+    {
+      "epoch": 0.09153122326775022,
+      "grad_norm": 0.9052416100597438,
+      "learning_rate": 9.901717663568323e-06,
+      "loss": 0.0221,
+      "step": 321
+    },
+    {
+      "epoch": 0.09181636726546906,
+      "grad_norm": 1.5417885755673133,
+      "learning_rate": 9.900804325171052e-06,
+      "loss": 0.0357,
+      "step": 322
+    },
+    {
+      "epoch": 0.09210151126318791,
+      "grad_norm": 1.1483931076981324,
+      "learning_rate": 9.899886805064723e-06,
+      "loss": 0.042,
+      "step": 323
+    },
+    {
+      "epoch": 0.09238665526090675,
+      "grad_norm": 0.8264374345543277,
+      "learning_rate": 9.89896510403223e-06,
+      "loss": 0.0135,
+      "step": 324
+    },
+    {
+      "epoch": 0.0926717992586256,
+      "grad_norm": 3.028712934159954,
+      "learning_rate": 9.898039222860032e-06,
+      "loss": 0.096,
+      "step": 325
+    },
+    {
+      "epoch": 0.09295694325634446,
+      "grad_norm": 3.908443382779185,
+      "learning_rate": 9.897109162338157e-06,
+      "loss": 0.0678,
+      "step": 326
+    },
+    {
+      "epoch": 0.0932420872540633,
+      "grad_norm": 1.4504578551621745,
+      "learning_rate": 9.896174923260198e-06,
+      "loss": 0.028,
+      "step": 327
+    },
+    {
+      "epoch": 0.09352723125178215,
+      "grad_norm": 1.0918597452992596,
+      "learning_rate": 9.89523650642331e-06,
+      "loss": 0.0352,
+      "step": 328
+    },
+    {
+      "epoch": 0.09381237524950099,
+      "grad_norm": 1.3624134945746003,
+      "learning_rate": 9.89429391262822e-06,
+      "loss": 0.034,
+      "step": 329
+    },
+    {
+      "epoch": 0.09409751924721985,
+      "grad_norm": 1.9386090064569985,
+      "learning_rate": 9.893347142679211e-06,
+      "loss": 0.0213,
+      "step": 330
+    },
+    {
+      "epoch": 0.0943826632449387,
+      "grad_norm": 2.442430474342043,
+      "learning_rate": 9.892396197384135e-06,
+      "loss": 0.0274,
+      "step": 331
+    },
+    {
+      "epoch": 0.09466780724265754,
+      "grad_norm": 2.7994856234060372,
+      "learning_rate": 9.891441077554405e-06,
+      "loss": 0.0496,
+      "step": 332
+    },
+    {
+      "epoch": 0.0949529512403764,
+      "grad_norm": 1.0457890691940244,
+      "learning_rate": 9.890481784004998e-06,
+      "loss": 0.0197,
+      "step": 333
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 6.1017012731170865,
+      "learning_rate": 9.889518317554446e-06,
+      "loss": 0.047,
+      "step": 334
+    },
+    {
+      "epoch": 0.09552323923581409,
+      "grad_norm": 2.090726587494363,
+      "learning_rate": 9.88855067902485e-06,
+      "loss": 0.0326,
+      "step": 335
+    },
+    {
+      "epoch": 0.09580838323353294,
+      "grad_norm": 0.699893313587921,
+      "learning_rate": 9.887578869241866e-06,
+      "loss": 0.0164,
+      "step": 336
+    },
+    {
+      "epoch": 0.09609352723125178,
+      "grad_norm": 3.507183876219138,
+      "learning_rate": 9.886602889034709e-06,
+      "loss": 0.0416,
+      "step": 337
+    },
+    {
+      "epoch": 0.09637867122897063,
+      "grad_norm": 0.5771253616906034,
+      "learning_rate": 9.885622739236154e-06,
+      "loss": 0.0161,
+      "step": 338
+    },
+    {
+      "epoch": 0.09666381522668947,
+      "grad_norm": 1.231417111947857,
+      "learning_rate": 9.884638420682534e-06,
+      "loss": 0.0269,
+      "step": 339
+    },
+    {
+      "epoch": 0.09694895922440833,
+      "grad_norm": 1.7586310033147416,
+      "learning_rate": 9.883649934213738e-06,
+      "loss": 0.0303,
+      "step": 340
+    },
+    {
+      "epoch": 0.09723410322212718,
+      "grad_norm": 1.6088417745313677,
+      "learning_rate": 9.882657280673212e-06,
+      "loss": 0.0277,
+      "step": 341
+    },
+    {
+      "epoch": 0.09751924721984602,
+      "grad_norm": 1.8001784084893846,
+      "learning_rate": 9.881660460907957e-06,
+      "loss": 0.0232,
+      "step": 342
+    },
+    {
+      "epoch": 0.09780439121756487,
+      "grad_norm": 3.549553604790123,
+      "learning_rate": 9.880659475768526e-06,
+      "loss": 0.1015,
+      "step": 343
+    },
+    {
+      "epoch": 0.09808953521528371,
+      "grad_norm": 1.0138498172858956,
+      "learning_rate": 9.879654326109037e-06,
+      "loss": 0.0219,
+      "step": 344
+    },
+    {
+      "epoch": 0.09837467921300257,
+      "grad_norm": 0.5962872642185872,
+      "learning_rate": 9.878645012787149e-06,
+      "loss": 0.0073,
+      "step": 345
+    },
+    {
+      "epoch": 0.09865982321072142,
+      "grad_norm": 1.5344142515023333,
+      "learning_rate": 9.87763153666408e-06,
+      "loss": 0.0188,
+      "step": 346
+    },
+    {
+      "epoch": 0.09894496720844026,
+      "grad_norm": 2.6171999253216014,
+      "learning_rate": 9.8766138986046e-06,
+      "loss": 0.0448,
+      "step": 347
+    },
+    {
+      "epoch": 0.09923011120615911,
+      "grad_norm": 2.0090501521581885,
+      "learning_rate": 9.875592099477025e-06,
+      "loss": 0.0481,
+      "step": 348
+    },
+    {
+      "epoch": 0.09951525520387795,
+      "grad_norm": 2.3354612758931386,
+      "learning_rate": 9.874566140153228e-06,
+      "loss": 0.0417,
+      "step": 349
+    },
+    {
+      "epoch": 0.0998003992015968,
+      "grad_norm": 1.5781598957515535,
+      "learning_rate": 9.87353602150863e-06,
+      "loss": 0.0188,
+      "step": 350
+    },
+    {
+      "epoch": 0.10008554319931566,
+      "grad_norm": 1.928184575603502,
+      "learning_rate": 9.8725017444222e-06,
+      "loss": 0.0234,
+      "step": 351
+    },
+    {
+      "epoch": 0.1003706871970345,
+      "grad_norm": 1.0876133323980701,
+      "learning_rate": 9.871463309776455e-06,
+      "loss": 0.009,
+      "step": 352
+    },
+    {
+      "epoch": 0.10065583119475335,
+      "grad_norm": 11.324227273778364,
+      "learning_rate": 9.870420718457458e-06,
+      "loss": 0.011,
+      "step": 353
+    },
+    {
+      "epoch": 0.1009409751924722,
+      "grad_norm": 2.3272703667045516,
+      "learning_rate": 9.869373971354826e-06,
+      "loss": 0.0427,
+      "step": 354
+    },
+    {
+      "epoch": 0.10122611919019105,
+      "grad_norm": 1.2649728160919362,
+      "learning_rate": 9.868323069361712e-06,
+      "loss": 0.0259,
+      "step": 355
+    },
+    {
+      "epoch": 0.1015112631879099,
+      "grad_norm": 2.0664662137667467,
+      "learning_rate": 9.867268013374822e-06,
+      "loss": 0.0226,
+      "step": 356
+    },
+    {
+      "epoch": 0.10179640718562874,
+      "grad_norm": 0.9365874475796875,
+      "learning_rate": 9.866208804294401e-06,
+      "loss": 0.0107,
+      "step": 357
+    },
+    {
+      "epoch": 0.1020815511833476,
+      "grad_norm": 0.5822681312488291,
+      "learning_rate": 9.865145443024243e-06,
+      "loss": 0.0073,
+      "step": 358
+    },
+    {
+      "epoch": 0.10236669518106643,
+      "grad_norm": 0.5773090002383701,
+      "learning_rate": 9.86407793047168e-06,
+      "loss": 0.0064,
+      "step": 359
+    },
+    {
+      "epoch": 0.10265183917878529,
+      "grad_norm": 1.298097020704675,
+      "learning_rate": 9.863006267547591e-06,
+      "loss": 0.0293,
+      "step": 360
+    },
+    {
+      "epoch": 0.10293698317650414,
+      "grad_norm": 2.01105743317194,
+      "learning_rate": 9.861930455166392e-06,
+      "loss": 0.0423,
+      "step": 361
+    },
+    {
+      "epoch": 0.10322212717422298,
+      "grad_norm": 3.854044900737908,
+      "learning_rate": 9.86085049424604e-06,
+      "loss": 0.0646,
+      "step": 362
+    },
+    {
+      "epoch": 0.10350727117194183,
+      "grad_norm": 3.1428521017863393,
+      "learning_rate": 9.859766385708035e-06,
+      "loss": 0.0603,
+      "step": 363
+    },
+    {
+      "epoch": 0.10379241516966067,
+      "grad_norm": 1.1415144666210106,
+      "learning_rate": 9.858678130477415e-06,
+      "loss": 0.0089,
+      "step": 364
+    },
+    {
+      "epoch": 0.10407755916737953,
+      "grad_norm": 1.4442748658078282,
+      "learning_rate": 9.857585729482753e-06,
+      "loss": 0.0271,
+      "step": 365
+    },
+    {
+      "epoch": 0.10436270316509838,
+      "grad_norm": 0.8289767742462578,
+      "learning_rate": 9.856489183656163e-06,
+      "loss": 0.0159,
+      "step": 366
+    },
+    {
+      "epoch": 0.10464784716281722,
+      "grad_norm": 1.7124474237003318,
+      "learning_rate": 9.855388493933298e-06,
+      "loss": 0.0527,
+      "step": 367
+    },
+    {
+      "epoch": 0.10493299116053607,
+      "grad_norm": 2.981703921126902,
+      "learning_rate": 9.854283661253338e-06,
+      "loss": 0.0357,
+      "step": 368
+    },
+    {
+      "epoch": 0.10521813515825491,
+      "grad_norm": 2.1585606753904965,
+      "learning_rate": 9.853174686559006e-06,
+      "loss": 0.0393,
+      "step": 369
+    },
+    {
+      "epoch": 0.10550327915597377,
+      "grad_norm": 1.5348910023395037,
+      "learning_rate": 9.852061570796557e-06,
+      "loss": 0.0398,
+      "step": 370
+    },
+    {
+      "epoch": 0.10578842315369262,
+      "grad_norm": 1.535951948243033,
+      "learning_rate": 9.85094431491578e-06,
+      "loss": 0.0208,
+      "step": 371
+    },
+    {
+      "epoch": 0.10607356715141146,
+      "grad_norm": 1.0394143122475457,
+      "learning_rate": 9.849822919869993e-06,
+      "loss": 0.0127,
+      "step": 372
+    },
+    {
+      "epoch": 0.10635871114913031,
+      "grad_norm": 1.8866287812876101,
+      "learning_rate": 9.848697386616052e-06,
+      "loss": 0.0243,
+      "step": 373
+    },
+    {
+      "epoch": 0.10664385514684915,
+      "grad_norm": 2.152613819596744,
+      "learning_rate": 9.847567716114339e-06,
+      "loss": 0.0475,
+      "step": 374
+    },
+    {
+      "epoch": 0.10692899914456801,
+      "grad_norm": 1.8499798292082466,
+      "learning_rate": 9.846433909328768e-06,
+      "loss": 0.0482,
+      "step": 375
+    },
+    {
+      "epoch": 0.10721414314228686,
+      "grad_norm": 2.1403198834038255,
+      "learning_rate": 9.845295967226782e-06,
+      "loss": 0.0359,
+      "step": 376
+    },
+    {
+      "epoch": 0.1074992871400057,
+      "grad_norm": 0.679326626935762,
+      "learning_rate": 9.844153890779352e-06,
+      "loss": 0.014,
+      "step": 377
+    },
+    {
+      "epoch": 0.10778443113772455,
+      "grad_norm": 2.4149761091081277,
+      "learning_rate": 9.84300768096098e-06,
+      "loss": 0.035,
+      "step": 378
+    },
+    {
+      "epoch": 0.1080695751354434,
+      "grad_norm": 1.1906254157269311,
+      "learning_rate": 9.841857338749693e-06,
+      "loss": 0.0321,
+      "step": 379
+    },
+    {
+      "epoch": 0.10835471913316225,
+      "grad_norm": 0.5918164701274412,
+      "learning_rate": 9.840702865127039e-06,
+      "loss": 0.0201,
+      "step": 380
+    },
+    {
+      "epoch": 0.1086398631308811,
+      "grad_norm": 1.7497628829871825,
+      "learning_rate": 9.839544261078099e-06,
+      "loss": 0.0211,
+      "step": 381
+    },
+    {
+      "epoch": 0.10892500712859994,
+      "grad_norm": 1.3983788980140046,
+      "learning_rate": 9.838381527591475e-06,
+      "loss": 0.0236,
+      "step": 382
+    },
+    {
+      "epoch": 0.1092101511263188,
+      "grad_norm": 3.1851074927752623,
+      "learning_rate": 9.83721466565929e-06,
+      "loss": 0.0372,
+      "step": 383
+    },
+    {
+      "epoch": 0.10949529512403763,
+      "grad_norm": 1.4171014306009728,
+      "learning_rate": 9.836043676277195e-06,
+      "loss": 0.0252,
+      "step": 384
+    },
+    {
+      "epoch": 0.10978043912175649,
+      "grad_norm": 0.9261846959930898,
+      "learning_rate": 9.83486856044436e-06,
+      "loss": 0.0191,
+      "step": 385
+    },
+    {
+      "epoch": 0.11006558311947534,
+      "grad_norm": 1.6296524588712908,
+      "learning_rate": 9.833689319163473e-06,
+      "loss": 0.0576,
+      "step": 386
+    },
+    {
+      "epoch": 0.11035072711719418,
+      "grad_norm": 0.8019311699722914,
+      "learning_rate": 9.832505953440748e-06,
+      "loss": 0.0202,
+      "step": 387
+    },
+    {
+      "epoch": 0.11063587111491303,
+      "grad_norm": 1.5113990773927573,
+      "learning_rate": 9.831318464285914e-06,
+      "loss": 0.046,
+      "step": 388
+    },
+    {
+      "epoch": 0.11092101511263187,
+      "grad_norm": 2.728229672787799,
+      "learning_rate": 9.83012685271222e-06,
+      "loss": 0.0537,
+      "step": 389
+    },
+    {
+      "epoch": 0.11120615911035073,
+      "grad_norm": 0.913884511065414,
+      "learning_rate": 9.828931119736435e-06,
+      "loss": 0.0102,
+      "step": 390
+    },
+    {
+      "epoch": 0.11149130310806958,
+      "grad_norm": 4.01446357656375,
+      "learning_rate": 9.827731266378839e-06,
+      "loss": 0.061,
+      "step": 391
+    },
+    {
+      "epoch": 0.11177644710578842,
+      "grad_norm": 1.3370546467948305,
+      "learning_rate": 9.82652729366323e-06,
+      "loss": 0.0371,
+      "step": 392
+    },
+    {
+      "epoch": 0.11206159110350727,
+      "grad_norm": 1.3728857509837673,
+      "learning_rate": 9.825319202616926e-06,
+      "loss": 0.0258,
+      "step": 393
+    },
+    {
+      "epoch": 0.11234673510122611,
+      "grad_norm": 1.8638416100609516,
+      "learning_rate": 9.82410699427075e-06,
+      "loss": 0.0386,
+      "step": 394
+    },
+    {
+      "epoch": 0.11263187909894497,
+      "grad_norm": 0.7022867943321849,
+      "learning_rate": 9.822890669659044e-06,
+      "loss": 0.0144,
+      "step": 395
+    },
+    {
+      "epoch": 0.11291702309666382,
+      "grad_norm": 1.4752375376718947,
+      "learning_rate": 9.821670229819663e-06,
+      "loss": 0.049,
+      "step": 396
+    },
+    {
+      "epoch": 0.11320216709438266,
+      "grad_norm": 0.7023676759664046,
+      "learning_rate": 9.820445675793973e-06,
+      "loss": 0.0241,
+      "step": 397
+    },
+    {
+      "epoch": 0.11348731109210151,
+      "grad_norm": 0.49880724393845305,
+      "learning_rate": 9.819217008626847e-06,
+      "loss": 0.0241,
+      "step": 398
+    },
+    {
+      "epoch": 0.11377245508982035,
+      "grad_norm": 2.064257083055477,
+      "learning_rate": 9.817984229366669e-06,
+      "loss": 0.058,
+      "step": 399
+    },
+    {
+      "epoch": 0.11405759908753921,
+      "grad_norm": 0.5760537754446529,
+      "learning_rate": 9.816747339065333e-06,
+      "loss": 0.0233,
+      "step": 400
+    },
+    {
+      "epoch": 0.11434274308525806,
+      "grad_norm": 1.2124787291362593,
+      "learning_rate": 9.81550633877824e-06,
+      "loss": 0.0614,
+      "step": 401
+    },
+    {
+      "epoch": 0.1146278870829769,
+      "grad_norm": 1.5895231814003101,
+      "learning_rate": 9.8142612295643e-06,
+      "loss": 0.0569,
+      "step": 402
+    },
+    {
+      "epoch": 0.11491303108069575,
+      "grad_norm": 1.0484784015774489,
+      "learning_rate": 9.813012012485925e-06,
+      "loss": 0.0161,
+      "step": 403
+    },
+    {
+      "epoch": 0.1151981750784146,
+      "grad_norm": 2.0015024924134988,
+      "learning_rate": 9.811758688609036e-06,
+      "loss": 0.0288,
+      "step": 404
+    },
+    {
+      "epoch": 0.11548331907613345,
+      "grad_norm": 1.0183121982064807,
+      "learning_rate": 9.810501259003058e-06,
+      "loss": 0.0272,
+      "step": 405
+    },
+    {
+      "epoch": 0.1157684630738523,
+      "grad_norm": 0.7040323962320703,
+      "learning_rate": 9.809239724740913e-06,
+      "loss": 0.0127,
+      "step": 406
+    },
+    {
+      "epoch": 0.11605360707157114,
+      "grad_norm": 1.2738847744090185,
+      "learning_rate": 9.807974086899037e-06,
+      "loss": 0.0326,
+      "step": 407
+    },
+    {
+      "epoch": 0.11633875106929,
+      "grad_norm": 1.6928730590425576,
+      "learning_rate": 9.806704346557354e-06,
+      "loss": 0.0467,
+      "step": 408
+    },
+    {
+      "epoch": 0.11662389506700883,
+      "grad_norm": 1.5237688876565934,
+      "learning_rate": 9.8054305047993e-06,
+      "loss": 0.0428,
+      "step": 409
+    },
+    {
+      "epoch": 0.11690903906472769,
+      "grad_norm": 2.5601247962488047,
+      "learning_rate": 9.804152562711804e-06,
+      "loss": 0.0312,
+      "step": 410
+    },
+    {
+      "epoch": 0.11719418306244654,
+      "grad_norm": 1.6898452734577185,
+      "learning_rate": 9.802870521385295e-06,
+      "loss": 0.0165,
+      "step": 411
+    },
+    {
+      "epoch": 0.11747932706016538,
+      "grad_norm": 3.202435062260781,
+      "learning_rate": 9.801584381913702e-06,
+      "loss": 0.0365,
+      "step": 412
+    },
+    {
+      "epoch": 0.11776447105788423,
+      "grad_norm": 2.4208311464375267,
+      "learning_rate": 9.800294145394449e-06,
+      "loss": 0.0456,
+      "step": 413
+    },
+    {
+      "epoch": 0.11804961505560307,
+      "grad_norm": 1.8464478511128473,
+      "learning_rate": 9.798999812928454e-06,
+      "loss": 0.0275,
+      "step": 414
+    },
+    {
+      "epoch": 0.11833475905332193,
+      "grad_norm": 1.7466714628297064,
+      "learning_rate": 9.797701385620135e-06,
+      "loss": 0.0664,
+      "step": 415
+    },
+    {
+      "epoch": 0.11861990305104078,
+      "grad_norm": 1.374398177525167,
+      "learning_rate": 9.796398864577398e-06,
+      "loss": 0.0249,
+      "step": 416
+    },
+    {
+      "epoch": 0.11890504704875962,
+      "grad_norm": 0.682379469399641,
+      "learning_rate": 9.795092250911646e-06,
+      "loss": 0.0324,
+      "step": 417
+    },
+    {
+      "epoch": 0.11919019104647847,
+      "grad_norm": 1.345099432009825,
+      "learning_rate": 9.793781545737775e-06,
+      "loss": 0.0304,
+      "step": 418
+    },
+    {
+      "epoch": 0.11947533504419731,
+      "grad_norm": 2.459208618588478,
+      "learning_rate": 9.79246675017417e-06,
+      "loss": 0.038,
+      "step": 419
+    },
+    {
+      "epoch": 0.11976047904191617,
+      "grad_norm": 1.1582497240978238,
+      "learning_rate": 9.791147865342703e-06,
+      "loss": 0.0239,
+      "step": 420
+    },
+    {
+      "epoch": 0.12004562303963502,
+      "grad_norm": 0.6873012288522815,
+      "learning_rate": 9.789824892368742e-06,
+      "loss": 0.0219,
+      "step": 421
+    },
+    {
+      "epoch": 0.12033076703735386,
+      "grad_norm": 1.4503792824156856,
+      "learning_rate": 9.78849783238114e-06,
+      "loss": 0.039,
+      "step": 422
+    },
+    {
+      "epoch": 0.12061591103507271,
+      "grad_norm": 1.4695551960581217,
+      "learning_rate": 9.787166686512237e-06,
+      "loss": 0.0225,
+      "step": 423
+    },
+    {
+      "epoch": 0.12090105503279155,
+      "grad_norm": 1.0755209767496783,
+      "learning_rate": 9.785831455897859e-06,
+      "loss": 0.0187,
+      "step": 424
+    },
+    {
+      "epoch": 0.12118619903051041,
+      "grad_norm": 2.8435317285492223,
+      "learning_rate": 9.784492141677318e-06,
+      "loss": 0.0377,
+      "step": 425
+    },
+    {
+      "epoch": 0.12147134302822926,
+      "grad_norm": 0.6194176889222343,
+      "learning_rate": 9.783148744993413e-06,
+      "loss": 0.014,
+      "step": 426
+    },
+    {
+      "epoch": 0.1217564870259481,
+      "grad_norm": 1.197556570522883,
+      "learning_rate": 9.781801266992421e-06,
+      "loss": 0.0144,
+      "step": 427
+    },
+    {
+      "epoch": 0.12204163102366695,
+      "grad_norm": 1.509780270356931,
+      "learning_rate": 9.780449708824107e-06,
+      "loss": 0.0149,
+      "step": 428
+    },
+    {
+      "epoch": 0.1223267750213858,
+      "grad_norm": 1.0425109169214848,
+      "learning_rate": 9.779094071641712e-06,
+      "loss": 0.0287,
+      "step": 429
+    },
+    {
+      "epoch": 0.12261191901910465,
+      "grad_norm": 2.9496669420120383,
+      "learning_rate": 9.777734356601964e-06,
+      "loss": 0.0152,
+      "step": 430
+    },
+    {
+      "epoch": 0.1228970630168235,
+      "grad_norm": 1.4441615240576289,
+      "learning_rate": 9.776370564865066e-06,
+      "loss": 0.0265,
+      "step": 431
+    },
+    {
+      "epoch": 0.12318220701454234,
+      "grad_norm": 2.8617609288687773,
+      "learning_rate": 9.775002697594696e-06,
+      "loss": 0.0259,
+      "step": 432
+    },
+    {
+      "epoch": 0.1234673510122612,
+      "grad_norm": 2.019269634607327,
+      "learning_rate": 9.773630755958021e-06,
+      "loss": 0.0419,
+      "step": 433
+    },
+    {
+      "epoch": 0.12375249500998003,
+      "grad_norm": 3.762297631517556,
+      "learning_rate": 9.772254741125672e-06,
+      "loss": 0.0328,
+      "step": 434
+    },
+    {
+      "epoch": 0.12403763900769889,
+      "grad_norm": 1.9417548664384192,
+      "learning_rate": 9.770874654271768e-06,
+      "loss": 0.0572,
+      "step": 435
+    },
+    {
+      "epoch": 0.12432278300541774,
+      "grad_norm": 1.7851777881726447,
+      "learning_rate": 9.769490496573886e-06,
+      "loss": 0.049,
+      "step": 436
+    },
+    {
+      "epoch": 0.12460792700313658,
+      "grad_norm": 2.477982800391376,
+      "learning_rate": 9.768102269213093e-06,
+      "loss": 0.0705,
+      "step": 437
+    },
+    {
+      "epoch": 0.12489307100085544,
+      "grad_norm": 0.3032358120092834,
+      "learning_rate": 9.76670997337392e-06,
+      "loss": 0.0047,
+      "step": 438
+    },
+    {
+      "epoch": 0.1251782149985743,
+      "grad_norm": 1.483642571394467,
+      "learning_rate": 9.765313610244372e-06,
+      "loss": 0.0151,
+      "step": 439
+    },
+    {
+      "epoch": 0.12546335899629313,
+      "grad_norm": 1.814367508425431,
+      "learning_rate": 9.763913181015923e-06,
+      "loss": 0.0459,
+      "step": 440
+    },
+    {
+      "epoch": 0.12574850299401197,
+      "grad_norm": 0.7952597090870602,
+      "learning_rate": 9.762508686883515e-06,
+      "loss": 0.0201,
+      "step": 441
+    },
+    {
+      "epoch": 0.12603364699173084,
+      "grad_norm": 3.185384390855697,
+      "learning_rate": 9.761100129045565e-06,
+      "loss": 0.0432,
+      "step": 442
+    },
+    {
+      "epoch": 0.12631879098944968,
+      "grad_norm": 1.0661696789911552,
+      "learning_rate": 9.759687508703948e-06,
+      "loss": 0.0139,
+      "step": 443
+    },
+    {
+      "epoch": 0.12660393498716851,
+      "grad_norm": 3.7725165197415964,
+      "learning_rate": 9.758270827064016e-06,
+      "loss": 0.0483,
+      "step": 444
+    },
+    {
+      "epoch": 0.12688907898488735,
+      "grad_norm": 2.746968288043148,
+      "learning_rate": 9.756850085334576e-06,
+      "loss": 0.0574,
+      "step": 445
+    },
+    {
+      "epoch": 0.12717422298260622,
+      "grad_norm": 0.926770487815625,
+      "learning_rate": 9.755425284727908e-06,
+      "loss": 0.018,
+      "step": 446
+    },
+    {
+      "epoch": 0.12745936698032506,
+      "grad_norm": 0.9082707028939784,
+      "learning_rate": 9.753996426459748e-06,
+      "loss": 0.0263,
+      "step": 447
+    },
+    {
+      "epoch": 0.1277445109780439,
+      "grad_norm": 1.164510189232619,
+      "learning_rate": 9.752563511749301e-06,
+      "loss": 0.0337,
+      "step": 448
+    },
+    {
+      "epoch": 0.12802965497576277,
+      "grad_norm": 2.3418912893557704,
+      "learning_rate": 9.75112654181923e-06,
+      "loss": 0.0416,
+      "step": 449
+    },
+    {
+      "epoch": 0.1283147989734816,
+      "grad_norm": 4.056148524420801,
+      "learning_rate": 9.749685517895654e-06,
+      "loss": 0.0539,
+      "step": 450
+    },
+    {
+      "epoch": 0.12859994297120045,
+      "grad_norm": 1.3521227970804115,
+      "learning_rate": 9.748240441208158e-06,
+      "loss": 0.0395,
+      "step": 451
+    },
+    {
+      "epoch": 0.12888508696891932,
+      "grad_norm": 1.9419704386778984,
+      "learning_rate": 9.746791312989785e-06,
+      "loss": 0.0393,
+      "step": 452
+    },
+    {
+      "epoch": 0.12917023096663816,
+      "grad_norm": 1.0595383076344087,
+      "learning_rate": 9.745338134477031e-06,
+      "loss": 0.0214,
+      "step": 453
+    },
+    {
+      "epoch": 0.129455374964357,
+      "grad_norm": 1.4679175053059728,
+      "learning_rate": 9.743880906909849e-06,
+      "loss": 0.0525,
+      "step": 454
+    },
+    {
+      "epoch": 0.12974051896207583,
+      "grad_norm": 1.4310123455755626,
+      "learning_rate": 9.742419631531647e-06,
+      "loss": 0.0158,
+      "step": 455
+    },
+    {
+      "epoch": 0.1300256629597947,
+      "grad_norm": 0.6588067935353689,
+      "learning_rate": 9.740954309589288e-06,
+      "loss": 0.0277,
+      "step": 456
+    },
+    {
+      "epoch": 0.13031080695751354,
+      "grad_norm": 0.9397591945163838,
+      "learning_rate": 9.739484942333087e-06,
+      "loss": 0.0312,
+      "step": 457
+    },
+    {
+      "epoch": 0.13059595095523238,
+      "grad_norm": 1.1729515696135353,
+      "learning_rate": 9.738011531016809e-06,
+      "loss": 0.0369,
+      "step": 458
+    },
+    {
+      "epoch": 0.13088109495295125,
+      "grad_norm": 2.9106198831897734,
+      "learning_rate": 9.736534076897676e-06,
+      "loss": 0.0698,
+      "step": 459
+    },
+    {
+      "epoch": 0.1311662389506701,
+      "grad_norm": 2.418685278178657,
+      "learning_rate": 9.735052581236353e-06,
+      "loss": 0.0529,
+      "step": 460
+    },
+    {
+      "epoch": 0.13145138294838893,
+      "grad_norm": 1.1991557936515589,
+      "learning_rate": 9.733567045296955e-06,
+      "loss": 0.019,
+      "step": 461
+    },
+    {
+      "epoch": 0.1317365269461078,
+      "grad_norm": 2.168170383166244,
+      "learning_rate": 9.732077470347043e-06,
+      "loss": 0.0336,
+      "step": 462
+    },
+    {
+      "epoch": 0.13202167094382664,
+      "grad_norm": 0.8382241005884289,
+      "learning_rate": 9.730583857657632e-06,
+      "loss": 0.0241,
+      "step": 463
+    },
+    {
+      "epoch": 0.13230681494154548,
+      "grad_norm": 1.1602419211992863,
+      "learning_rate": 9.729086208503174e-06,
+      "loss": 0.0285,
+      "step": 464
+    },
+    {
+      "epoch": 0.13259195893926431,
+      "grad_norm": 2.3952533816912234,
+      "learning_rate": 9.727584524161568e-06,
+      "loss": 0.062,
+      "step": 465
+    },
+    {
+      "epoch": 0.13287710293698318,
+      "grad_norm": 3.1634854012983693,
+      "learning_rate": 9.726078805914156e-06,
+      "loss": 0.0585,
+      "step": 466
+    },
+    {
+      "epoch": 0.13316224693470202,
+      "grad_norm": 2.081821935988785,
+      "learning_rate": 9.724569055045722e-06,
+      "loss": 0.0497,
+      "step": 467
+    },
+    {
+      "epoch": 0.13344739093242086,
+      "grad_norm": 1.7133129071243365,
+      "learning_rate": 9.723055272844492e-06,
+      "loss": 0.0231,
+      "step": 468
+    },
+    {
+      "epoch": 0.13373253493013973,
+      "grad_norm": 1.3379025505204591,
+      "learning_rate": 9.72153746060213e-06,
+      "loss": 0.0296,
+      "step": 469
+    },
+    {
+      "epoch": 0.13401767892785857,
+      "grad_norm": 2.4905984184567176,
+      "learning_rate": 9.720015619613738e-06,
+      "loss": 0.0341,
+      "step": 470
+    },
+    {
+      "epoch": 0.1343028229255774,
+      "grad_norm": 1.7732686088505376,
+      "learning_rate": 9.718489751177863e-06,
+      "loss": 0.046,
+      "step": 471
+    },
+    {
+      "epoch": 0.13458796692329628,
+      "grad_norm": 1.4099272711066784,
+      "learning_rate": 9.716959856596476e-06,
+      "loss": 0.0517,
+      "step": 472
+    },
+    {
+      "epoch": 0.13487311092101512,
+      "grad_norm": 0.5759751962800088,
+      "learning_rate": 9.715425937174992e-06,
+      "loss": 0.0205,
+      "step": 473
+    },
+    {
+      "epoch": 0.13515825491873396,
+      "grad_norm": 1.8213929340301374,
+      "learning_rate": 9.71388799422226e-06,
+      "loss": 0.0424,
+      "step": 474
+    },
+    {
+      "epoch": 0.1354433989164528,
+      "grad_norm": 1.1189923917553266,
+      "learning_rate": 9.712346029050561e-06,
+      "loss": 0.0261,
+      "step": 475
+    },
+    {
+      "epoch": 0.13572854291417166,
+      "grad_norm": 0.7740540942029566,
+      "learning_rate": 9.710800042975604e-06,
+      "loss": 0.0229,
+      "step": 476
+    },
+    {
+      "epoch": 0.1360136869118905,
+      "grad_norm": 0.7017604368884549,
+      "learning_rate": 9.709250037316535e-06,
+      "loss": 0.0189,
+      "step": 477
+    },
+    {
+      "epoch": 0.13629883090960934,
+      "grad_norm": 0.8260830471935292,
+      "learning_rate": 9.707696013395929e-06,
+      "loss": 0.0222,
+      "step": 478
+    },
+    {
+      "epoch": 0.1365839749073282,
+      "grad_norm": 1.2766002745562819,
+      "learning_rate": 9.706137972539784e-06,
+      "loss": 0.0264,
+      "step": 479
+    },
+    {
+      "epoch": 0.13686911890504705,
+      "grad_norm": 1.316995242325395,
+      "learning_rate": 9.70457591607753e-06,
+      "loss": 0.029,
+      "step": 480
+    },
+    {
+      "epoch": 0.1371542629027659,
+      "grad_norm": 1.3031733188507375,
+      "learning_rate": 9.703009845342027e-06,
+      "loss": 0.0306,
+      "step": 481
+    },
+    {
+      "epoch": 0.13743940690048476,
+      "grad_norm": 2.0441243406454515,
+      "learning_rate": 9.701439761669551e-06,
+      "loss": 0.0384,
+      "step": 482
+    },
+    {
+      "epoch": 0.1377245508982036,
+      "grad_norm": 4.4329827478789765,
+      "learning_rate": 9.69986566639981e-06,
+      "loss": 0.0669,
+      "step": 483
+    },
+    {
+      "epoch": 0.13800969489592244,
+      "grad_norm": 0.550863641831382,
+      "learning_rate": 9.698287560875932e-06,
+      "loss": 0.011,
+      "step": 484
+    },
+    {
+      "epoch": 0.13829483889364128,
+      "grad_norm": 1.367528196902721,
+      "learning_rate": 9.696705446444465e-06,
+      "loss": 0.0475,
+      "step": 485
+    },
+    {
+      "epoch": 0.13857998289136014,
+      "grad_norm": 2.9049463748248625,
+      "learning_rate": 9.695119324455383e-06,
+      "loss": 0.048,
+      "step": 486
+    },
+    {
+      "epoch": 0.13886512688907898,
+      "grad_norm": 2.152275954468813,
+      "learning_rate": 9.693529196262073e-06,
+      "loss": 0.0426,
+      "step": 487
+    },
+    {
+      "epoch": 0.13915027088679782,
+      "grad_norm": 1.4509315149948045,
+      "learning_rate": 9.691935063221347e-06,
+      "loss": 0.0371,
+      "step": 488
+    },
+    {
+      "epoch": 0.1394354148845167,
+      "grad_norm": 0.924165717850074,
+      "learning_rate": 9.690336926693427e-06,
+      "loss": 0.0165,
+      "step": 489
+    },
+    {
+      "epoch": 0.13972055888223553,
+      "grad_norm": 1.5645440774065995,
+      "learning_rate": 9.688734788041958e-06,
+      "loss": 0.0222,
+      "step": 490
+    },
+    {
+      "epoch": 0.14000570287995437,
+      "grad_norm": 1.8857089742319397,
+      "learning_rate": 9.687128648633995e-06,
+      "loss": 0.0338,
+      "step": 491
+    },
+    {
+      "epoch": 0.14029084687767324,
+      "grad_norm": 1.2496068613960891,
+      "learning_rate": 9.685518509840008e-06,
+      "loss": 0.0393,
+      "step": 492
+    },
+    {
+      "epoch": 0.14057599087539208,
+      "grad_norm": 2.2609376738497797,
+      "learning_rate": 9.683904373033884e-06,
+      "loss": 0.0424,
+      "step": 493
+    },
+    {
+      "epoch": 0.14086113487311092,
+      "grad_norm": 0.6716431956995677,
+      "learning_rate": 9.682286239592912e-06,
+      "loss": 0.0189,
+      "step": 494
+    },
+    {
+      "epoch": 0.14114627887082976,
+      "grad_norm": 1.8445797069209673,
+      "learning_rate": 9.6806641108978e-06,
+      "loss": 0.0467,
+      "step": 495
+    },
+    {
+      "epoch": 0.14143142286854862,
+      "grad_norm": 1.039428374248318,
+      "learning_rate": 9.67903798833266e-06,
+      "loss": 0.0296,
+      "step": 496
+    },
+    {
+      "epoch": 0.14171656686626746,
+      "grad_norm": 2.241000783231958,
+      "learning_rate": 9.677407873285016e-06,
+      "loss": 0.021,
+      "step": 497
+    },
+    {
+      "epoch": 0.1420017108639863,
+      "grad_norm": 2.1580510158063904,
+      "learning_rate": 9.675773767145795e-06,
+      "loss": 0.0404,
+      "step": 498
+    },
+    {
+      "epoch": 0.14228685486170517,
+      "grad_norm": 2.1848379943693854,
+      "learning_rate": 9.674135671309329e-06,
+      "loss": 0.0392,
+      "step": 499
+    },
+    {
+      "epoch": 0.142571998859424,
+      "grad_norm": 1.2914001435408309,
+      "learning_rate": 9.672493587173356e-06,
+      "loss": 0.0193,
+      "step": 500
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 1.6915143202313678,
+      "learning_rate": 9.670847516139019e-06,
+      "loss": 0.0287,
+      "step": 501
+    },
+    {
+      "epoch": 0.14314228685486172,
+      "grad_norm": 0.6349880464190206,
+      "learning_rate": 9.66919745961086e-06,
+      "loss": 0.011,
+      "step": 502
+    },
+    {
+      "epoch": 0.14342743085258056,
+      "grad_norm": 1.3888680526489037,
+      "learning_rate": 9.667543418996824e-06,
+      "loss": 0.0206,
+      "step": 503
+    },
+    {
+      "epoch": 0.1437125748502994,
+      "grad_norm": 1.9166676274285204,
+      "learning_rate": 9.665885395708252e-06,
+      "loss": 0.0321,
+      "step": 504
+    },
+    {
+      "epoch": 0.14399771884801824,
+      "grad_norm": 0.6986695674839958,
+      "learning_rate": 9.664223391159885e-06,
+      "loss": 0.0222,
+      "step": 505
+    },
+    {
+      "epoch": 0.1442828628457371,
+      "grad_norm": 0.9132750549727724,
+      "learning_rate": 9.662557406769865e-06,
+      "loss": 0.026,
+      "step": 506
+    },
+    {
+      "epoch": 0.14456800684345594,
+      "grad_norm": 1.5540547156511098,
+      "learning_rate": 9.660887443959726e-06,
+      "loss": 0.039,
+      "step": 507
+    },
+    {
+      "epoch": 0.14485315084117478,
+      "grad_norm": 0.9598545046429632,
+      "learning_rate": 9.659213504154393e-06,
+      "loss": 0.0211,
+      "step": 508
+    },
+    {
+      "epoch": 0.14513829483889365,
+      "grad_norm": 2.2737404573757862,
+      "learning_rate": 9.65753558878219e-06,
+      "loss": 0.0239,
+      "step": 509
+    },
+    {
+      "epoch": 0.1454234388366125,
+      "grad_norm": 1.5062825135884002,
+      "learning_rate": 9.655853699274834e-06,
+      "loss": 0.0345,
+      "step": 510
+    },
+    {
+      "epoch": 0.14570858283433133,
+      "grad_norm": 1.6018765019531456,
+      "learning_rate": 9.65416783706743e-06,
+      "loss": 0.0209,
+      "step": 511
+    },
+    {
+      "epoch": 0.1459937268320502,
+      "grad_norm": 1.2442982344382103,
+      "learning_rate": 9.652478003598471e-06,
+      "loss": 0.0285,
+      "step": 512
+    },
+    {
+      "epoch": 0.14627887082976904,
+      "grad_norm": 1.9216622846090483,
+      "learning_rate": 9.650784200309847e-06,
+      "loss": 0.0339,
+      "step": 513
+    },
+    {
+      "epoch": 0.14656401482748788,
+      "grad_norm": 2.3825991592561717,
+      "learning_rate": 9.64908642864682e-06,
+      "loss": 0.044,
+      "step": 514
+    },
+    {
+      "epoch": 0.14684915882520672,
+      "grad_norm": 2.5818323385820916,
+      "learning_rate": 9.647384690058058e-06,
+      "loss": 0.0353,
+      "step": 515
+    },
+    {
+      "epoch": 0.14713430282292558,
+      "grad_norm": 1.5120506419436006,
+      "learning_rate": 9.645678985995597e-06,
+      "loss": 0.0184,
+      "step": 516
+    },
+    {
+      "epoch": 0.14741944682064442,
+      "grad_norm": 3.092529038403767,
+      "learning_rate": 9.643969317914865e-06,
+      "loss": 0.0459,
+      "step": 517
+    },
+    {
+      "epoch": 0.14770459081836326,
+      "grad_norm": 1.6305582095574176,
+      "learning_rate": 9.642255687274669e-06,
+      "loss": 0.0745,
+      "step": 518
+    },
+    {
+      "epoch": 0.14798973481608213,
+      "grad_norm": 1.9745427124068613,
+      "learning_rate": 9.6405380955372e-06,
+      "loss": 0.0431,
+      "step": 519
+    },
+    {
+      "epoch": 0.14827487881380097,
+      "grad_norm": 1.1060350887232473,
+      "learning_rate": 9.638816544168027e-06,
+      "loss": 0.0401,
+      "step": 520
+    },
+    {
+      "epoch": 0.1485600228115198,
+      "grad_norm": 1.48122125532716,
+      "learning_rate": 9.637091034636097e-06,
+      "loss": 0.0523,
+      "step": 521
+    },
+    {
+      "epoch": 0.14884516680923868,
+      "grad_norm": 1.7854176889849915,
+      "learning_rate": 9.635361568413739e-06,
+      "loss": 0.0392,
+      "step": 522
+    },
+    {
+      "epoch": 0.14913031080695752,
+      "grad_norm": 1.5771118563507494,
+      "learning_rate": 9.633628146976649e-06,
+      "loss": 0.0291,
+      "step": 523
+    },
+    {
+      "epoch": 0.14941545480467636,
+      "grad_norm": 1.170813729239735,
+      "learning_rate": 9.631890771803909e-06,
+      "loss": 0.0324,
+      "step": 524
+    },
+    {
+      "epoch": 0.1497005988023952,
+      "grad_norm": 1.1607083768274689,
+      "learning_rate": 9.630149444377964e-06,
+      "loss": 0.027,
+      "step": 525
+    },
+    {
+      "epoch": 0.14998574280011406,
+      "grad_norm": 2.6420375275270946,
+      "learning_rate": 9.628404166184639e-06,
+      "loss": 0.0392,
+      "step": 526
+    },
+    {
+      "epoch": 0.1502708867978329,
+      "grad_norm": 1.5977547719023326,
+      "learning_rate": 9.626654938713128e-06,
+      "loss": 0.039,
+      "step": 527
+    },
+    {
+      "epoch": 0.15055603079555174,
+      "grad_norm": 2.4449410251819055,
+      "learning_rate": 9.624901763455994e-06,
+      "loss": 0.1085,
+      "step": 528
+    },
+    {
+      "epoch": 0.1508411747932706,
+      "grad_norm": 1.66447527220779,
+      "learning_rate": 9.623144641909167e-06,
+      "loss": 0.0453,
+      "step": 529
+    },
+    {
+      "epoch": 0.15112631879098945,
+      "grad_norm": 3.13057218774973,
+      "learning_rate": 9.621383575571948e-06,
+      "loss": 0.057,
+      "step": 530
+    },
+    {
+      "epoch": 0.1514114627887083,
+      "grad_norm": 1.2160935811246358,
+      "learning_rate": 9.619618565947e-06,
+      "loss": 0.0424,
+      "step": 531
+    },
+    {
+      "epoch": 0.15169660678642716,
+      "grad_norm": 1.4641008759404945,
+      "learning_rate": 9.617849614540356e-06,
+      "loss": 0.048,
+      "step": 532
+    },
+    {
+      "epoch": 0.151981750784146,
+      "grad_norm": 1.4923618640634606,
+      "learning_rate": 9.616076722861406e-06,
+      "loss": 0.035,
+      "step": 533
+    },
+    {
+      "epoch": 0.15226689478186484,
+      "grad_norm": 0.7354465216068986,
+      "learning_rate": 9.614299892422905e-06,
+      "loss": 0.0328,
+      "step": 534
+    },
+    {
+      "epoch": 0.15255203877958368,
+      "grad_norm": 1.3156709351195746,
+      "learning_rate": 9.61251912474097e-06,
+      "loss": 0.0295,
+      "step": 535
+    },
+    {
+      "epoch": 0.15283718277730254,
+      "grad_norm": 2.465389798366464,
+      "learning_rate": 9.610734421335078e-06,
+      "loss": 0.0355,
+      "step": 536
+    },
+    {
+      "epoch": 0.15312232677502138,
+      "grad_norm": 2.107752927710834,
+      "learning_rate": 9.608945783728061e-06,
+      "loss": 0.0304,
+      "step": 537
+    },
+    {
+      "epoch": 0.15340747077274022,
+      "grad_norm": 0.8382376238641022,
+      "learning_rate": 9.60715321344611e-06,
+      "loss": 0.0292,
+      "step": 538
+    },
+    {
+      "epoch": 0.1536926147704591,
+      "grad_norm": 1.3467398331487281,
+      "learning_rate": 9.605356712018773e-06,
+      "loss": 0.0385,
+      "step": 539
+    },
+    {
+      "epoch": 0.15397775876817793,
+      "grad_norm": 2.0809206260900552,
+      "learning_rate": 9.603556280978947e-06,
+      "loss": 0.0572,
+      "step": 540
+    },
+    {
+      "epoch": 0.15426290276589677,
+      "grad_norm": 1.2901980818849303,
+      "learning_rate": 9.60175192186289e-06,
+      "loss": 0.0193,
+      "step": 541
+    },
+    {
+      "epoch": 0.15454804676361564,
+      "grad_norm": 0.736806847215193,
+      "learning_rate": 9.599943636210204e-06,
+      "loss": 0.0284,
+      "step": 542
+    },
+    {
+      "epoch": 0.15483319076133448,
+      "grad_norm": 1.5679254655317187,
+      "learning_rate": 9.598131425563847e-06,
+      "loss": 0.0317,
+      "step": 543
+    },
+    {
+      "epoch": 0.15511833475905332,
+      "grad_norm": 0.7132612878107385,
+      "learning_rate": 9.596315291470122e-06,
+      "loss": 0.0114,
+      "step": 544
+    },
+    {
+      "epoch": 0.15540347875677216,
+      "grad_norm": 2.593670926612122,
+      "learning_rate": 9.594495235478685e-06,
+      "loss": 0.0463,
+      "step": 545
+    },
+    {
+      "epoch": 0.15568862275449102,
+      "grad_norm": 1.1955510290910427,
+      "learning_rate": 9.59267125914253e-06,
+      "loss": 0.0254,
+      "step": 546
+    },
+    {
+      "epoch": 0.15597376675220986,
+      "grad_norm": 1.045769334119973,
+      "learning_rate": 9.590843364018005e-06,
+      "loss": 0.029,
+      "step": 547
+    },
+    {
+      "epoch": 0.1562589107499287,
+      "grad_norm": 1.6801520699380361,
+      "learning_rate": 9.589011551664797e-06,
+      "loss": 0.0572,
+      "step": 548
+    },
+    {
+      "epoch": 0.15654405474764757,
+      "grad_norm": 1.5891642039931364,
+      "learning_rate": 9.587175823645936e-06,
+      "loss": 0.0296,
+      "step": 549
+    },
+    {
+      "epoch": 0.1568291987453664,
+      "grad_norm": 1.4409334872176962,
+      "learning_rate": 9.585336181527795e-06,
+      "loss": 0.0264,
+      "step": 550
+    },
+    {
+      "epoch": 0.15711434274308525,
+      "grad_norm": 2.2586423562130116,
+      "learning_rate": 9.583492626880082e-06,
+      "loss": 0.058,
+      "step": 551
+    },
+    {
+      "epoch": 0.15739948674080412,
+      "grad_norm": 1.661999933017668,
+      "learning_rate": 9.581645161275852e-06,
+      "loss": 0.0551,
+      "step": 552
+    },
+    {
+      "epoch": 0.15768463073852296,
+      "grad_norm": 1.7655597655232174,
+      "learning_rate": 9.579793786291486e-06,
+      "loss": 0.0278,
+      "step": 553
+    },
+    {
+      "epoch": 0.1579697747362418,
+      "grad_norm": 3.589911729513461,
+      "learning_rate": 9.577938503506712e-06,
+      "loss": 0.0517,
+      "step": 554
+    },
+    {
+      "epoch": 0.15825491873396064,
+      "grad_norm": 3.001411920745496,
+      "learning_rate": 9.576079314504584e-06,
+      "loss": 0.0449,
+      "step": 555
+    },
+    {
+      "epoch": 0.1585400627316795,
+      "grad_norm": 1.604218093067663,
+      "learning_rate": 9.574216220871492e-06,
+      "loss": 0.0246,
+      "step": 556
+    },
+    {
+      "epoch": 0.15882520672939834,
+      "grad_norm": 0.935156656620672,
+      "learning_rate": 9.57234922419716e-06,
+      "loss": 0.0161,
+      "step": 557
+    },
+    {
+      "epoch": 0.15911035072711718,
+      "grad_norm": 1.1513995231538736,
+      "learning_rate": 9.570478326074638e-06,
+      "loss": 0.0283,
+      "step": 558
+    },
+    {
+      "epoch": 0.15939549472483605,
+      "grad_norm": 1.4012387052969366,
+      "learning_rate": 9.568603528100306e-06,
+      "loss": 0.0325,
+      "step": 559
+    },
+    {
+      "epoch": 0.1596806387225549,
+      "grad_norm": 1.5162153854252265,
+      "learning_rate": 9.566724831873876e-06,
+      "loss": 0.0225,
+      "step": 560
+    },
+    {
+      "epoch": 0.15996578272027373,
+      "grad_norm": 0.846516598739829,
+      "learning_rate": 9.564842238998381e-06,
+      "loss": 0.0099,
+      "step": 561
+    },
+    {
+      "epoch": 0.1602509267179926,
+      "grad_norm": 2.647135598055157,
+      "learning_rate": 9.562955751080183e-06,
+      "loss": 0.0532,
+      "step": 562
+    },
+    {
+      "epoch": 0.16053607071571144,
+      "grad_norm": 1.073278252413763,
+      "learning_rate": 9.561065369728963e-06,
+      "loss": 0.0355,
+      "step": 563
+    },
+    {
+      "epoch": 0.16082121471343028,
+      "grad_norm": 1.170548007164538,
+      "learning_rate": 9.559171096557728e-06,
+      "loss": 0.029,
+      "step": 564
+    },
+    {
+      "epoch": 0.16110635871114912,
+      "grad_norm": 1.9939002606252199,
+      "learning_rate": 9.557272933182804e-06,
+      "loss": 0.036,
+      "step": 565
+    },
+    {
+      "epoch": 0.16139150270886798,
+      "grad_norm": 2.1480044032388466,
+      "learning_rate": 9.555370881223837e-06,
+      "loss": 0.0436,
+      "step": 566
+    },
+    {
+      "epoch": 0.16167664670658682,
+      "grad_norm": 1.325637196128801,
+      "learning_rate": 9.55346494230379e-06,
+      "loss": 0.0157,
+      "step": 567
+    },
+    {
+      "epoch": 0.16196179070430566,
+      "grad_norm": 0.7564851499202597,
+      "learning_rate": 9.551555118048943e-06,
+      "loss": 0.0249,
+      "step": 568
+    },
+    {
+      "epoch": 0.16224693470202453,
+      "grad_norm": 0.9722401904633631,
+      "learning_rate": 9.549641410088895e-06,
+      "loss": 0.0231,
+      "step": 569
+    },
+    {
+      "epoch": 0.16253207869974337,
+      "grad_norm": 0.8478045120928479,
+      "learning_rate": 9.547723820056552e-06,
+      "loss": 0.0105,
+      "step": 570
+    },
+    {
+      "epoch": 0.1628172226974622,
+      "grad_norm": 1.351787113285877,
+      "learning_rate": 9.545802349588136e-06,
+      "loss": 0.0311,
+      "step": 571
+    },
+    {
+      "epoch": 0.16310236669518108,
+      "grad_norm": 0.836648012433114,
+      "learning_rate": 9.543877000323181e-06,
+      "loss": 0.023,
+      "step": 572
+    },
+    {
+      "epoch": 0.16338751069289992,
+      "grad_norm": 0.7901380885568204,
+      "learning_rate": 9.54194777390453e-06,
+      "loss": 0.014,
+      "step": 573
+    },
+    {
+      "epoch": 0.16367265469061876,
+      "grad_norm": 1.2938393184716226,
+      "learning_rate": 9.540014671978335e-06,
+      "loss": 0.0632,
+      "step": 574
+    },
+    {
+      "epoch": 0.1639577986883376,
+      "grad_norm": 1.2836949271527522,
+      "learning_rate": 9.53807769619405e-06,
+      "loss": 0.0231,
+      "step": 575
+    },
+    {
+      "epoch": 0.16424294268605646,
+      "grad_norm": 2.27381818725564,
+      "learning_rate": 9.536136848204443e-06,
+      "loss": 0.0418,
+      "step": 576
+    },
+    {
+      "epoch": 0.1645280866837753,
+      "grad_norm": 2.4569515266903026,
+      "learning_rate": 9.534192129665578e-06,
+      "loss": 0.0141,
+      "step": 577
+    },
+    {
+      "epoch": 0.16481323068149414,
+      "grad_norm": 1.7019979551998075,
+      "learning_rate": 9.532243542236826e-06,
+      "loss": 0.0159,
+      "step": 578
+    },
+    {
+      "epoch": 0.165098374679213,
+      "grad_norm": 0.7491727489330064,
+      "learning_rate": 9.530291087580857e-06,
+      "loss": 0.011,
+      "step": 579
+    },
+    {
+      "epoch": 0.16538351867693185,
+      "grad_norm": 1.0697725490475511,
+      "learning_rate": 9.528334767363643e-06,
+      "loss": 0.0316,
+      "step": 580
+    },
+    {
+      "epoch": 0.1656686626746507,
+      "grad_norm": 1.514917067065579,
+      "learning_rate": 9.526374583254454e-06,
+      "loss": 0.0247,
+      "step": 581
+    },
+    {
+      "epoch": 0.16595380667236956,
+      "grad_norm": 1.6526372073335776,
+      "learning_rate": 9.524410536925854e-06,
+      "loss": 0.0242,
+      "step": 582
+    },
+    {
+      "epoch": 0.1662389506700884,
+      "grad_norm": 1.154206493649726,
+      "learning_rate": 9.522442630053708e-06,
+      "loss": 0.0151,
+      "step": 583
+    },
+    {
+      "epoch": 0.16652409466780724,
+      "grad_norm": 2.220783739275899,
+      "learning_rate": 9.520470864317169e-06,
+      "loss": 0.0405,
+      "step": 584
+    },
+    {
+      "epoch": 0.16680923866552608,
+      "grad_norm": 2.1478234313482005,
+      "learning_rate": 9.518495241398684e-06,
+      "loss": 0.0255,
+      "step": 585
+    },
+    {
+      "epoch": 0.16709438266324494,
+      "grad_norm": 1.888990019043023,
+      "learning_rate": 9.516515762983996e-06,
+      "loss": 0.023,
+      "step": 586
+    },
+    {
+      "epoch": 0.16737952666096378,
+      "grad_norm": 0.568362822177166,
+      "learning_rate": 9.514532430762133e-06,
+      "loss": 0.0234,
+      "step": 587
+    },
+    {
+      "epoch": 0.16766467065868262,
+      "grad_norm": 2.666053376987735,
+      "learning_rate": 9.512545246425416e-06,
+      "loss": 0.0334,
+      "step": 588
+    },
+    {
+      "epoch": 0.1679498146564015,
+      "grad_norm": 0.9889048397741379,
+      "learning_rate": 9.510554211669443e-06,
+      "loss": 0.0272,
+      "step": 589
+    },
+    {
+      "epoch": 0.16823495865412033,
+      "grad_norm": 2.663367144657973,
+      "learning_rate": 9.50855932819311e-06,
+      "loss": 0.0863,
+      "step": 590
+    },
+    {
+      "epoch": 0.16852010265183917,
+      "grad_norm": 3.0842378390540564,
+      "learning_rate": 9.506560597698588e-06,
+      "loss": 0.0544,
+      "step": 591
+    },
+    {
+      "epoch": 0.16880524664955804,
+      "grad_norm": 2.2352938418828945,
+      "learning_rate": 9.504558021891335e-06,
+      "loss": 0.0597,
+      "step": 592
+    },
+    {
+      "epoch": 0.16909039064727688,
+      "grad_norm": 0.9718298379039145,
+      "learning_rate": 9.502551602480087e-06,
+      "loss": 0.0243,
+      "step": 593
+    },
+    {
+      "epoch": 0.16937553464499572,
+      "grad_norm": 0.9679204345252653,
+      "learning_rate": 9.500541341176865e-06,
+      "loss": 0.0226,
+      "step": 594
+    },
+    {
+      "epoch": 0.16966067864271456,
+      "grad_norm": 1.9624333575904165,
+      "learning_rate": 9.498527239696962e-06,
+      "loss": 0.0303,
+      "step": 595
+    },
+    {
+      "epoch": 0.16994582264043342,
+      "grad_norm": 1.0983438075432492,
+      "learning_rate": 9.496509299758949e-06,
+      "loss": 0.0263,
+      "step": 596
+    },
+    {
+      "epoch": 0.17023096663815226,
+      "grad_norm": 1.485540944433959,
+      "learning_rate": 9.494487523084676e-06,
+      "loss": 0.018,
+      "step": 597
+    },
+    {
+      "epoch": 0.1705161106358711,
+      "grad_norm": 0.7546809410329396,
+      "learning_rate": 9.492461911399265e-06,
+      "loss": 0.0208,
+      "step": 598
+    },
+    {
+      "epoch": 0.17080125463358997,
+      "grad_norm": 1.2028050799273944,
+      "learning_rate": 9.490432466431107e-06,
+      "loss": 0.0126,
+      "step": 599
+    },
+    {
+      "epoch": 0.1710863986313088,
+      "grad_norm": 2.3029305869735377,
+      "learning_rate": 9.488399189911866e-06,
+      "loss": 0.0396,
+      "step": 600
+    },
+    {
+      "epoch": 0.17137154262902765,
+      "grad_norm": 1.7854430171080835,
+      "learning_rate": 9.486362083576479e-06,
+      "loss": 0.0228,
+      "step": 601
+    },
+    {
+      "epoch": 0.17165668662674652,
+      "grad_norm": 1.2139930054544301,
+      "learning_rate": 9.484321149163145e-06,
+      "loss": 0.053,
+      "step": 602
+    },
+    {
+      "epoch": 0.17194183062446536,
+      "grad_norm": 1.8652637627111008,
+      "learning_rate": 9.482276388413331e-06,
+      "loss": 0.0243,
+      "step": 603
+    },
+    {
+      "epoch": 0.1722269746221842,
+      "grad_norm": 1.733593841920375,
+      "learning_rate": 9.480227803071775e-06,
+      "loss": 0.0305,
+      "step": 604
+    },
+    {
+      "epoch": 0.17251211861990304,
+      "grad_norm": 2.218846595287222,
+      "learning_rate": 9.478175394886469e-06,
+      "loss": 0.0345,
+      "step": 605
+    },
+    {
+      "epoch": 0.1727972626176219,
+      "grad_norm": 2.4849800012227345,
+      "learning_rate": 9.47611916560867e-06,
+      "loss": 0.0359,
+      "step": 606
+    },
+    {
+      "epoch": 0.17308240661534074,
+      "grad_norm": 3.832248662370049,
+      "learning_rate": 9.474059116992901e-06,
+      "loss": 0.0489,
+      "step": 607
+    },
+    {
+      "epoch": 0.17336755061305958,
+      "grad_norm": 1.2556167721835865,
+      "learning_rate": 9.471995250796936e-06,
+      "loss": 0.0224,
+      "step": 608
+    },
+    {
+      "epoch": 0.17365269461077845,
+      "grad_norm": 1.3725885311717074,
+      "learning_rate": 9.469927568781814e-06,
+      "loss": 0.0246,
+      "step": 609
+    },
+    {
+      "epoch": 0.1739378386084973,
+      "grad_norm": 2.0268680892167312,
+      "learning_rate": 9.467856072711821e-06,
+      "loss": 0.0377,
+      "step": 610
+    },
+    {
+      "epoch": 0.17422298260621613,
+      "grad_norm": 2.13735531583806,
+      "learning_rate": 9.465780764354505e-06,
+      "loss": 0.032,
+      "step": 611
+    },
+    {
+      "epoch": 0.174508126603935,
+      "grad_norm": 1.0077175687597997,
+      "learning_rate": 9.463701645480665e-06,
+      "loss": 0.0125,
+      "step": 612
+    },
+    {
+      "epoch": 0.17479327060165384,
+      "grad_norm": 1.4433624480865979,
+      "learning_rate": 9.46161871786435e-06,
+      "loss": 0.0213,
+      "step": 613
+    },
+    {
+      "epoch": 0.17507841459937268,
+      "grad_norm": 1.7363894985553223,
+      "learning_rate": 9.459531983282858e-06,
+      "loss": 0.0312,
+      "step": 614
+    },
+    {
+      "epoch": 0.17536355859709152,
+      "grad_norm": 1.1480559254413287,
+      "learning_rate": 9.45744144351674e-06,
+      "loss": 0.0126,
+      "step": 615
+    },
+    {
+      "epoch": 0.17564870259481039,
+      "grad_norm": 1.0144730105958946,
+      "learning_rate": 9.455347100349785e-06,
+      "loss": 0.0465,
+      "step": 616
+    },
+    {
+      "epoch": 0.17593384659252922,
+      "grad_norm": 2.3315897375286956,
+      "learning_rate": 9.453248955569041e-06,
+      "loss": 0.0375,
+      "step": 617
+    },
+    {
+      "epoch": 0.17621899059024806,
+      "grad_norm": 1.517116537675244,
+      "learning_rate": 9.451147010964786e-06,
+      "loss": 0.0255,
+      "step": 618
+    },
+    {
+      "epoch": 0.17650413458796693,
+      "grad_norm": 1.3539214290926733,
+      "learning_rate": 9.449041268330549e-06,
+      "loss": 0.0175,
+      "step": 619
+    },
+    {
+      "epoch": 0.17678927858568577,
+      "grad_norm": 0.8250630045444727,
+      "learning_rate": 9.446931729463093e-06,
+      "loss": 0.0189,
+      "step": 620
+    },
+    {
+      "epoch": 0.1770744225834046,
+      "grad_norm": 1.6566428427949393,
+      "learning_rate": 9.44481839616243e-06,
+      "loss": 0.0263,
+      "step": 621
+    },
+    {
+      "epoch": 0.17735956658112348,
+      "grad_norm": 1.296549496531528,
+      "learning_rate": 9.442701270231799e-06,
+      "loss": 0.0357,
+      "step": 622
+    },
+    {
+      "epoch": 0.17764471057884232,
+      "grad_norm": 1.2623177453839831,
+      "learning_rate": 9.440580353477682e-06,
+      "loss": 0.0138,
+      "step": 623
+    },
+    {
+      "epoch": 0.17792985457656116,
+      "grad_norm": 0.9401424954471435,
+      "learning_rate": 9.438455647709794e-06,
+      "loss": 0.0344,
+      "step": 624
+    },
+    {
+      "epoch": 0.17821499857428,
+      "grad_norm": 1.2061707620169295,
+      "learning_rate": 9.436327154741082e-06,
+      "loss": 0.0204,
+      "step": 625
+    },
+    {
+      "epoch": 0.17850014257199887,
+      "grad_norm": 3.5812415668133295,
+      "learning_rate": 9.434194876387723e-06,
+      "loss": 0.0348,
+      "step": 626
+    },
+    {
+      "epoch": 0.1787852865697177,
+      "grad_norm": 2.179594593613911,
+      "learning_rate": 9.43205881446913e-06,
+      "loss": 0.0405,
+      "step": 627
+    },
+    {
+      "epoch": 0.17907043056743654,
+      "grad_norm": 3.133447888160429,
+      "learning_rate": 9.429918970807939e-06,
+      "loss": 0.0883,
+      "step": 628
+    },
+    {
+      "epoch": 0.1793555745651554,
+      "grad_norm": 1.1894173921814282,
+      "learning_rate": 9.427775347230013e-06,
+      "loss": 0.0459,
+      "step": 629
+    },
+    {
+      "epoch": 0.17964071856287425,
+      "grad_norm": 1.220795451723468,
+      "learning_rate": 9.425627945564442e-06,
+      "loss": 0.0276,
+      "step": 630
+    },
+    {
+      "epoch": 0.1799258625605931,
+      "grad_norm": 0.9556676101760306,
+      "learning_rate": 9.423476767643539e-06,
+      "loss": 0.0408,
+      "step": 631
+    },
+    {
+      "epoch": 0.18021100655831196,
+      "grad_norm": 1.3012772103524362,
+      "learning_rate": 9.42132181530284e-06,
+      "loss": 0.0147,
+      "step": 632
+    },
+    {
+      "epoch": 0.1804961505560308,
+      "grad_norm": 0.8734603524960712,
+      "learning_rate": 9.419163090381102e-06,
+      "loss": 0.0232,
+      "step": 633
+    },
+    {
+      "epoch": 0.18078129455374964,
+      "grad_norm": 1.2200670729025649,
+      "learning_rate": 9.4170005947203e-06,
+      "loss": 0.0329,
+      "step": 634
+    },
+    {
+      "epoch": 0.18106643855146848,
+      "grad_norm": 1.1685662362242049,
+      "learning_rate": 9.414834330165626e-06,
+      "loss": 0.0372,
+      "step": 635
+    },
+    {
+      "epoch": 0.18135158254918735,
+      "grad_norm": 0.6949628469219308,
+      "learning_rate": 9.412664298565486e-06,
+      "loss": 0.0246,
+      "step": 636
+    },
+    {
+      "epoch": 0.18163672654690619,
+      "grad_norm": 1.771280848314573,
+      "learning_rate": 9.410490501771507e-06,
+      "loss": 0.0306,
+      "step": 637
+    },
+    {
+      "epoch": 0.18192187054462503,
+      "grad_norm": 1.2498455618438535,
+      "learning_rate": 9.408312941638522e-06,
+      "loss": 0.0301,
+      "step": 638
+    },
+    {
+      "epoch": 0.1822070145423439,
+      "grad_norm": 2.278447179095516,
+      "learning_rate": 9.406131620024576e-06,
+      "loss": 0.0466,
+      "step": 639
+    },
+    {
+      "epoch": 0.18249215854006273,
+      "grad_norm": 1.0355835464758893,
+      "learning_rate": 9.403946538790931e-06,
+      "loss": 0.023,
+      "step": 640
+    },
+    {
+      "epoch": 0.18277730253778157,
+      "grad_norm": 0.683840225577905,
+      "learning_rate": 9.401757699802046e-06,
+      "loss": 0.0171,
+      "step": 641
+    },
+    {
+      "epoch": 0.18306244653550044,
+      "grad_norm": 2.090222100453515,
+      "learning_rate": 9.399565104925591e-06,
+      "loss": 0.0425,
+      "step": 642
+    },
+    {
+      "epoch": 0.18334759053321928,
+      "grad_norm": 2.036391317556621,
+      "learning_rate": 9.397368756032445e-06,
+      "loss": 0.0266,
+      "step": 643
+    },
+    {
+      "epoch": 0.18363273453093812,
+      "grad_norm": 1.738387885682237,
+      "learning_rate": 9.395168654996685e-06,
+      "loss": 0.0519,
+      "step": 644
+    },
+    {
+      "epoch": 0.18391787852865696,
+      "grad_norm": 2.117881545376208,
+      "learning_rate": 9.392964803695592e-06,
+      "loss": 0.032,
+      "step": 645
+    },
+    {
+      "epoch": 0.18420302252637583,
+      "grad_norm": 0.6059383955679531,
+      "learning_rate": 9.390757204009644e-06,
+      "loss": 0.0083,
+      "step": 646
+    },
+    {
+      "epoch": 0.18448816652409467,
+      "grad_norm": 0.5792275998008327,
+      "learning_rate": 9.38854585782252e-06,
+      "loss": 0.0127,
+      "step": 647
+    },
+    {
+      "epoch": 0.1847733105218135,
+      "grad_norm": 0.9221379221878636,
+      "learning_rate": 9.386330767021098e-06,
+      "loss": 0.0255,
+      "step": 648
+    },
+    {
+      "epoch": 0.18505845451953237,
+      "grad_norm": 0.6920928264035611,
+      "learning_rate": 9.384111933495442e-06,
+      "loss": 0.0084,
+      "step": 649
+    },
+    {
+      "epoch": 0.1853435985172512,
+      "grad_norm": 1.3255760857611412,
+      "learning_rate": 9.381889359138823e-06,
+      "loss": 0.0191,
+      "step": 650
+    },
+    {
+      "epoch": 0.18562874251497005,
+      "grad_norm": 1.056441268866933,
+      "learning_rate": 9.379663045847693e-06,
+      "loss": 0.0479,
+      "step": 651
+    },
+    {
+      "epoch": 0.18591388651268892,
+      "grad_norm": 2.0724011021981594,
+      "learning_rate": 9.377432995521701e-06,
+      "loss": 0.0248,
+      "step": 652
+    },
+    {
+      "epoch": 0.18619903051040776,
+      "grad_norm": 1.7914795536241521,
+      "learning_rate": 9.375199210063676e-06,
+      "loss": 0.0342,
+      "step": 653
+    },
+    {
+      "epoch": 0.1864841745081266,
+      "grad_norm": 1.540949023356334,
+      "learning_rate": 9.372961691379649e-06,
+      "loss": 0.0599,
+      "step": 654
+    },
+    {
+      "epoch": 0.18676931850584544,
+      "grad_norm": 0.6650657087107381,
+      "learning_rate": 9.37072044137882e-06,
+      "loss": 0.0148,
+      "step": 655
+    },
+    {
+      "epoch": 0.1870544625035643,
+      "grad_norm": 1.9248212315910453,
+      "learning_rate": 9.368475461973582e-06,
+      "loss": 0.0375,
+      "step": 656
+    },
+    {
+      "epoch": 0.18733960650128315,
+      "grad_norm": 2.808110344533141,
+      "learning_rate": 9.366226755079513e-06,
+      "loss": 0.0422,
+      "step": 657
+    },
+    {
+      "epoch": 0.18762475049900199,
+      "grad_norm": 2.1184357826543194,
+      "learning_rate": 9.36397432261536e-06,
+      "loss": 0.0282,
+      "step": 658
+    },
+    {
+      "epoch": 0.18790989449672085,
+      "grad_norm": 1.5031809588854805,
+      "learning_rate": 9.361718166503062e-06,
+      "loss": 0.0151,
+      "step": 659
+    },
+    {
+      "epoch": 0.1881950384944397,
+      "grad_norm": 0.9601384811433088,
+      "learning_rate": 9.359458288667725e-06,
+      "loss": 0.0414,
+      "step": 660
+    },
+    {
+      "epoch": 0.18848018249215853,
+      "grad_norm": 2.0522251099402053,
+      "learning_rate": 9.357194691037637e-06,
+      "loss": 0.0436,
+      "step": 661
+    },
+    {
+      "epoch": 0.1887653264898774,
+      "grad_norm": 0.8416753384643524,
+      "learning_rate": 9.354927375544256e-06,
+      "loss": 0.0124,
+      "step": 662
+    },
+    {
+      "epoch": 0.18905047048759624,
+      "grad_norm": 1.0668298335424686,
+      "learning_rate": 9.352656344122216e-06,
+      "loss": 0.0439,
+      "step": 663
+    },
+    {
+      "epoch": 0.18933561448531508,
+      "grad_norm": 1.34820663049341,
+      "learning_rate": 9.350381598709319e-06,
+      "loss": 0.0305,
+      "step": 664
+    },
+    {
+      "epoch": 0.18962075848303392,
+      "grad_norm": 1.5705169220785646,
+      "learning_rate": 9.348103141246538e-06,
+      "loss": 0.0191,
+      "step": 665
+    },
+    {
+      "epoch": 0.1899059024807528,
+      "grad_norm": 2.322186069862557,
+      "learning_rate": 9.345820973678011e-06,
+      "loss": 0.0284,
+      "step": 666
+    },
+    {
+      "epoch": 0.19019104647847163,
+      "grad_norm": 1.344929313625686,
+      "learning_rate": 9.343535097951044e-06,
+      "loss": 0.0343,
+      "step": 667
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 1.1700086372233123,
+      "learning_rate": 9.341245516016105e-06,
+      "loss": 0.018,
+      "step": 668
+    },
+    {
+      "epoch": 0.19076133447390933,
+      "grad_norm": 1.0706753371606232,
+      "learning_rate": 9.338952229826825e-06,
+      "loss": 0.0273,
+      "step": 669
+    },
+    {
+      "epoch": 0.19104647847162817,
+      "grad_norm": 1.6254449347628146,
+      "learning_rate": 9.336655241339999e-06,
+      "loss": 0.0367,
+      "step": 670
+    },
+    {
+      "epoch": 0.191331622469347,
+      "grad_norm": 67.18866812915265,
+      "learning_rate": 9.334354552515576e-06,
+      "loss": 0.0859,
+      "step": 671
+    },
+    {
+      "epoch": 0.19161676646706588,
+      "grad_norm": 191.39881973611224,
+      "learning_rate": 9.332050165316664e-06,
+      "loss": 0.5247,
+      "step": 672
+    },
+    {
+      "epoch": 0.19190191046478472,
+      "grad_norm": 33.27059153720094,
+      "learning_rate": 9.32974208170953e-06,
+      "loss": 0.0762,
+      "step": 673
+    },
+    {
+      "epoch": 0.19218705446250356,
+      "grad_norm": 2.1745024710329655,
+      "learning_rate": 9.327430303663589e-06,
+      "loss": 0.0331,
+      "step": 674
+    },
+    {
+      "epoch": 0.1924721984602224,
+      "grad_norm": 0.958496001953392,
+      "learning_rate": 9.325114833151414e-06,
+      "loss": 0.0174,
+      "step": 675
+    },
+    {
+      "epoch": 0.19275734245794127,
+      "grad_norm": 1.5630060617545107,
+      "learning_rate": 9.322795672148726e-06,
+      "loss": 0.0434,
+      "step": 676
+    },
+    {
+      "epoch": 0.1930424864556601,
+      "grad_norm": 4.90486633391519,
+      "learning_rate": 9.320472822634395e-06,
+      "loss": 0.0542,
+      "step": 677
+    },
+    {
+      "epoch": 0.19332763045337895,
+      "grad_norm": 0.5026753644154698,
+      "learning_rate": 9.31814628659044e-06,
+      "loss": 0.0108,
+      "step": 678
+    },
+    {
+      "epoch": 0.1936127744510978,
+      "grad_norm": 2.318017474873092,
+      "learning_rate": 9.315816066002024e-06,
+      "loss": 0.0408,
+      "step": 679
+    },
+    {
+      "epoch": 0.19389791844881665,
+      "grad_norm": 2.903171610551692,
+      "learning_rate": 9.313482162857452e-06,
+      "loss": 0.0557,
+      "step": 680
+    },
+    {
+      "epoch": 0.1941830624465355,
+      "grad_norm": 1.4218762758102965,
+      "learning_rate": 9.311144579148175e-06,
+      "loss": 0.0159,
+      "step": 681
+    },
+    {
+      "epoch": 0.19446820644425436,
+      "grad_norm": 2.009139671030263,
+      "learning_rate": 9.308803316868783e-06,
+      "loss": 0.0458,
+      "step": 682
+    },
+    {
+      "epoch": 0.1947533504419732,
+      "grad_norm": 1.6113043050375866,
+      "learning_rate": 9.306458378017004e-06,
+      "loss": 0.0678,
+      "step": 683
+    },
+    {
+      "epoch": 0.19503849443969204,
+      "grad_norm": 3.4623487395539505,
+      "learning_rate": 9.304109764593705e-06,
+      "loss": 0.0398,
+      "step": 684
+    },
+    {
+      "epoch": 0.19532363843741088,
+      "grad_norm": 2.3381560335801646,
+      "learning_rate": 9.301757478602886e-06,
+      "loss": 0.0445,
+      "step": 685
+    },
+    {
+      "epoch": 0.19560878243512975,
+      "grad_norm": 64.94755869755691,
+      "learning_rate": 9.299401522051685e-06,
+      "loss": 1.2332,
+      "step": 686
+    },
+    {
+      "epoch": 0.1958939264328486,
+      "grad_norm": 18.353718727926776,
+      "learning_rate": 9.297041896950365e-06,
+      "loss": 0.1454,
+      "step": 687
+    },
+    {
+      "epoch": 0.19617907043056743,
+      "grad_norm": 0.6617552332955904,
+      "learning_rate": 9.294678605312323e-06,
+      "loss": 0.0201,
+      "step": 688
+    },
+    {
+      "epoch": 0.1964642144282863,
+      "grad_norm": 1.1973919773785329,
+      "learning_rate": 9.292311649154088e-06,
+      "loss": 0.0133,
+      "step": 689
+    },
+    {
+      "epoch": 0.19674935842600513,
+      "grad_norm": 2.181331281045211,
+      "learning_rate": 9.289941030495313e-06,
+      "loss": 0.0279,
+      "step": 690
+    },
+    {
+      "epoch": 0.19703450242372397,
+      "grad_norm": 11.102336408880397,
+      "learning_rate": 9.287566751358773e-06,
+      "loss": 0.0318,
+      "step": 691
+    },
+    {
+      "epoch": 0.19731964642144284,
+      "grad_norm": 0.9201931408854377,
+      "learning_rate": 9.285188813770368e-06,
+      "loss": 0.025,
+      "step": 692
+    },
+    {
+      "epoch": 0.19760479041916168,
+      "grad_norm": 1.508525890818712,
+      "learning_rate": 9.282807219759123e-06,
+      "loss": 0.0414,
+      "step": 693
+    },
+    {
+      "epoch": 0.19788993441688052,
+      "grad_norm": 2.464653813173033,
+      "learning_rate": 9.280421971357181e-06,
+      "loss": 0.0389,
+      "step": 694
+    },
+    {
+      "epoch": 0.19817507841459936,
+      "grad_norm": 0.8743425835485196,
+      "learning_rate": 9.2780330705998e-06,
+      "loss": 0.0165,
+      "step": 695
+    },
+    {
+      "epoch": 0.19846022241231823,
+      "grad_norm": 1.8205084856824216,
+      "learning_rate": 9.27564051952536e-06,
+      "loss": 0.0286,
+      "step": 696
+    },
+    {
+      "epoch": 0.19874536641003707,
+      "grad_norm": 1.3224132996211422,
+      "learning_rate": 9.273244320175352e-06,
+      "loss": 0.0333,
+      "step": 697
+    },
+    {
+      "epoch": 0.1990305104077559,
+      "grad_norm": 1.0636884242784166,
+      "learning_rate": 9.270844474594381e-06,
+      "loss": 0.0165,
+      "step": 698
+    },
+    {
+      "epoch": 0.19931565440547477,
+      "grad_norm": 1.0517832694626335,
+      "learning_rate": 9.268440984830163e-06,
+      "loss": 0.0186,
+      "step": 699
+    },
+    {
+      "epoch": 0.1996007984031936,
+      "grad_norm": 1.0627875282801584,
+      "learning_rate": 9.266033852933525e-06,
+      "loss": 0.0299,
+      "step": 700
+    },
+    {
+      "epoch": 0.19988594240091245,
+      "grad_norm": 0.7272225161607353,
+      "learning_rate": 9.263623080958398e-06,
+      "loss": 0.0062,
+      "step": 701
+    },
+    {
+      "epoch": 0.20017108639863132,
+      "grad_norm": 1.847720921414192,
+      "learning_rate": 9.26120867096182e-06,
+      "loss": 0.0506,
+      "step": 702
+    },
+    {
+      "epoch": 0.20045623039635016,
+      "grad_norm": 1.6823362190282132,
+      "learning_rate": 9.258790625003939e-06,
+      "loss": 0.0653,
+      "step": 703
+    },
+    {
+      "epoch": 0.200741374394069,
+      "grad_norm": 1.4787235791879039,
+      "learning_rate": 9.256368945147998e-06,
+      "loss": 0.0538,
+      "step": 704
+    },
+    {
+      "epoch": 0.20102651839178784,
+      "grad_norm": 1.2040529131598932,
+      "learning_rate": 9.253943633460344e-06,
+      "loss": 0.0175,
+      "step": 705
+    },
+    {
+      "epoch": 0.2013116623895067,
+      "grad_norm": 1.6910714189347542,
+      "learning_rate": 9.251514692010423e-06,
+      "loss": 0.0349,
+      "step": 706
+    },
+    {
+      "epoch": 0.20159680638722555,
+      "grad_norm": 1.472948498746407,
+      "learning_rate": 9.249082122870779e-06,
+      "loss": 0.0369,
+      "step": 707
+    },
+    {
+      "epoch": 0.2018819503849444,
+      "grad_norm": 0.8762692780359592,
+      "learning_rate": 9.246645928117047e-06,
+      "loss": 0.0224,
+      "step": 708
+    },
+    {
+      "epoch": 0.20216709438266325,
+      "grad_norm": 2.4349769570385162,
+      "learning_rate": 9.24420610982796e-06,
+      "loss": 0.0254,
+      "step": 709
+    },
+    {
+      "epoch": 0.2024522383803821,
+      "grad_norm": 1.9112438219894465,
+      "learning_rate": 9.241762670085343e-06,
+      "loss": 0.0455,
+      "step": 710
+    },
+    {
+      "epoch": 0.20273738237810093,
+      "grad_norm": 2.434206504065968,
+      "learning_rate": 9.239315610974109e-06,
+      "loss": 0.0428,
+      "step": 711
+    },
+    {
+      "epoch": 0.2030225263758198,
+      "grad_norm": 1.6611178690442459,
+      "learning_rate": 9.236864934582259e-06,
+      "loss": 0.0311,
+      "step": 712
+    },
+    {
+      "epoch": 0.20330767037353864,
+      "grad_norm": 2.7886821146690237,
+      "learning_rate": 9.234410643000884e-06,
+      "loss": 0.0471,
+      "step": 713
+    },
+    {
+      "epoch": 0.20359281437125748,
+      "grad_norm": 1.7169627167940775,
+      "learning_rate": 9.231952738324155e-06,
+      "loss": 0.0302,
+      "step": 714
+    },
+    {
+      "epoch": 0.20387795836897632,
+      "grad_norm": 2.5621764097108652,
+      "learning_rate": 9.229491222649328e-06,
+      "loss": 0.0485,
+      "step": 715
+    },
+    {
+      "epoch": 0.2041631023666952,
+      "grad_norm": 1.17459682130259,
+      "learning_rate": 9.227026098076742e-06,
+      "loss": 0.0264,
+      "step": 716
+    },
+    {
+      "epoch": 0.20444824636441403,
+      "grad_norm": 1.3071201080016008,
+      "learning_rate": 9.224557366709813e-06,
+      "loss": 0.0255,
+      "step": 717
+    },
+    {
+      "epoch": 0.20473339036213287,
+      "grad_norm": 2.051560890960995,
+      "learning_rate": 9.222085030655035e-06,
+      "loss": 0.0476,
+      "step": 718
+    },
+    {
+      "epoch": 0.20501853435985173,
+      "grad_norm": 1.019980133767295,
+      "learning_rate": 9.219609092021976e-06,
+      "loss": 0.0202,
+      "step": 719
+    },
+    {
+      "epoch": 0.20530367835757057,
+      "grad_norm": 0.8053908676837997,
+      "learning_rate": 9.217129552923287e-06,
+      "loss": 0.0144,
+      "step": 720
+    },
+    {
+      "epoch": 0.2055888223552894,
+      "grad_norm": 0.44915310802273345,
+      "learning_rate": 9.214646415474676e-06,
+      "loss": 0.0188,
+      "step": 721
+    },
+    {
+      "epoch": 0.20587396635300828,
+      "grad_norm": 2.0651332669928664,
+      "learning_rate": 9.212159681794935e-06,
+      "loss": 0.0576,
+      "step": 722
+    },
+    {
+      "epoch": 0.20615911035072712,
+      "grad_norm": 1.1845182088251505,
+      "learning_rate": 9.209669354005915e-06,
+      "loss": 0.0251,
+      "step": 723
+    },
+    {
+      "epoch": 0.20644425434844596,
+      "grad_norm": 0.8224946140289474,
+      "learning_rate": 9.20717543423254e-06,
+      "loss": 0.0397,
+      "step": 724
+    },
+    {
+      "epoch": 0.2067293983461648,
+      "grad_norm": 0.922555862579926,
+      "learning_rate": 9.204677924602799e-06,
+      "loss": 0.0228,
+      "step": 725
+    },
+    {
+      "epoch": 0.20701454234388367,
+      "grad_norm": 0.7195816655824018,
+      "learning_rate": 9.202176827247739e-06,
+      "loss": 0.0234,
+      "step": 726
+    },
+    {
+      "epoch": 0.2072996863416025,
+      "grad_norm": 1.5777936992378956,
+      "learning_rate": 9.19967214430147e-06,
+      "loss": 0.047,
+      "step": 727
+    },
+    {
+      "epoch": 0.20758483033932135,
+      "grad_norm": 2.115593197700014,
+      "learning_rate": 9.197163877901167e-06,
+      "loss": 0.0804,
+      "step": 728
+    },
+    {
+      "epoch": 0.20786997433704021,
+      "grad_norm": 1.7316346105054923,
+      "learning_rate": 9.194652030187055e-06,
+      "loss": 0.0452,
+      "step": 729
+    },
+    {
+      "epoch": 0.20815511833475905,
+      "grad_norm": 1.732715865988259,
+      "learning_rate": 9.19213660330242e-06,
+      "loss": 0.0288,
+      "step": 730
+    },
+    {
+      "epoch": 0.2084402623324779,
+      "grad_norm": 2.9257895919417085,
+      "learning_rate": 9.1896175993936e-06,
+      "loss": 0.0551,
+      "step": 731
+    },
+    {
+      "epoch": 0.20872540633019676,
+      "grad_norm": 1.2005441423398024,
+      "learning_rate": 9.187095020609982e-06,
+      "loss": 0.0591,
+      "step": 732
+    },
+    {
+      "epoch": 0.2090105503279156,
+      "grad_norm": 1.1442567758861109,
+      "learning_rate": 9.18456886910401e-06,
+      "loss": 0.0316,
+      "step": 733
+    },
+    {
+      "epoch": 0.20929569432563444,
+      "grad_norm": 1.442435487183275,
+      "learning_rate": 9.182039147031174e-06,
+      "loss": 0.0357,
+      "step": 734
+    },
+    {
+      "epoch": 0.20958083832335328,
+      "grad_norm": 0.7389338008585332,
+      "learning_rate": 9.179505856550006e-06,
+      "loss": 0.0213,
+      "step": 735
+    },
+    {
+      "epoch": 0.20986598232107215,
+      "grad_norm": 0.5763570642711087,
+      "learning_rate": 9.176968999822091e-06,
+      "loss": 0.0261,
+      "step": 736
+    },
+    {
+      "epoch": 0.210151126318791,
+      "grad_norm": 2.9913704549741467,
+      "learning_rate": 9.174428579012051e-06,
+      "loss": 0.0466,
+      "step": 737
+    },
+    {
+      "epoch": 0.21043627031650983,
+      "grad_norm": 0.9514972565621266,
+      "learning_rate": 9.171884596287548e-06,
+      "loss": 0.0316,
+      "step": 738
+    },
+    {
+      "epoch": 0.2107214143142287,
+      "grad_norm": 1.1166267794394404,
+      "learning_rate": 9.16933705381929e-06,
+      "loss": 0.0425,
+      "step": 739
+    },
+    {
+      "epoch": 0.21100655831194753,
+      "grad_norm": 2.6475916702209794,
+      "learning_rate": 9.166785953781017e-06,
+      "loss": 0.0383,
+      "step": 740
+    },
+    {
+      "epoch": 0.21129170230966637,
+      "grad_norm": 0.5035517781871393,
+      "learning_rate": 9.164231298349505e-06,
+      "loss": 0.0116,
+      "step": 741
+    },
+    {
+      "epoch": 0.21157684630738524,
+      "grad_norm": 2.735386210598088,
+      "learning_rate": 9.161673089704565e-06,
+      "loss": 0.0439,
+      "step": 742
+    },
+    {
+      "epoch": 0.21186199030510408,
+      "grad_norm": 1.7132023378847019,
+      "learning_rate": 9.159111330029041e-06,
+      "loss": 0.0428,
+      "step": 743
+    },
+    {
+      "epoch": 0.21214713430282292,
+      "grad_norm": 1.2102951850569579,
+      "learning_rate": 9.156546021508803e-06,
+      "loss": 0.0181,
+      "step": 744
+    },
+    {
+      "epoch": 0.21243227830054176,
+      "grad_norm": 1.822168441101455,
+      "learning_rate": 9.153977166332756e-06,
+      "loss": 0.038,
+      "step": 745
+    },
+    {
+      "epoch": 0.21271742229826063,
+      "grad_norm": 3.1595776380649334,
+      "learning_rate": 9.151404766692822e-06,
+      "loss": 0.0676,
+      "step": 746
+    },
+    {
+      "epoch": 0.21300256629597947,
+      "grad_norm": 2.4059297211150947,
+      "learning_rate": 9.148828824783956e-06,
+      "loss": 0.0418,
+      "step": 747
+    },
+    {
+      "epoch": 0.2132877102936983,
+      "grad_norm": 1.1863131598693795,
+      "learning_rate": 9.146249342804128e-06,
+      "loss": 0.0179,
+      "step": 748
+    },
+    {
+      "epoch": 0.21357285429141717,
+      "grad_norm": 1.6629250296893985,
+      "learning_rate": 9.143666322954336e-06,
+      "loss": 0.0321,
+      "step": 749
+    },
+    {
+      "epoch": 0.21385799828913601,
+      "grad_norm": 0.702256216039878,
+      "learning_rate": 9.141079767438592e-06,
+      "loss": 0.0272,
+      "step": 750
+    },
+    {
+      "epoch": 0.21414314228685485,
+      "grad_norm": 1.0421114336526116,
+      "learning_rate": 9.138489678463927e-06,
+      "loss": 0.0288,
+      "step": 751
+    },
+    {
+      "epoch": 0.21442828628457372,
+      "grad_norm": 1.398071589217176,
+      "learning_rate": 9.135896058240384e-06,
+      "loss": 0.0164,
+      "step": 752
+    },
+    {
+      "epoch": 0.21471343028229256,
+      "grad_norm": 1.180627226683535,
+      "learning_rate": 9.133298908981021e-06,
+      "loss": 0.0302,
+      "step": 753
+    },
+    {
+      "epoch": 0.2149985742800114,
+      "grad_norm": 2.12849533813395,
+      "learning_rate": 9.13069823290191e-06,
+      "loss": 0.034,
+      "step": 754
+    },
+    {
+      "epoch": 0.21528371827773024,
+      "grad_norm": 1.1520445091389948,
+      "learning_rate": 9.12809403222213e-06,
+      "loss": 0.02,
+      "step": 755
+    },
+    {
+      "epoch": 0.2155688622754491,
+      "grad_norm": 0.9314697391415853,
+      "learning_rate": 9.125486309163764e-06,
+      "loss": 0.022,
+      "step": 756
+    },
+    {
+      "epoch": 0.21585400627316795,
+      "grad_norm": 0.4807671860197455,
+      "learning_rate": 9.122875065951907e-06,
+      "loss": 0.0129,
+      "step": 757
+    },
+    {
+      "epoch": 0.2161391502708868,
+      "grad_norm": 1.0284661378275857,
+      "learning_rate": 9.12026030481465e-06,
+      "loss": 0.0198,
+      "step": 758
+    },
+    {
+      "epoch": 0.21642429426860565,
+      "grad_norm": 1.3711724506820158,
+      "learning_rate": 9.117642027983096e-06,
+      "loss": 0.0275,
+      "step": 759
+    },
+    {
+      "epoch": 0.2167094382663245,
+      "grad_norm": 1.2056627878364008,
+      "learning_rate": 9.115020237691336e-06,
+      "loss": 0.0233,
+      "step": 760
+    },
+    {
+      "epoch": 0.21699458226404333,
+      "grad_norm": 1.5615068937947674,
+      "learning_rate": 9.11239493617647e-06,
+      "loss": 0.0472,
+      "step": 761
+    },
+    {
+      "epoch": 0.2172797262617622,
+      "grad_norm": 1.053111712057514,
+      "learning_rate": 9.109766125678585e-06,
+      "loss": 0.0236,
+      "step": 762
+    },
+    {
+      "epoch": 0.21756487025948104,
+      "grad_norm": 1.492926709547356,
+      "learning_rate": 9.107133808440767e-06,
+      "loss": 0.0278,
+      "step": 763
+    },
+    {
+      "epoch": 0.21785001425719988,
+      "grad_norm": 0.7321893976881936,
+      "learning_rate": 9.104497986709096e-06,
+      "loss": 0.011,
+      "step": 764
+    },
+    {
+      "epoch": 0.21813515825491872,
+      "grad_norm": 0.972053791654552,
+      "learning_rate": 9.101858662732635e-06,
+      "loss": 0.0284,
+      "step": 765
+    },
+    {
+      "epoch": 0.2184203022526376,
+      "grad_norm": 0.384726545541271,
+      "learning_rate": 9.099215838763444e-06,
+      "loss": 0.0091,
+      "step": 766
+    },
+    {
+      "epoch": 0.21870544625035643,
+      "grad_norm": 1.4964596931892333,
+      "learning_rate": 9.096569517056562e-06,
+      "loss": 0.0155,
+      "step": 767
+    },
+    {
+      "epoch": 0.21899059024807527,
+      "grad_norm": 1.4119719189435824,
+      "learning_rate": 9.093919699870017e-06,
+      "loss": 0.0278,
+      "step": 768
+    },
+    {
+      "epoch": 0.21927573424579413,
+      "grad_norm": 2.196204389526829,
+      "learning_rate": 9.091266389464818e-06,
+      "loss": 0.0621,
+      "step": 769
+    },
+    {
+      "epoch": 0.21956087824351297,
+      "grad_norm": 1.437174200010841,
+      "learning_rate": 9.088609588104958e-06,
+      "loss": 0.0207,
+      "step": 770
+    },
+    {
+      "epoch": 0.21984602224123181,
+      "grad_norm": 2.1986672533499307,
+      "learning_rate": 9.085949298057402e-06,
+      "loss": 0.0357,
+      "step": 771
+    },
+    {
+      "epoch": 0.22013116623895068,
+      "grad_norm": 2.0638899957028642,
+      "learning_rate": 9.083285521592097e-06,
+      "loss": 0.0473,
+      "step": 772
+    },
+    {
+      "epoch": 0.22041631023666952,
+      "grad_norm": 1.1744345395272808,
+      "learning_rate": 9.080618260981964e-06,
+      "loss": 0.0155,
+      "step": 773
+    },
+    {
+      "epoch": 0.22070145423438836,
+      "grad_norm": 2.306169669849833,
+      "learning_rate": 9.077947518502894e-06,
+      "loss": 0.033,
+      "step": 774
+    },
+    {
+      "epoch": 0.2209865982321072,
+      "grad_norm": 1.9124745318350345,
+      "learning_rate": 9.075273296433753e-06,
+      "loss": 0.0379,
+      "step": 775
+    },
+    {
+      "epoch": 0.22127174222982607,
+      "grad_norm": 2.0494348960481785,
+      "learning_rate": 9.072595597056375e-06,
+      "loss": 0.0671,
+      "step": 776
+    },
+    {
+      "epoch": 0.2215568862275449,
+      "grad_norm": 1.6680009681455517,
+      "learning_rate": 9.069914422655559e-06,
+      "loss": 0.0301,
+      "step": 777
+    },
+    {
+      "epoch": 0.22184203022526375,
+      "grad_norm": 1.6771017256370313,
+      "learning_rate": 9.067229775519071e-06,
+      "loss": 0.0229,
+      "step": 778
+    },
+    {
+      "epoch": 0.22212717422298262,
+      "grad_norm": 2.0275590166360806,
+      "learning_rate": 9.064541657937641e-06,
+      "loss": 0.0685,
+      "step": 779
+    },
+    {
+      "epoch": 0.22241231822070145,
+      "grad_norm": 2.360895367226193,
+      "learning_rate": 9.061850072204958e-06,
+      "loss": 0.0299,
+      "step": 780
+    },
+    {
+      "epoch": 0.2226974622184203,
+      "grad_norm": 1.5539467978062291,
+      "learning_rate": 9.05915502061767e-06,
+      "loss": 0.0403,
+      "step": 781
+    },
+    {
+      "epoch": 0.22298260621613916,
+      "grad_norm": 0.7670508938888166,
+      "learning_rate": 9.056456505475385e-06,
+      "loss": 0.0105,
+      "step": 782
+    },
+    {
+      "epoch": 0.223267750213858,
+      "grad_norm": 0.7937880675570071,
+      "learning_rate": 9.053754529080664e-06,
+      "loss": 0.0176,
+      "step": 783
+    },
+    {
+      "epoch": 0.22355289421157684,
+      "grad_norm": 2.157831660066521,
+      "learning_rate": 9.051049093739023e-06,
+      "loss": 0.0336,
+      "step": 784
+    },
+    {
+      "epoch": 0.22383803820929568,
+      "grad_norm": 1.1663071318976364,
+      "learning_rate": 9.048340201758929e-06,
+      "loss": 0.0252,
+      "step": 785
+    },
+    {
+      "epoch": 0.22412318220701455,
+      "grad_norm": 1.292838079011928,
+      "learning_rate": 9.045627855451797e-06,
+      "loss": 0.0363,
+      "step": 786
+    },
+    {
+      "epoch": 0.2244083262047334,
+      "grad_norm": 1.5733550446218494,
+      "learning_rate": 9.04291205713199e-06,
+      "loss": 0.0213,
+      "step": 787
+    },
+    {
+      "epoch": 0.22469347020245223,
+      "grad_norm": 1.1259643693921413,
+      "learning_rate": 9.04019280911682e-06,
+      "loss": 0.0149,
+      "step": 788
+    },
+    {
+      "epoch": 0.2249786142001711,
+      "grad_norm": 1.0083139262164555,
+      "learning_rate": 9.037470113726537e-06,
+      "loss": 0.0172,
+      "step": 789
+    },
+    {
+      "epoch": 0.22526375819788994,
+      "grad_norm": 1.3511474029563015,
+      "learning_rate": 9.034743973284337e-06,
+      "loss": 0.0215,
+      "step": 790
+    },
+    {
+      "epoch": 0.22554890219560877,
+      "grad_norm": 0.8363495828122983,
+      "learning_rate": 9.032014390116351e-06,
+      "loss": 0.0266,
+      "step": 791
+    },
+    {
+      "epoch": 0.22583404619332764,
+      "grad_norm": 3.1833632224433996,
+      "learning_rate": 9.029281366551654e-06,
+      "loss": 0.0254,
+      "step": 792
+    },
+    {
+      "epoch": 0.22611919019104648,
+      "grad_norm": 2.1181522596055435,
+      "learning_rate": 9.02654490492225e-06,
+      "loss": 0.034,
+      "step": 793
+    },
+    {
+      "epoch": 0.22640433418876532,
+      "grad_norm": 1.125558580379027,
+      "learning_rate": 9.02380500756308e-06,
+      "loss": 0.0244,
+      "step": 794
+    },
+    {
+      "epoch": 0.22668947818648416,
+      "grad_norm": 1.2351475059799766,
+      "learning_rate": 9.021061676812016e-06,
+      "loss": 0.0175,
+      "step": 795
+    },
+    {
+      "epoch": 0.22697462218420303,
+      "grad_norm": 2.0823417135384426,
+      "learning_rate": 9.01831491500986e-06,
+      "loss": 0.0314,
+      "step": 796
+    },
+    {
+      "epoch": 0.22725976618192187,
+      "grad_norm": 2.6273804667622955,
+      "learning_rate": 9.015564724500343e-06,
+      "loss": 0.0439,
+      "step": 797
+    },
+    {
+      "epoch": 0.2275449101796407,
+      "grad_norm": 0.5129484219330254,
+      "learning_rate": 9.012811107630118e-06,
+      "loss": 0.0101,
+      "step": 798
+    },
+    {
+      "epoch": 0.22783005417735958,
+      "grad_norm": 1.1938003324425006,
+      "learning_rate": 9.010054066748764e-06,
+      "loss": 0.0379,
+      "step": 799
+    },
+    {
+      "epoch": 0.22811519817507842,
+      "grad_norm": 3.097898875063007,
+      "learning_rate": 9.00729360420878e-06,
+      "loss": 0.0781,
+      "step": 800
+    },
+    {
+      "epoch": 0.22840034217279725,
+      "grad_norm": 1.0145281059549673,
+      "learning_rate": 9.004529722365585e-06,
+      "loss": 0.0137,
+      "step": 801
+    },
+    {
+      "epoch": 0.22868548617051612,
+      "grad_norm": 0.9369644953910446,
+      "learning_rate": 9.001762423577521e-06,
+      "loss": 0.0129,
+      "step": 802
+    },
+    {
+      "epoch": 0.22897063016823496,
+      "grad_norm": 1.6897375760595272,
+      "learning_rate": 8.998991710205837e-06,
+      "loss": 0.0358,
+      "step": 803
+    },
+    {
+      "epoch": 0.2292557741659538,
+      "grad_norm": 1.5083615821843614,
+      "learning_rate": 8.996217584614702e-06,
+      "loss": 0.0299,
+      "step": 804
+    },
+    {
+      "epoch": 0.22954091816367264,
+      "grad_norm": 1.5739940630915592,
+      "learning_rate": 8.99344004917119e-06,
+      "loss": 0.0256,
+      "step": 805
+    },
+    {
+      "epoch": 0.2298260621613915,
+      "grad_norm": 3.8075761182654593,
+      "learning_rate": 8.990659106245292e-06,
+      "loss": 0.0503,
+      "step": 806
+    },
+    {
+      "epoch": 0.23011120615911035,
+      "grad_norm": 0.679718812603675,
+      "learning_rate": 8.9878747582099e-06,
+      "loss": 0.0152,
+      "step": 807
+    },
+    {
+      "epoch": 0.2303963501568292,
+      "grad_norm": 1.6636057899743955,
+      "learning_rate": 8.98508700744082e-06,
+      "loss": 0.0236,
+      "step": 808
+    },
+    {
+      "epoch": 0.23068149415454806,
+      "grad_norm": 1.0031959304118039,
+      "learning_rate": 8.98229585631675e-06,
+      "loss": 0.0223,
+      "step": 809
+    },
+    {
+      "epoch": 0.2309666381522669,
+      "grad_norm": 0.49804281066136624,
+      "learning_rate": 8.979501307219298e-06,
+      "loss": 0.0126,
+      "step": 810
+    },
+    {
+      "epoch": 0.23125178214998574,
+      "grad_norm": 2.390993893561002,
+      "learning_rate": 8.976703362532971e-06,
+      "loss": 0.0316,
+      "step": 811
+    },
+    {
+      "epoch": 0.2315369261477046,
+      "grad_norm": 1.9859352958275294,
+      "learning_rate": 8.973902024645165e-06,
+      "loss": 0.0164,
+      "step": 812
+    },
+    {
+      "epoch": 0.23182207014542344,
+      "grad_norm": 2.609134116114303,
+      "learning_rate": 8.971097295946183e-06,
+      "loss": 0.0511,
+      "step": 813
+    },
+    {
+      "epoch": 0.23210721414314228,
+      "grad_norm": 1.3455604419784448,
+      "learning_rate": 8.968289178829214e-06,
+      "loss": 0.0232,
+      "step": 814
+    },
+    {
+      "epoch": 0.23239235814086112,
+      "grad_norm": 2.843848539977979,
+      "learning_rate": 8.96547767569034e-06,
+      "loss": 0.0263,
+      "step": 815
+    },
+    {
+      "epoch": 0.23267750213858,
+      "grad_norm": 3.7860602043709983,
+      "learning_rate": 8.962662788928531e-06,
+      "loss": 0.0428,
+      "step": 816
+    },
+    {
+      "epoch": 0.23296264613629883,
+      "grad_norm": 2.4050408956018763,
+      "learning_rate": 8.959844520945646e-06,
+      "loss": 0.0357,
+      "step": 817
+    },
+    {
+      "epoch": 0.23324779013401767,
+      "grad_norm": 2.7114207248581113,
+      "learning_rate": 8.957022874146429e-06,
+      "loss": 0.0348,
+      "step": 818
+    },
+    {
+      "epoch": 0.23353293413173654,
+      "grad_norm": 2.5893678936272444,
+      "learning_rate": 8.954197850938506e-06,
+      "loss": 0.0257,
+      "step": 819
+    },
+    {
+      "epoch": 0.23381807812945538,
+      "grad_norm": 1.1802173880782538,
+      "learning_rate": 8.951369453732386e-06,
+      "loss": 0.0354,
+      "step": 820
+    },
+    {
+      "epoch": 0.23410322212717422,
+      "grad_norm": 0.5538609744161046,
+      "learning_rate": 8.948537684941452e-06,
+      "loss": 0.0121,
+      "step": 821
+    },
+    {
+      "epoch": 0.23438836612489308,
+      "grad_norm": 0.7526259552569778,
+      "learning_rate": 8.94570254698197e-06,
+      "loss": 0.0348,
+      "step": 822
+    },
+    {
+      "epoch": 0.23467351012261192,
+      "grad_norm": 0.38116984197805204,
+      "learning_rate": 8.942864042273075e-06,
+      "loss": 0.0036,
+      "step": 823
+    },
+    {
+      "epoch": 0.23495865412033076,
+      "grad_norm": 1.31857014571736,
+      "learning_rate": 8.94002217323678e-06,
+      "loss": 0.0163,
+      "step": 824
+    },
+    {
+      "epoch": 0.2352437981180496,
+      "grad_norm": 2.921327945172875,
+      "learning_rate": 8.937176942297968e-06,
+      "loss": 0.0399,
+      "step": 825
+    },
+    {
+      "epoch": 0.23552894211576847,
+      "grad_norm": 1.5144558599662687,
+      "learning_rate": 8.934328351884386e-06,
+      "loss": 0.0382,
+      "step": 826
+    },
+    {
+      "epoch": 0.2358140861134873,
+      "grad_norm": 1.2344815677516623,
+      "learning_rate": 8.931476404426653e-06,
+      "loss": 0.0332,
+      "step": 827
+    },
+    {
+      "epoch": 0.23609923011120615,
+      "grad_norm": 0.7033882183737051,
+      "learning_rate": 8.928621102358248e-06,
+      "loss": 0.0118,
+      "step": 828
+    },
+    {
+      "epoch": 0.23638437410892502,
+      "grad_norm": 6.45022989283852,
+      "learning_rate": 8.925762448115516e-06,
+      "loss": 0.0202,
+      "step": 829
+    },
+    {
+      "epoch": 0.23666951810664386,
+      "grad_norm": 2.5229595774880087,
+      "learning_rate": 8.92290044413766e-06,
+      "loss": 0.0368,
+      "step": 830
+    },
+    {
+      "epoch": 0.2369546621043627,
+      "grad_norm": 1.258824021540127,
+      "learning_rate": 8.92003509286674e-06,
+      "loss": 0.0202,
+      "step": 831
+    },
+    {
+      "epoch": 0.23723980610208156,
+      "grad_norm": 2.6905988020090956,
+      "learning_rate": 8.917166396747681e-06,
+      "loss": 0.0397,
+      "step": 832
+    },
+    {
+      "epoch": 0.2375249500998004,
+      "grad_norm": 1.4334425139307863,
+      "learning_rate": 8.914294358228245e-06,
+      "loss": 0.0233,
+      "step": 833
+    },
+    {
+      "epoch": 0.23781009409751924,
+      "grad_norm": 3.3639351642610373,
+      "learning_rate": 8.911418979759066e-06,
+      "loss": 0.035,
+      "step": 834
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 1.0725408705087607,
+      "learning_rate": 8.908540263793611e-06,
+      "loss": 0.0357,
+      "step": 835
+    },
+    {
+      "epoch": 0.23838038209295695,
+      "grad_norm": 0.8357183521879652,
+      "learning_rate": 8.905658212788207e-06,
+      "loss": 0.0134,
+      "step": 836
+    },
+    {
+      "epoch": 0.2386655260906758,
+      "grad_norm": 2.0411720213460347,
+      "learning_rate": 8.902772829202015e-06,
+      "loss": 0.0561,
+      "step": 837
+    },
+    {
+      "epoch": 0.23895067008839463,
+      "grad_norm": 0.4962525894662451,
+      "learning_rate": 8.899884115497053e-06,
+      "loss": 0.0089,
+      "step": 838
+    },
+    {
+      "epoch": 0.2392358140861135,
+      "grad_norm": 5.584077909965726,
+      "learning_rate": 8.896992074138171e-06,
+      "loss": 0.0478,
+      "step": 839
+    },
+    {
+      "epoch": 0.23952095808383234,
+      "grad_norm": 1.8286855246809188,
+      "learning_rate": 8.89409670759306e-06,
+      "loss": 0.0356,
+      "step": 840
+    },
+    {
+      "epoch": 0.23980610208155118,
+      "grad_norm": 0.7242311973923197,
+      "learning_rate": 8.89119801833225e-06,
+      "loss": 0.0196,
+      "step": 841
+    },
+    {
+      "epoch": 0.24009124607927004,
+      "grad_norm": 1.6524840292900185,
+      "learning_rate": 8.888296008829106e-06,
+      "loss": 0.0313,
+      "step": 842
+    },
+    {
+      "epoch": 0.24037639007698888,
+      "grad_norm": 1.6025270226398667,
+      "learning_rate": 8.885390681559822e-06,
+      "loss": 0.0589,
+      "step": 843
+    },
+    {
+      "epoch": 0.24066153407470772,
+      "grad_norm": 0.6828303452986196,
+      "learning_rate": 8.88248203900343e-06,
+      "loss": 0.0129,
+      "step": 844
+    },
+    {
+      "epoch": 0.24094667807242656,
+      "grad_norm": 1.0847428212810786,
+      "learning_rate": 8.879570083641788e-06,
+      "loss": 0.0169,
+      "step": 845
+    },
+    {
+      "epoch": 0.24123182207014543,
+      "grad_norm": 1.3682289401821128,
+      "learning_rate": 8.87665481795958e-06,
+      "loss": 0.0253,
+      "step": 846
+    },
+    {
+      "epoch": 0.24151696606786427,
+      "grad_norm": 3.7461358288629434,
+      "learning_rate": 8.873736244444311e-06,
+      "loss": 0.0483,
+      "step": 847
+    },
+    {
+      "epoch": 0.2418021100655831,
+      "grad_norm": 1.9237644513844145,
+      "learning_rate": 8.870814365586315e-06,
+      "loss": 0.0205,
+      "step": 848
+    },
+    {
+      "epoch": 0.24208725406330198,
+      "grad_norm": 1.2152994567992417,
+      "learning_rate": 8.867889183878742e-06,
+      "loss": 0.0268,
+      "step": 849
+    },
+    {
+      "epoch": 0.24237239806102082,
+      "grad_norm": 1.91329866781686,
+      "learning_rate": 8.864960701817564e-06,
+      "loss": 0.0178,
+      "step": 850
+    },
+    {
+      "epoch": 0.24265754205873966,
+      "grad_norm": 1.9431409917366391,
+      "learning_rate": 8.862028921901563e-06,
+      "loss": 0.033,
+      "step": 851
+    },
+    {
+      "epoch": 0.24294268605645852,
+      "grad_norm": 1.2283113327719135,
+      "learning_rate": 8.859093846632343e-06,
+      "loss": 0.0295,
+      "step": 852
+    },
+    {
+      "epoch": 0.24322783005417736,
+      "grad_norm": 0.9973988056561106,
+      "learning_rate": 8.856155478514313e-06,
+      "loss": 0.0212,
+      "step": 853
+    },
+    {
+      "epoch": 0.2435129740518962,
+      "grad_norm": 2.4421795168152753,
+      "learning_rate": 8.853213820054693e-06,
+      "loss": 0.0172,
+      "step": 854
+    },
+    {
+      "epoch": 0.24379811804961504,
+      "grad_norm": 0.6853186509231574,
+      "learning_rate": 8.850268873763514e-06,
+      "loss": 0.0071,
+      "step": 855
+    },
+    {
+      "epoch": 0.2440832620473339,
+      "grad_norm": 0.20714634708932925,
+      "learning_rate": 8.84732064215361e-06,
+      "loss": 0.0081,
+      "step": 856
+    },
+    {
+      "epoch": 0.24436840604505275,
+      "grad_norm": 1.4038906766225647,
+      "learning_rate": 8.844369127740617e-06,
+      "loss": 0.0591,
+      "step": 857
+    },
+    {
+      "epoch": 0.2446535500427716,
+      "grad_norm": 3.2053223620330944,
+      "learning_rate": 8.841414333042975e-06,
+      "loss": 0.0932,
+      "step": 858
+    },
+    {
+      "epoch": 0.24493869404049046,
+      "grad_norm": 0.7632677420756766,
+      "learning_rate": 8.83845626058192e-06,
+      "loss": 0.016,
+      "step": 859
+    },
+    {
+      "epoch": 0.2452238380382093,
+      "grad_norm": 1.4585990037580068,
+      "learning_rate": 8.835494912881487e-06,
+      "loss": 0.0264,
+      "step": 860
+    },
+    {
+      "epoch": 0.24550898203592814,
+      "grad_norm": 0.9392293367012389,
+      "learning_rate": 8.832530292468509e-06,
+      "loss": 0.0176,
+      "step": 861
+    },
+    {
+      "epoch": 0.245794126033647,
+      "grad_norm": 0.9313249492245359,
+      "learning_rate": 8.8295624018726e-06,
+      "loss": 0.0239,
+      "step": 862
+    },
+    {
+      "epoch": 0.24607927003136584,
+      "grad_norm": 1.8503666333746411,
+      "learning_rate": 8.826591243626178e-06,
+      "loss": 0.0468,
+      "step": 863
+    },
+    {
+      "epoch": 0.24636441402908468,
+      "grad_norm": 1.5599510249156554,
+      "learning_rate": 8.82361682026444e-06,
+      "loss": 0.0466,
+      "step": 864
+    },
+    {
+      "epoch": 0.24664955802680352,
+      "grad_norm": 0.7932008240844486,
+      "learning_rate": 8.820639134325371e-06,
+      "loss": 0.0174,
+      "step": 865
+    },
+    {
+      "epoch": 0.2469347020245224,
+      "grad_norm": 1.7519462558310501,
+      "learning_rate": 8.817658188349745e-06,
+      "loss": 0.0308,
+      "step": 866
+    },
+    {
+      "epoch": 0.24721984602224123,
+      "grad_norm": 0.7374567291630424,
+      "learning_rate": 8.81467398488111e-06,
+      "loss": 0.0268,
+      "step": 867
+    },
+    {
+      "epoch": 0.24750499001996007,
+      "grad_norm": 2.2486150692979185,
+      "learning_rate": 8.811686526465799e-06,
+      "loss": 0.0364,
+      "step": 868
+    },
+    {
+      "epoch": 0.24779013401767894,
+      "grad_norm": 1.743439371955327,
+      "learning_rate": 8.808695815652922e-06,
+      "loss": 0.0414,
+      "step": 869
+    },
+    {
+      "epoch": 0.24807527801539778,
+      "grad_norm": 1.0726966014600394,
+      "learning_rate": 8.805701854994358e-06,
+      "loss": 0.0208,
+      "step": 870
+    },
+    {
+      "epoch": 0.24836042201311662,
+      "grad_norm": 2.634942103615967,
+      "learning_rate": 8.802704647044766e-06,
+      "loss": 0.0335,
+      "step": 871
+    },
+    {
+      "epoch": 0.24864556601083548,
+      "grad_norm": 1.2580046690380893,
+      "learning_rate": 8.799704194361575e-06,
+      "loss": 0.0253,
+      "step": 872
+    },
+    {
+      "epoch": 0.24893071000855432,
+      "grad_norm": 1.096234047219748,
+      "learning_rate": 8.79670049950498e-06,
+      "loss": 0.0328,
+      "step": 873
+    },
+    {
+      "epoch": 0.24921585400627316,
+      "grad_norm": 1.547027261492061,
+      "learning_rate": 8.793693565037942e-06,
+      "loss": 0.034,
+      "step": 874
+    },
+    {
+      "epoch": 0.249500998003992,
+      "grad_norm": 0.5892064156881954,
+      "learning_rate": 8.790683393526192e-06,
+      "loss": 0.0164,
+      "step": 875
+    },
+    {
+      "epoch": 0.24978614200171087,
+      "grad_norm": 1.3901316717407268,
+      "learning_rate": 8.787669987538214e-06,
+      "loss": 0.0405,
+      "step": 876
+    },
+    {
+      "epoch": 0.25007128599942974,
+      "grad_norm": 1.7467544871345946,
+      "learning_rate": 8.784653349645259e-06,
+      "loss": 0.0464,
+      "step": 877
+    },
+    {
+      "epoch": 0.2503564299971486,
+      "grad_norm": 0.8752141247160915,
+      "learning_rate": 8.781633482421333e-06,
+      "loss": 0.0186,
+      "step": 878
+    },
+    {
+      "epoch": 0.2506415739948674,
+      "grad_norm": 2.4276853957481612,
+      "learning_rate": 8.7786103884432e-06,
+      "loss": 0.0584,
+      "step": 879
+    },
+    {
+      "epoch": 0.25092671799258626,
+      "grad_norm": 1.3072797801662137,
+      "learning_rate": 8.775584070290373e-06,
+      "loss": 0.0283,
+      "step": 880
+    },
+    {
+      "epoch": 0.2512118619903051,
+      "grad_norm": 0.676361478133077,
+      "learning_rate": 8.772554530545118e-06,
+      "loss": 0.0127,
+      "step": 881
+    },
+    {
+      "epoch": 0.25149700598802394,
+      "grad_norm": 2.1577548832377618,
+      "learning_rate": 8.769521771792453e-06,
+      "loss": 0.0389,
+      "step": 882
+    },
+    {
+      "epoch": 0.2517821499857428,
+      "grad_norm": 2.3856943630410496,
+      "learning_rate": 8.766485796620135e-06,
+      "loss": 0.0634,
+      "step": 883
+    },
+    {
+      "epoch": 0.25206729398346167,
+      "grad_norm": 1.0305070102089764,
+      "learning_rate": 8.763446607618675e-06,
+      "loss": 0.0202,
+      "step": 884
+    },
+    {
+      "epoch": 0.2523524379811805,
+      "grad_norm": 1.2830717231149895,
+      "learning_rate": 8.76040420738132e-06,
+      "loss": 0.0284,
+      "step": 885
+    },
+    {
+      "epoch": 0.25263758197889935,
+      "grad_norm": 1.6411073612068923,
+      "learning_rate": 8.75735859850406e-06,
+      "loss": 0.0219,
+      "step": 886
+    },
+    {
+      "epoch": 0.2529227259766182,
+      "grad_norm": 1.1756235902210506,
+      "learning_rate": 8.754309783585619e-06,
+      "loss": 0.0349,
+      "step": 887
+    },
+    {
+      "epoch": 0.25320786997433703,
+      "grad_norm": 0.6847813278483827,
+      "learning_rate": 8.751257765227462e-06,
+      "loss": 0.0206,
+      "step": 888
+    },
+    {
+      "epoch": 0.25349301397205587,
+      "grad_norm": 0.5892100476513646,
+      "learning_rate": 8.748202546033781e-06,
+      "loss": 0.0202,
+      "step": 889
+    },
+    {
+      "epoch": 0.2537781579697747,
+      "grad_norm": 0.6949316617452543,
+      "learning_rate": 8.745144128611506e-06,
+      "loss": 0.021,
+      "step": 890
+    },
+    {
+      "epoch": 0.2540633019674936,
+      "grad_norm": 0.9729222569070698,
+      "learning_rate": 8.742082515570291e-06,
+      "loss": 0.0194,
+      "step": 891
+    },
+    {
+      "epoch": 0.25434844596521244,
+      "grad_norm": 0.45568034642641875,
+      "learning_rate": 8.739017709522519e-06,
+      "loss": 0.0106,
+      "step": 892
+    },
+    {
+      "epoch": 0.2546335899629313,
+      "grad_norm": 1.2410486423220612,
+      "learning_rate": 8.735949713083295e-06,
+      "loss": 0.0118,
+      "step": 893
+    },
+    {
+      "epoch": 0.2549187339606501,
+      "grad_norm": 1.9597718656925054,
+      "learning_rate": 8.732878528870452e-06,
+      "loss": 0.0271,
+      "step": 894
+    },
+    {
+      "epoch": 0.25520387795836896,
+      "grad_norm": 1.7196417456122466,
+      "learning_rate": 8.729804159504537e-06,
+      "loss": 0.0498,
+      "step": 895
+    },
+    {
+      "epoch": 0.2554890219560878,
+      "grad_norm": 0.9101465781142756,
+      "learning_rate": 8.726726607608817e-06,
+      "loss": 0.009,
+      "step": 896
+    },
+    {
+      "epoch": 0.2557741659538067,
+      "grad_norm": 1.5268690830009926,
+      "learning_rate": 8.723645875809274e-06,
+      "loss": 0.0311,
+      "step": 897
+    },
+    {
+      "epoch": 0.25605930995152554,
+      "grad_norm": 1.2252517246259709,
+      "learning_rate": 8.720561966734604e-06,
+      "loss": 0.0102,
+      "step": 898
+    },
+    {
+      "epoch": 0.2563444539492444,
+      "grad_norm": 1.2280625351519197,
+      "learning_rate": 8.717474883016214e-06,
+      "loss": 0.0085,
+      "step": 899
+    },
+    {
+      "epoch": 0.2566295979469632,
+      "grad_norm": 3.117132439752605,
+      "learning_rate": 8.714384627288222e-06,
+      "loss": 0.0791,
+      "step": 900
+    },
+    {
+      "epoch": 0.25691474194468206,
+      "grad_norm": 1.4949161090600118,
+      "learning_rate": 8.711291202187447e-06,
+      "loss": 0.0322,
+      "step": 901
+    },
+    {
+      "epoch": 0.2571998859424009,
+      "grad_norm": 1.8302629323751376,
+      "learning_rate": 8.708194610353418e-06,
+      "loss": 0.0617,
+      "step": 902
+    },
+    {
+      "epoch": 0.25748502994011974,
+      "grad_norm": 2.308044393671751,
+      "learning_rate": 8.705094854428362e-06,
+      "loss": 0.0443,
+      "step": 903
+    },
+    {
+      "epoch": 0.25777017393783863,
+      "grad_norm": 1.6208323890738643,
+      "learning_rate": 8.701991937057211e-06,
+      "loss": 0.0229,
+      "step": 904
+    },
+    {
+      "epoch": 0.25805531793555747,
+      "grad_norm": 0.9063595773842802,
+      "learning_rate": 8.698885860887587e-06,
+      "loss": 0.0076,
+      "step": 905
+    },
+    {
+      "epoch": 0.2583404619332763,
+      "grad_norm": 0.7271393808586717,
+      "learning_rate": 8.695776628569813e-06,
+      "loss": 0.0305,
+      "step": 906
+    },
+    {
+      "epoch": 0.25862560593099515,
+      "grad_norm": 1.9625970380708422,
+      "learning_rate": 8.692664242756902e-06,
+      "loss": 0.0415,
+      "step": 907
+    },
+    {
+      "epoch": 0.258910749928714,
+      "grad_norm": 0.5666328525418538,
+      "learning_rate": 8.689548706104564e-06,
+      "loss": 0.0159,
+      "step": 908
+    },
+    {
+      "epoch": 0.25919589392643283,
+      "grad_norm": 2.8145375900053806,
+      "learning_rate": 8.68643002127119e-06,
+      "loss": 0.0574,
+      "step": 909
+    },
+    {
+      "epoch": 0.25948103792415167,
+      "grad_norm": 1.5275888603420062,
+      "learning_rate": 8.683308190917857e-06,
+      "loss": 0.0341,
+      "step": 910
+    },
+    {
+      "epoch": 0.25976618192187056,
+      "grad_norm": 3.1426880240693214,
+      "learning_rate": 8.680183217708334e-06,
+      "loss": 0.0549,
+      "step": 911
+    },
+    {
+      "epoch": 0.2600513259195894,
+      "grad_norm": 0.7773518110968769,
+      "learning_rate": 8.677055104309062e-06,
+      "loss": 0.0145,
+      "step": 912
+    },
+    {
+      "epoch": 0.26033646991730824,
+      "grad_norm": 1.9823459424727548,
+      "learning_rate": 8.673923853389172e-06,
+      "loss": 0.0437,
+      "step": 913
+    },
+    {
+      "epoch": 0.2606216139150271,
+      "grad_norm": 1.1164470316091455,
+      "learning_rate": 8.670789467620461e-06,
+      "loss": 0.0231,
+      "step": 914
+    },
+    {
+      "epoch": 0.2609067579127459,
+      "grad_norm": 0.6832591626715077,
+      "learning_rate": 8.667651949677409e-06,
+      "loss": 0.0163,
+      "step": 915
+    },
+    {
+      "epoch": 0.26119190191046476,
+      "grad_norm": 1.876547933242045,
+      "learning_rate": 8.664511302237164e-06,
+      "loss": 0.0279,
+      "step": 916
+    },
+    {
+      "epoch": 0.26147704590818366,
+      "grad_norm": 1.152597026479379,
+      "learning_rate": 8.661367527979547e-06,
+      "loss": 0.0161,
+      "step": 917
+    },
+    {
+      "epoch": 0.2617621899059025,
+      "grad_norm": 1.9164628134925323,
+      "learning_rate": 8.658220629587046e-06,
+      "loss": 0.0457,
+      "step": 918
+    },
+    {
+      "epoch": 0.26204733390362134,
+      "grad_norm": 1.5725610393465852,
+      "learning_rate": 8.655070609744816e-06,
+      "loss": 0.0271,
+      "step": 919
+    },
+    {
+      "epoch": 0.2623324779013402,
+      "grad_norm": 1.6679600131417485,
+      "learning_rate": 8.651917471140673e-06,
+      "loss": 0.0562,
+      "step": 920
+    },
+    {
+      "epoch": 0.262617621899059,
+      "grad_norm": 1.7758731689450964,
+      "learning_rate": 8.648761216465096e-06,
+      "loss": 0.0299,
+      "step": 921
+    },
+    {
+      "epoch": 0.26290276589677786,
+      "grad_norm": 1.3868053732760177,
+      "learning_rate": 8.64560184841122e-06,
+      "loss": 0.0178,
+      "step": 922
+    },
+    {
+      "epoch": 0.2631879098944967,
+      "grad_norm": 1.9904157646953466,
+      "learning_rate": 8.642439369674845e-06,
+      "loss": 0.0239,
+      "step": 923
+    },
+    {
+      "epoch": 0.2634730538922156,
+      "grad_norm": 1.7623169430202168,
+      "learning_rate": 8.639273782954412e-06,
+      "loss": 0.0263,
+      "step": 924
+    },
+    {
+      "epoch": 0.26375819788993443,
+      "grad_norm": 1.0216670315613425,
+      "learning_rate": 8.636105090951022e-06,
+      "loss": 0.0286,
+      "step": 925
+    },
+    {
+      "epoch": 0.26404334188765327,
+      "grad_norm": 1.0887871495489123,
+      "learning_rate": 8.63293329636843e-06,
+      "loss": 0.0395,
+      "step": 926
+    },
+    {
+      "epoch": 0.2643284858853721,
+      "grad_norm": 1.0890011420632484,
+      "learning_rate": 8.629758401913027e-06,
+      "loss": 0.0269,
+      "step": 927
+    },
+    {
+      "epoch": 0.26461362988309095,
+      "grad_norm": 1.0417579612272687,
+      "learning_rate": 8.626580410293859e-06,
+      "loss": 0.0133,
+      "step": 928
+    },
+    {
+      "epoch": 0.2648987738808098,
+      "grad_norm": 1.236301382418515,
+      "learning_rate": 8.623399324222608e-06,
+      "loss": 0.0164,
+      "step": 929
+    },
+    {
+      "epoch": 0.26518391787852863,
+      "grad_norm": 1.4547427099574102,
+      "learning_rate": 8.620215146413603e-06,
+      "loss": 0.0193,
+      "step": 930
+    },
+    {
+      "epoch": 0.2654690618762475,
+      "grad_norm": 0.9108297216575995,
+      "learning_rate": 8.617027879583801e-06,
+      "loss": 0.0211,
+      "step": 931
+    },
+    {
+      "epoch": 0.26575420587396636,
+      "grad_norm": 1.0697944180054113,
+      "learning_rate": 8.613837526452806e-06,
+      "loss": 0.0295,
+      "step": 932
+    },
+    {
+      "epoch": 0.2660393498716852,
+      "grad_norm": 0.6751528441766895,
+      "learning_rate": 8.61064408974285e-06,
+      "loss": 0.0156,
+      "step": 933
+    },
+    {
+      "epoch": 0.26632449386940404,
+      "grad_norm": 1.8490910731349444,
+      "learning_rate": 8.607447572178796e-06,
+      "loss": 0.0642,
+      "step": 934
+    },
+    {
+      "epoch": 0.2666096378671229,
+      "grad_norm": 0.7855324683250933,
+      "learning_rate": 8.604247976488137e-06,
+      "loss": 0.0292,
+      "step": 935
+    },
+    {
+      "epoch": 0.2668947818648417,
+      "grad_norm": 1.731820383728147,
+      "learning_rate": 8.601045305400988e-06,
+      "loss": 0.0373,
+      "step": 936
+    },
+    {
+      "epoch": 0.2671799258625606,
+      "grad_norm": 2.1895639341457716,
+      "learning_rate": 8.597839561650096e-06,
+      "loss": 0.0577,
+      "step": 937
+    },
+    {
+      "epoch": 0.26746506986027946,
+      "grad_norm": 0.891986333285142,
+      "learning_rate": 8.594630747970824e-06,
+      "loss": 0.0315,
+      "step": 938
+    },
+    {
+      "epoch": 0.2677502138579983,
+      "grad_norm": 0.6734024951645515,
+      "learning_rate": 8.591418867101158e-06,
+      "loss": 0.0169,
+      "step": 939
+    },
+    {
+      "epoch": 0.26803535785571714,
+      "grad_norm": 0.7790384279219437,
+      "learning_rate": 8.588203921781699e-06,
+      "loss": 0.024,
+      "step": 940
+    },
+    {
+      "epoch": 0.268320501853436,
+      "grad_norm": 1.8442659646304564,
+      "learning_rate": 8.584985914755663e-06,
+      "loss": 0.0476,
+      "step": 941
+    },
+    {
+      "epoch": 0.2686056458511548,
+      "grad_norm": 1.1596418180170927,
+      "learning_rate": 8.581764848768878e-06,
+      "loss": 0.0444,
+      "step": 942
+    },
+    {
+      "epoch": 0.26889078984887366,
+      "grad_norm": 1.4812227170866508,
+      "learning_rate": 8.578540726569782e-06,
+      "loss": 0.0379,
+      "step": 943
+    },
+    {
+      "epoch": 0.26917593384659255,
+      "grad_norm": 0.778765028574822,
+      "learning_rate": 8.575313550909424e-06,
+      "loss": 0.017,
+      "step": 944
+    },
+    {
+      "epoch": 0.2694610778443114,
+      "grad_norm": 0.6214418267264455,
+      "learning_rate": 8.572083324541454e-06,
+      "loss": 0.0193,
+      "step": 945
+    },
+    {
+      "epoch": 0.26974622184203023,
+      "grad_norm": 0.5537904546033956,
+      "learning_rate": 8.568850050222129e-06,
+      "loss": 0.0156,
+      "step": 946
+    },
+    {
+      "epoch": 0.27003136583974907,
+      "grad_norm": 1.9665330817864057,
+      "learning_rate": 8.565613730710303e-06,
+      "loss": 0.0363,
+      "step": 947
+    },
+    {
+      "epoch": 0.2703165098374679,
+      "grad_norm": 1.7179405656998434,
+      "learning_rate": 8.56237436876743e-06,
+      "loss": 0.0282,
+      "step": 948
+    },
+    {
+      "epoch": 0.27060165383518675,
+      "grad_norm": 1.591795064619829,
+      "learning_rate": 8.559131967157561e-06,
+      "loss": 0.0381,
+      "step": 949
+    },
+    {
+      "epoch": 0.2708867978329056,
+      "grad_norm": 1.4855162991515705,
+      "learning_rate": 8.55588652864734e-06,
+      "loss": 0.0319,
+      "step": 950
+    },
+    {
+      "epoch": 0.2711719418306245,
+      "grad_norm": 1.23660525088581,
+      "learning_rate": 8.552638056006004e-06,
+      "loss": 0.0211,
+      "step": 951
+    },
+    {
+      "epoch": 0.2714570858283433,
+      "grad_norm": 1.3975270889745315,
+      "learning_rate": 8.549386552005375e-06,
+      "loss": 0.0134,
+      "step": 952
+    },
+    {
+      "epoch": 0.27174222982606216,
+      "grad_norm": 1.9571395698904834,
+      "learning_rate": 8.546132019419862e-06,
+      "loss": 0.0248,
+      "step": 953
+    },
+    {
+      "epoch": 0.272027373823781,
+      "grad_norm": 2.1677184607215527,
+      "learning_rate": 8.542874461026462e-06,
+      "loss": 0.0502,
+      "step": 954
+    },
+    {
+      "epoch": 0.27231251782149984,
+      "grad_norm": 1.699300392038293,
+      "learning_rate": 8.539613879604751e-06,
+      "loss": 0.049,
+      "step": 955
+    },
+    {
+      "epoch": 0.2725976618192187,
+      "grad_norm": 1.5110559904082717,
+      "learning_rate": 8.536350277936887e-06,
+      "loss": 0.0361,
+      "step": 956
+    },
+    {
+      "epoch": 0.2728828058169376,
+      "grad_norm": 1.694634659292682,
+      "learning_rate": 8.533083658807601e-06,
+      "loss": 0.0602,
+      "step": 957
+    },
+    {
+      "epoch": 0.2731679498146564,
+      "grad_norm": 0.6948943012372113,
+      "learning_rate": 8.529814025004202e-06,
+      "loss": 0.0208,
+      "step": 958
+    },
+    {
+      "epoch": 0.27345309381237526,
+      "grad_norm": 1.2290435954640724,
+      "learning_rate": 8.526541379316569e-06,
+      "loss": 0.037,
+      "step": 959
+    },
+    {
+      "epoch": 0.2737382378100941,
+      "grad_norm": 0.9917533429856523,
+      "learning_rate": 8.523265724537153e-06,
+      "loss": 0.0326,
+      "step": 960
+    },
+    {
+      "epoch": 0.27402338180781294,
+      "grad_norm": 0.5603987383374495,
+      "learning_rate": 8.519987063460973e-06,
+      "loss": 0.0135,
+      "step": 961
+    },
+    {
+      "epoch": 0.2743085258055318,
+      "grad_norm": 0.7544090611895435,
+      "learning_rate": 8.51670539888561e-06,
+      "loss": 0.0346,
+      "step": 962
+    },
+    {
+      "epoch": 0.2745936698032506,
+      "grad_norm": 1.2787254933712948,
+      "learning_rate": 8.513420733611212e-06,
+      "loss": 0.0456,
+      "step": 963
+    },
+    {
+      "epoch": 0.2748788138009695,
+      "grad_norm": 1.186340995532642,
+      "learning_rate": 8.510133070440483e-06,
+      "loss": 0.023,
+      "step": 964
+    },
+    {
+      "epoch": 0.27516395779868835,
+      "grad_norm": 1.1520864734843645,
+      "learning_rate": 8.506842412178688e-06,
+      "loss": 0.0204,
+      "step": 965
+    },
+    {
+      "epoch": 0.2754491017964072,
+      "grad_norm": 1.3801482600792678,
+      "learning_rate": 8.503548761633646e-06,
+      "loss": 0.0513,
+      "step": 966
+    },
+    {
+      "epoch": 0.27573424579412603,
+      "grad_norm": 1.4900548822914077,
+      "learning_rate": 8.500252121615733e-06,
+      "loss": 0.0357,
+      "step": 967
+    },
+    {
+      "epoch": 0.27601938979184487,
+      "grad_norm": 1.3136520745656428,
+      "learning_rate": 8.496952494937869e-06,
+      "loss": 0.0287,
+      "step": 968
+    },
+    {
+      "epoch": 0.2763045337895637,
+      "grad_norm": 1.1985944568873772,
+      "learning_rate": 8.493649884415529e-06,
+      "loss": 0.0288,
+      "step": 969
+    },
+    {
+      "epoch": 0.27658967778728255,
+      "grad_norm": 1.3680126216074568,
+      "learning_rate": 8.490344292866728e-06,
+      "loss": 0.042,
+      "step": 970
+    },
+    {
+      "epoch": 0.27687482178500145,
+      "grad_norm": 1.3159870085955017,
+      "learning_rate": 8.487035723112033e-06,
+      "loss": 0.0205,
+      "step": 971
+    },
+    {
+      "epoch": 0.2771599657827203,
+      "grad_norm": 1.0078868077879315,
+      "learning_rate": 8.483724177974543e-06,
+      "loss": 0.0172,
+      "step": 972
+    },
+    {
+      "epoch": 0.2774451097804391,
+      "grad_norm": 1.1647342396721414,
+      "learning_rate": 8.480409660279903e-06,
+      "loss": 0.0187,
+      "step": 973
+    },
+    {
+      "epoch": 0.27773025377815796,
+      "grad_norm": 2.31581097184704,
+      "learning_rate": 8.477092172856287e-06,
+      "loss": 0.0386,
+      "step": 974
+    },
+    {
+      "epoch": 0.2780153977758768,
+      "grad_norm": 2.810835578979634,
+      "learning_rate": 8.473771718534411e-06,
+      "loss": 0.0757,
+      "step": 975
+    },
+    {
+      "epoch": 0.27830054177359564,
+      "grad_norm": 2.79707883978544,
+      "learning_rate": 8.47044830014752e-06,
+      "loss": 0.0477,
+      "step": 976
+    },
+    {
+      "epoch": 0.27858568577131454,
+      "grad_norm": 0.9667581419870983,
+      "learning_rate": 8.467121920531383e-06,
+      "loss": 0.0242,
+      "step": 977
+    },
+    {
+      "epoch": 0.2788708297690334,
+      "grad_norm": 0.6681742594213469,
+      "learning_rate": 8.463792582524302e-06,
+      "loss": 0.0096,
+      "step": 978
+    },
+    {
+      "epoch": 0.2791559737667522,
+      "grad_norm": 1.4545154240977922,
+      "learning_rate": 8.460460288967101e-06,
+      "loss": 0.0217,
+      "step": 979
+    },
+    {
+      "epoch": 0.27944111776447106,
+      "grad_norm": 0.3637082821635198,
+      "learning_rate": 8.457125042703124e-06,
+      "loss": 0.0067,
+      "step": 980
+    },
+    {
+      "epoch": 0.2797262617621899,
+      "grad_norm": 1.7754271211354633,
+      "learning_rate": 8.45378684657824e-06,
+      "loss": 0.0622,
+      "step": 981
+    },
+    {
+      "epoch": 0.28001140575990874,
+      "grad_norm": 0.512206035741963,
+      "learning_rate": 8.45044570344083e-06,
+      "loss": 0.0079,
+      "step": 982
+    },
+    {
+      "epoch": 0.2802965497576276,
+      "grad_norm": 0.287836877286698,
+      "learning_rate": 8.44710161614179e-06,
+      "loss": 0.0067,
+      "step": 983
+    },
+    {
+      "epoch": 0.2805816937553465,
+      "grad_norm": 0.5981293412787716,
+      "learning_rate": 8.443754587534529e-06,
+      "loss": 0.0182,
+      "step": 984
+    },
+    {
+      "epoch": 0.2808668377530653,
+      "grad_norm": 1.5761429513855125,
+      "learning_rate": 8.440404620474967e-06,
+      "loss": 0.0237,
+      "step": 985
+    },
+    {
+      "epoch": 0.28115198175078415,
+      "grad_norm": 1.3666585287674748,
+      "learning_rate": 8.43705171782153e-06,
+      "loss": 0.0425,
+      "step": 986
+    },
+    {
+      "epoch": 0.281437125748503,
+      "grad_norm": 1.697391675178744,
+      "learning_rate": 8.43369588243515e-06,
+      "loss": 0.0236,
+      "step": 987
+    },
+    {
+      "epoch": 0.28172226974622183,
+      "grad_norm": 3.0482492895545796,
+      "learning_rate": 8.430337117179259e-06,
+      "loss": 0.0409,
+      "step": 988
+    },
+    {
+      "epoch": 0.28200741374394067,
+      "grad_norm": 0.7229976500220648,
+      "learning_rate": 8.426975424919791e-06,
+      "loss": 0.01,
+      "step": 989
+    },
+    {
+      "epoch": 0.2822925577416595,
+      "grad_norm": 1.4158032540007655,
+      "learning_rate": 8.423610808525177e-06,
+      "loss": 0.0186,
+      "step": 990
+    },
+    {
+      "epoch": 0.2825777017393784,
+      "grad_norm": 1.6615373983383692,
+      "learning_rate": 8.420243270866343e-06,
+      "loss": 0.0652,
+      "step": 991
+    },
+    {
+      "epoch": 0.28286284573709725,
+      "grad_norm": 2.307370170129627,
+      "learning_rate": 8.416872814816707e-06,
+      "loss": 0.047,
+      "step": 992
+    },
+    {
+      "epoch": 0.2831479897348161,
+      "grad_norm": 1.8461913330119348,
+      "learning_rate": 8.41349944325218e-06,
+      "loss": 0.0465,
+      "step": 993
+    },
+    {
+      "epoch": 0.2834331337325349,
+      "grad_norm": 1.8568994890741377,
+      "learning_rate": 8.410123159051155e-06,
+      "loss": 0.0205,
+      "step": 994
+    },
+    {
+      "epoch": 0.28371827773025377,
+      "grad_norm": 1.2212193684268584,
+      "learning_rate": 8.40674396509452e-06,
+      "loss": 0.0218,
+      "step": 995
+    },
+    {
+      "epoch": 0.2840034217279726,
+      "grad_norm": 1.1766960236291593,
+      "learning_rate": 8.40336186426563e-06,
+      "loss": 0.0457,
+      "step": 996
+    },
+    {
+      "epoch": 0.2842885657256915,
+      "grad_norm": 1.1251722175740966,
+      "learning_rate": 8.39997685945034e-06,
+      "loss": 0.0216,
+      "step": 997
+    },
+    {
+      "epoch": 0.28457370972341034,
+      "grad_norm": 2.3908402551223786,
+      "learning_rate": 8.396588953536968e-06,
+      "loss": 0.0403,
+      "step": 998
+    },
+    {
+      "epoch": 0.2848588537211292,
+      "grad_norm": 2.331842901417062,
+      "learning_rate": 8.393198149416311e-06,
+      "loss": 0.0701,
+      "step": 999
+    },
+    {
+      "epoch": 0.285143997718848,
+      "grad_norm": 1.9563319254857874,
+      "learning_rate": 8.389804449981645e-06,
+      "loss": 0.0171,
+      "step": 1000
+    },
+    {
+      "epoch": 0.28542914171656686,
+      "grad_norm": 1.5049554471669486,
+      "learning_rate": 8.386407858128707e-06,
+      "loss": 0.0368,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 2.3620183376762647,
+      "learning_rate": 8.383008376755707e-06,
+      "loss": 0.0298,
+      "step": 1002
+    },
+    {
+      "epoch": 0.28599942971200454,
+      "grad_norm": 1.5127708826099804,
+      "learning_rate": 8.379606008763325e-06,
+      "loss": 0.0384,
+      "step": 1003
+    },
+    {
+      "epoch": 0.28628457370972343,
+      "grad_norm": 1.9077680147517553,
+      "learning_rate": 8.376200757054695e-06,
+      "loss": 0.0313,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2865697177074423,
+      "grad_norm": 1.3689906306129866,
+      "learning_rate": 8.372792624535417e-06,
+      "loss": 0.0321,
+      "step": 1005
+    },
+    {
+      "epoch": 0.2868548617051611,
+      "grad_norm": 1.1238958908515264,
+      "learning_rate": 8.369381614113547e-06,
+      "loss": 0.014,
+      "step": 1006
+    },
+    {
+      "epoch": 0.28714000570287995,
+      "grad_norm": 1.2072815121718892,
+      "learning_rate": 8.365967728699602e-06,
+      "loss": 0.0179,
+      "step": 1007
+    },
+    {
+      "epoch": 0.2874251497005988,
+      "grad_norm": 0.4667146551541128,
+      "learning_rate": 8.362550971206543e-06,
+      "loss": 0.0146,
+      "step": 1008
+    },
+    {
+      "epoch": 0.28771029369831763,
+      "grad_norm": 1.6836956764062203,
+      "learning_rate": 8.359131344549788e-06,
+      "loss": 0.0299,
+      "step": 1009
+    },
+    {
+      "epoch": 0.28799543769603647,
+      "grad_norm": 0.8142984273923918,
+      "learning_rate": 8.355708851647202e-06,
+      "loss": 0.0226,
+      "step": 1010
+    },
+    {
+      "epoch": 0.28828058169375537,
+      "grad_norm": 1.6389536841060004,
+      "learning_rate": 8.352283495419096e-06,
+      "loss": 0.0308,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2885657256914742,
+      "grad_norm": 1.5444478808693596,
+      "learning_rate": 8.348855278788224e-06,
+      "loss": 0.0589,
+      "step": 1012
+    },
+    {
+      "epoch": 0.28885086968919305,
+      "grad_norm": 0.7259294284234575,
+      "learning_rate": 8.345424204679778e-06,
+      "loss": 0.0191,
+      "step": 1013
+    },
+    {
+      "epoch": 0.2891360136869119,
+      "grad_norm": 0.6088052137967528,
+      "learning_rate": 8.34199027602139e-06,
+      "loss": 0.0158,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2894211576846307,
+      "grad_norm": 0.6287296236873937,
+      "learning_rate": 8.338553495743132e-06,
+      "loss": 0.0218,
+      "step": 1015
+    },
+    {
+      "epoch": 0.28970630168234957,
+      "grad_norm": 1.7658297184074514,
+      "learning_rate": 8.335113866777502e-06,
+      "loss": 0.0441,
+      "step": 1016
+    },
+    {
+      "epoch": 0.28999144568006846,
+      "grad_norm": 0.5235221980156642,
+      "learning_rate": 8.331671392059433e-06,
+      "loss": 0.0166,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2902765896777873,
+      "grad_norm": 0.8033872378713534,
+      "learning_rate": 8.328226074526284e-06,
+      "loss": 0.0292,
+      "step": 1018
+    },
+    {
+      "epoch": 0.29056173367550614,
+      "grad_norm": 1.568699813214378,
+      "learning_rate": 8.324777917117843e-06,
+      "loss": 0.029,
+      "step": 1019
+    },
+    {
+      "epoch": 0.290846877673225,
+      "grad_norm": 0.8052329509177221,
+      "learning_rate": 8.321326922776321e-06,
+      "loss": 0.0231,
+      "step": 1020
+    },
+    {
+      "epoch": 0.2911320216709438,
+      "grad_norm": 0.5269396898380581,
+      "learning_rate": 8.31787309444634e-06,
+      "loss": 0.0174,
+      "step": 1021
+    },
+    {
+      "epoch": 0.29141716566866266,
+      "grad_norm": 1.6205080390020445,
+      "learning_rate": 8.314416435074956e-06,
+      "loss": 0.0197,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2917023096663815,
+      "grad_norm": 1.666001562463425,
+      "learning_rate": 8.31095694761163e-06,
+      "loss": 0.019,
+      "step": 1023
+    },
+    {
+      "epoch": 0.2919874536641004,
+      "grad_norm": 1.835698903031861,
+      "learning_rate": 8.307494635008237e-06,
+      "loss": 0.0214,
+      "step": 1024
+    },
+    {
+      "epoch": 0.29227259766181923,
+      "grad_norm": 0.8037987767198944,
+      "learning_rate": 8.304029500219064e-06,
+      "loss": 0.0168,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2925577416595381,
+      "grad_norm": 2.370065976232439,
+      "learning_rate": 8.300561546200812e-06,
+      "loss": 0.0247,
+      "step": 1026
+    },
+    {
+      "epoch": 0.2928428856572569,
+      "grad_norm": 1.5504716392290512,
+      "learning_rate": 8.297090775912574e-06,
+      "loss": 0.0391,
+      "step": 1027
+    },
+    {
+      "epoch": 0.29312802965497575,
+      "grad_norm": 1.2554327251719692,
+      "learning_rate": 8.293617192315859e-06,
+      "loss": 0.0517,
+      "step": 1028
+    },
+    {
+      "epoch": 0.2934131736526946,
+      "grad_norm": 0.790459152689314,
+      "learning_rate": 8.29014079837457e-06,
+      "loss": 0.0119,
+      "step": 1029
+    },
+    {
+      "epoch": 0.29369831765041343,
+      "grad_norm": 1.4149648173860672,
+      "learning_rate": 8.28666159705501e-06,
+      "loss": 0.0133,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2939834616481323,
+      "grad_norm": 0.9973359366143715,
+      "learning_rate": 8.283179591325879e-06,
+      "loss": 0.013,
+      "step": 1031
+    },
+    {
+      "epoch": 0.29426860564585117,
+      "grad_norm": 0.36926897766983774,
+      "learning_rate": 8.279694784158262e-06,
+      "loss": 0.008,
+      "step": 1032
+    },
+    {
+      "epoch": 0.29455374964357,
+      "grad_norm": 0.9623976721838673,
+      "learning_rate": 8.276207178525646e-06,
+      "loss": 0.0196,
+      "step": 1033
+    },
+    {
+      "epoch": 0.29483889364128885,
+      "grad_norm": 1.4554871378304686,
+      "learning_rate": 8.272716777403898e-06,
+      "loss": 0.041,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2951240376390077,
+      "grad_norm": 2.6162190558641463,
+      "learning_rate": 8.26922358377127e-06,
+      "loss": 0.0745,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2954091816367265,
+      "grad_norm": 0.8422156448296566,
+      "learning_rate": 8.265727600608401e-06,
+      "loss": 0.0194,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2956943256344454,
+      "grad_norm": 1.0120407410840147,
+      "learning_rate": 8.262228830898313e-06,
+      "loss": 0.0387,
+      "step": 1037
+    },
+    {
+      "epoch": 0.29597946963216426,
+      "grad_norm": 0.7723311949173021,
+      "learning_rate": 8.258727277626394e-06,
+      "loss": 0.0102,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2962646136298831,
+      "grad_norm": 0.9132982937599546,
+      "learning_rate": 8.255222943780419e-06,
+      "loss": 0.0236,
+      "step": 1039
+    },
+    {
+      "epoch": 0.29654975762760194,
+      "grad_norm": 1.4697397386034843,
+      "learning_rate": 8.251715832350526e-06,
+      "loss": 0.0406,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2968349016253208,
+      "grad_norm": 1.3125795625460037,
+      "learning_rate": 8.248205946329233e-06,
+      "loss": 0.0249,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2971200456230396,
+      "grad_norm": 2.0022314201473765,
+      "learning_rate": 8.244693288711416e-06,
+      "loss": 0.0389,
+      "step": 1042
+    },
+    {
+      "epoch": 0.29740518962075846,
+      "grad_norm": 0.9233574909374214,
+      "learning_rate": 8.241177862494323e-06,
+      "loss": 0.0177,
+      "step": 1043
+    },
+    {
+      "epoch": 0.29769033361847735,
+      "grad_norm": 1.118407928249881,
+      "learning_rate": 8.23765967067756e-06,
+      "loss": 0.0504,
+      "step": 1044
+    },
+    {
+      "epoch": 0.2979754776161962,
+      "grad_norm": 2.874005101086032,
+      "learning_rate": 8.234138716263095e-06,
+      "loss": 0.0629,
+      "step": 1045
+    },
+    {
+      "epoch": 0.29826062161391503,
+      "grad_norm": 1.7249213253113764,
+      "learning_rate": 8.230615002255254e-06,
+      "loss": 0.039,
+      "step": 1046
+    },
+    {
+      "epoch": 0.2985457656116339,
+      "grad_norm": 1.282268931205406,
+      "learning_rate": 8.227088531660712e-06,
+      "loss": 0.0313,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2988309096093527,
+      "grad_norm": 1.5416921550475424,
+      "learning_rate": 8.223559307488506e-06,
+      "loss": 0.0517,
+      "step": 1048
+    },
+    {
+      "epoch": 0.29911605360707155,
+      "grad_norm": 0.9912931776780044,
+      "learning_rate": 8.220027332750012e-06,
+      "loss": 0.0316,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2994011976047904,
+      "grad_norm": 0.6255964251801459,
+      "learning_rate": 8.21649261045896e-06,
+      "loss": 0.0218,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2996863416025093,
+      "grad_norm": 1.0299927686497676,
+      "learning_rate": 8.212955143631425e-06,
+      "loss": 0.0246,
+      "step": 1051
+    },
+    {
+      "epoch": 0.2999714856002281,
+      "grad_norm": 1.5809553916222134,
+      "learning_rate": 8.209414935285816e-06,
+      "loss": 0.036,
+      "step": 1052
+    },
+    {
+      "epoch": 0.30025662959794697,
+      "grad_norm": 1.0729865806566057,
+      "learning_rate": 8.20587198844289e-06,
+      "loss": 0.0227,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3005417735956658,
+      "grad_norm": 2.1564117236579454,
+      "learning_rate": 8.202326306125736e-06,
+      "loss": 0.0314,
+      "step": 1054
+    },
+    {
+      "epoch": 0.30082691759338465,
+      "grad_norm": 0.9502094736160007,
+      "learning_rate": 8.198777891359778e-06,
+      "loss": 0.0248,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3011120615911035,
+      "grad_norm": 0.9962396374644391,
+      "learning_rate": 8.19522674717277e-06,
+      "loss": 0.041,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3013972055888224,
+      "grad_norm": 0.6369816183985646,
+      "learning_rate": 8.1916728765948e-06,
+      "loss": 0.0237,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3016823495865412,
+      "grad_norm": 2.5117579204688623,
+      "learning_rate": 8.188116282658278e-06,
+      "loss": 0.0359,
+      "step": 1058
+    },
+    {
+      "epoch": 0.30196749358426006,
+      "grad_norm": 1.6764540415557843,
+      "learning_rate": 8.184556968397938e-06,
+      "loss": 0.0259,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3022526375819789,
+      "grad_norm": 0.9996801308305837,
+      "learning_rate": 8.180994936850834e-06,
+      "loss": 0.0242,
+      "step": 1060
+    },
+    {
+      "epoch": 0.30253778157969774,
+      "grad_norm": 0.7595100913126963,
+      "learning_rate": 8.177430191056346e-06,
+      "loss": 0.0173,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3028229255774166,
+      "grad_norm": 0.9654704679405917,
+      "learning_rate": 8.173862734056158e-06,
+      "loss": 0.0277,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3031080695751354,
+      "grad_norm": 1.5038824626171603,
+      "learning_rate": 8.170292568894278e-06,
+      "loss": 0.0237,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3033932135728543,
+      "grad_norm": 0.7973443754741695,
+      "learning_rate": 8.16671969861702e-06,
+      "loss": 0.027,
+      "step": 1064
+    },
+    {
+      "epoch": 0.30367835757057315,
+      "grad_norm": 1.8187366967593017,
+      "learning_rate": 8.163144126273004e-06,
+      "loss": 0.025,
+      "step": 1065
+    },
+    {
+      "epoch": 0.303963501568292,
+      "grad_norm": 0.9250825610118477,
+      "learning_rate": 8.159565854913162e-06,
+      "loss": 0.0175,
+      "step": 1066
+    },
+    {
+      "epoch": 0.30424864556601083,
+      "grad_norm": 1.506560630883658,
+      "learning_rate": 8.155984887590724e-06,
+      "loss": 0.0332,
+      "step": 1067
+    },
+    {
+      "epoch": 0.3045337895637297,
+      "grad_norm": 1.2512981744236478,
+      "learning_rate": 8.152401227361224e-06,
+      "loss": 0.0369,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3048189335614485,
+      "grad_norm": 1.8833596709136027,
+      "learning_rate": 8.148814877282487e-06,
+      "loss": 0.0292,
+      "step": 1069
+    },
+    {
+      "epoch": 0.30510407755916735,
+      "grad_norm": 0.7881838063455426,
+      "learning_rate": 8.145225840414641e-06,
+      "loss": 0.0101,
+      "step": 1070
+    },
+    {
+      "epoch": 0.30538922155688625,
+      "grad_norm": 1.6796194211057058,
+      "learning_rate": 8.141634119820101e-06,
+      "loss": 0.0287,
+      "step": 1071
+    },
+    {
+      "epoch": 0.3056743655546051,
+      "grad_norm": 1.2891380669074353,
+      "learning_rate": 8.138039718563578e-06,
+      "loss": 0.0492,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3059595095523239,
+      "grad_norm": 2.357993042502323,
+      "learning_rate": 8.134442639712063e-06,
+      "loss": 0.0456,
+      "step": 1073
+    },
+    {
+      "epoch": 0.30624465355004277,
+      "grad_norm": 0.5191000758439441,
+      "learning_rate": 8.130842886334837e-06,
+      "loss": 0.0083,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3065297975477616,
+      "grad_norm": 0.9634677131357932,
+      "learning_rate": 8.127240461503462e-06,
+      "loss": 0.0159,
+      "step": 1075
+    },
+    {
+      "epoch": 0.30681494154548045,
+      "grad_norm": 0.27264204707454004,
+      "learning_rate": 8.123635368291777e-06,
+      "loss": 0.0051,
+      "step": 1076
+    },
+    {
+      "epoch": 0.30710008554319934,
+      "grad_norm": 2.2718130414005446,
+      "learning_rate": 8.120027609775902e-06,
+      "loss": 0.04,
+      "step": 1077
+    },
+    {
+      "epoch": 0.3073852295409182,
+      "grad_norm": 1.2599259798035607,
+      "learning_rate": 8.116417189034227e-06,
+      "loss": 0.0234,
+      "step": 1078
+    },
+    {
+      "epoch": 0.307670373538637,
+      "grad_norm": 1.2003202010441418,
+      "learning_rate": 8.112804109147416e-06,
+      "loss": 0.0137,
+      "step": 1079
+    },
+    {
+      "epoch": 0.30795551753635586,
+      "grad_norm": 1.525563420275153,
+      "learning_rate": 8.1091883731984e-06,
+      "loss": 0.0452,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3082406615340747,
+      "grad_norm": 1.8645400213251933,
+      "learning_rate": 8.10556998427238e-06,
+      "loss": 0.0267,
+      "step": 1081
+    },
+    {
+      "epoch": 0.30852580553179354,
+      "grad_norm": 2.063747391446373,
+      "learning_rate": 8.10194894545682e-06,
+      "loss": 0.0357,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3088109495295124,
+      "grad_norm": 1.2254249014948786,
+      "learning_rate": 8.09832525984144e-06,
+      "loss": 0.0226,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3090960935272313,
+      "grad_norm": 1.4378801385084123,
+      "learning_rate": 8.094698930518224e-06,
+      "loss": 0.0207,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3093812375249501,
+      "grad_norm": 1.6399576082537426,
+      "learning_rate": 8.091069960581408e-06,
+      "loss": 0.0208,
+      "step": 1085
+    },
+    {
+      "epoch": 0.30966638152266895,
+      "grad_norm": 0.4934963525040569,
+      "learning_rate": 8.087438353127486e-06,
+      "loss": 0.0065,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3099515255203878,
+      "grad_norm": 1.009916460402547,
+      "learning_rate": 8.083804111255197e-06,
+      "loss": 0.0179,
+      "step": 1087
+    },
+    {
+      "epoch": 0.31023666951810663,
+      "grad_norm": 0.8948637487646737,
+      "learning_rate": 8.08016723806553e-06,
+      "loss": 0.0202,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3105218135158255,
+      "grad_norm": 1.3076962346304433,
+      "learning_rate": 8.076527736661724e-06,
+      "loss": 0.0362,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3108069575135443,
+      "grad_norm": 2.205028660931532,
+      "learning_rate": 8.072885610149251e-06,
+      "loss": 0.0309,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3110921015112632,
+      "grad_norm": 1.4313613753949053,
+      "learning_rate": 8.06924086163583e-06,
+      "loss": 0.021,
+      "step": 1091
+    },
+    {
+      "epoch": 0.31137724550898205,
+      "grad_norm": 2.054475718794255,
+      "learning_rate": 8.065593494231418e-06,
+      "loss": 0.0468,
+      "step": 1092
+    },
+    {
+      "epoch": 0.3116623895067009,
+      "grad_norm": 1.7740986866927582,
+      "learning_rate": 8.061943511048199e-06,
+      "loss": 0.0516,
+      "step": 1093
+    },
+    {
+      "epoch": 0.3119475335044197,
+      "grad_norm": 46.640258129930466,
+      "learning_rate": 8.058290915200597e-06,
+      "loss": 0.1971,
+      "step": 1094
+    },
+    {
+      "epoch": 0.31223267750213857,
+      "grad_norm": 0.781762262501865,
+      "learning_rate": 8.054635709805263e-06,
+      "loss": 0.0248,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3125178214998574,
+      "grad_norm": 1.082335674544725,
+      "learning_rate": 8.050977897981071e-06,
+      "loss": 0.0357,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3128029654975763,
+      "grad_norm": 1.7052642583096669,
+      "learning_rate": 8.047317482849124e-06,
+      "loss": 0.0305,
+      "step": 1097
+    },
+    {
+      "epoch": 0.31308810949529514,
+      "grad_norm": 1.9320064281915101,
+      "learning_rate": 8.043654467532744e-06,
+      "loss": 0.0486,
+      "step": 1098
+    },
+    {
+      "epoch": 0.313373253493014,
+      "grad_norm": 2.6104512138319564,
+      "learning_rate": 8.039988855157472e-06,
+      "loss": 0.0608,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3136583974907328,
+      "grad_norm": 0.906309465890772,
+      "learning_rate": 8.036320648851064e-06,
+      "loss": 0.0188,
+      "step": 1100
+    },
+    {
+      "epoch": 0.31394354148845166,
+      "grad_norm": 1.2528075290998113,
+      "learning_rate": 8.032649851743493e-06,
+      "loss": 0.0458,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3142286854861705,
+      "grad_norm": 1.7482113067259941,
+      "learning_rate": 8.028976466966934e-06,
+      "loss": 0.0369,
+      "step": 1102
+    },
+    {
+      "epoch": 0.31451382948388934,
+      "grad_norm": 2.9189918710398755,
+      "learning_rate": 8.025300497655783e-06,
+      "loss": 0.0377,
+      "step": 1103
+    },
+    {
+      "epoch": 0.31479897348160824,
+      "grad_norm": 1.0983245681069267,
+      "learning_rate": 8.021621946946628e-06,
+      "loss": 0.0242,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3150841174793271,
+      "grad_norm": 1.2121132358162416,
+      "learning_rate": 8.01794081797827e-06,
+      "loss": 0.0335,
+      "step": 1105
+    },
+    {
+      "epoch": 0.3153692614770459,
+      "grad_norm": 0.8067611328102654,
+      "learning_rate": 8.014257113891704e-06,
+      "loss": 0.0301,
+      "step": 1106
+    },
+    {
+      "epoch": 0.31565440547476475,
+      "grad_norm": 0.8529627858655344,
+      "learning_rate": 8.010570837830124e-06,
+      "loss": 0.0367,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3159395494724836,
+      "grad_norm": 0.8676313074444348,
+      "learning_rate": 8.00688199293892e-06,
+      "loss": 0.019,
+      "step": 1108
+    },
+    {
+      "epoch": 0.31622469347020243,
+      "grad_norm": 0.7427102352300117,
+      "learning_rate": 8.003190582365669e-06,
+      "loss": 0.0262,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3165098374679213,
+      "grad_norm": 0.9736769790238554,
+      "learning_rate": 7.999496609260144e-06,
+      "loss": 0.0184,
+      "step": 1110
+    },
+    {
+      "epoch": 0.31679498146564017,
+      "grad_norm": 1.27619488452561,
+      "learning_rate": 7.995800076774301e-06,
+      "loss": 0.0232,
+      "step": 1111
+    },
+    {
+      "epoch": 0.317080125463359,
+      "grad_norm": 2.0389975152307476,
+      "learning_rate": 7.992100988062277e-06,
+      "loss": 0.0393,
+      "step": 1112
+    },
+    {
+      "epoch": 0.31736526946107785,
+      "grad_norm": 0.7124345990268323,
+      "learning_rate": 7.988399346280398e-06,
+      "loss": 0.0157,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3176504134587967,
+      "grad_norm": 0.7642116938852253,
+      "learning_rate": 7.98469515458716e-06,
+      "loss": 0.0239,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3179355574565155,
+      "grad_norm": 0.8303694050770815,
+      "learning_rate": 7.980988416143239e-06,
+      "loss": 0.0188,
+      "step": 1115
+    },
+    {
+      "epoch": 0.31822070145423437,
+      "grad_norm": 1.7299262380985978,
+      "learning_rate": 7.977279134111487e-06,
+      "loss": 0.0384,
+      "step": 1116
+    },
+    {
+      "epoch": 0.31850584545195326,
+      "grad_norm": 0.821877902826892,
+      "learning_rate": 7.973567311656917e-06,
+      "loss": 0.0152,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3187909894496721,
+      "grad_norm": 0.841185794140611,
+      "learning_rate": 7.96985295194672e-06,
+      "loss": 0.0213,
+      "step": 1118
+    },
+    {
+      "epoch": 0.31907613344739094,
+      "grad_norm": 2.9117336343809694,
+      "learning_rate": 7.966136058150247e-06,
+      "loss": 0.0337,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3193612774451098,
+      "grad_norm": 0.37780063918483164,
+      "learning_rate": 7.962416633439008e-06,
+      "loss": 0.0143,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3196464214428286,
+      "grad_norm": 1.1660782504301215,
+      "learning_rate": 7.958694680986682e-06,
+      "loss": 0.0258,
+      "step": 1121
+    },
+    {
+      "epoch": 0.31993156544054746,
+      "grad_norm": 0.6348660624052973,
+      "learning_rate": 7.954970203969095e-06,
+      "loss": 0.0401,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3202167094382663,
+      "grad_norm": 0.5893537112468747,
+      "learning_rate": 7.951243205564234e-06,
+      "loss": 0.0106,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3205018534359852,
+      "grad_norm": 0.7759687299701817,
+      "learning_rate": 7.947513688952234e-06,
+      "loss": 0.0202,
+      "step": 1124
+    },
+    {
+      "epoch": 0.32078699743370404,
+      "grad_norm": 1.7145098320011822,
+      "learning_rate": 7.943781657315377e-06,
+      "loss": 0.0227,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3210721414314229,
+      "grad_norm": 1.8907573081270583,
+      "learning_rate": 7.940047113838096e-06,
+      "loss": 0.0218,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3213572854291417,
+      "grad_norm": 1.1764345975362007,
+      "learning_rate": 7.936310061706965e-06,
+      "loss": 0.0399,
+      "step": 1127
+    },
+    {
+      "epoch": 0.32164242942686055,
+      "grad_norm": 0.8821811693487115,
+      "learning_rate": 7.932570504110697e-06,
+      "loss": 0.0252,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3219275734245794,
+      "grad_norm": 0.8404345204905768,
+      "learning_rate": 7.928828444240144e-06,
+      "loss": 0.0273,
+      "step": 1129
+    },
+    {
+      "epoch": 0.32221271742229823,
+      "grad_norm": 0.714789071782589,
+      "learning_rate": 7.925083885288296e-06,
+      "loss": 0.0165,
+      "step": 1130
+    },
+    {
+      "epoch": 0.32249786142001713,
+      "grad_norm": 0.9439440999089335,
+      "learning_rate": 7.921336830450268e-06,
+      "loss": 0.0174,
+      "step": 1131
+    },
+    {
+      "epoch": 0.32278300541773597,
+      "grad_norm": 1.9591792214351356,
+      "learning_rate": 7.917587282923312e-06,
+      "loss": 0.0492,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3230681494154548,
+      "grad_norm": 0.9159150616345992,
+      "learning_rate": 7.913835245906805e-06,
+      "loss": 0.0174,
+      "step": 1133
+    },
+    {
+      "epoch": 0.32335329341317365,
+      "grad_norm": 1.6772121272739662,
+      "learning_rate": 7.910080722602245e-06,
+      "loss": 0.0219,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3236384374108925,
+      "grad_norm": 1.1616225108993936,
+      "learning_rate": 7.906323716213256e-06,
+      "loss": 0.0263,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3239235814086113,
+      "grad_norm": 1.151082996421702,
+      "learning_rate": 7.902564229945577e-06,
+      "loss": 0.0111,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3242087254063302,
+      "grad_norm": 1.2986081859094945,
+      "learning_rate": 7.898802267007067e-06,
+      "loss": 0.0116,
+      "step": 1137
+    },
+    {
+      "epoch": 0.32449386940404906,
+      "grad_norm": 1.3947759655941292,
+      "learning_rate": 7.895037830607692e-06,
+      "loss": 0.0389,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3247790134017679,
+      "grad_norm": 1.0484963530466218,
+      "learning_rate": 7.891270923959537e-06,
+      "loss": 0.014,
+      "step": 1139
+    },
+    {
+      "epoch": 0.32506415739948674,
+      "grad_norm": 1.504826560429216,
+      "learning_rate": 7.887501550276789e-06,
+      "loss": 0.0423,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3253493013972056,
+      "grad_norm": 1.5608082869525692,
+      "learning_rate": 7.88372971277574e-06,
+      "loss": 0.0684,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3256344453949244,
+      "grad_norm": 2.7628731807480653,
+      "learning_rate": 7.879955414674784e-06,
+      "loss": 0.0681,
+      "step": 1142
+    },
+    {
+      "epoch": 0.32591958939264326,
+      "grad_norm": 1.613849056092281,
+      "learning_rate": 7.87617865919442e-06,
+      "loss": 0.0463,
+      "step": 1143
+    },
+    {
+      "epoch": 0.32620473339036216,
+      "grad_norm": 2.5686764919053124,
+      "learning_rate": 7.872399449557238e-06,
+      "loss": 0.0536,
+      "step": 1144
+    },
+    {
+      "epoch": 0.326489877388081,
+      "grad_norm": 1.4183432016049962,
+      "learning_rate": 7.868617788987925e-06,
+      "loss": 0.0285,
+      "step": 1145
+    },
+    {
+      "epoch": 0.32677502138579984,
+      "grad_norm": 1.7124522563822497,
+      "learning_rate": 7.864833680713256e-06,
+      "loss": 0.0593,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3270601653835187,
+      "grad_norm": 1.2283635139885853,
+      "learning_rate": 7.861047127962099e-06,
+      "loss": 0.0321,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3273453093812375,
+      "grad_norm": 1.6249903419870981,
+      "learning_rate": 7.857258133965405e-06,
+      "loss": 0.0399,
+      "step": 1148
+    },
+    {
+      "epoch": 0.32763045337895635,
+      "grad_norm": 3.053147088970876,
+      "learning_rate": 7.853466701956208e-06,
+      "loss": 0.0531,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3279155973766752,
+      "grad_norm": 24.31024864362681,
+      "learning_rate": 7.849672835169625e-06,
+      "loss": 0.1359,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3282007413743941,
+      "grad_norm": 1.696805065592631,
+      "learning_rate": 7.845876536842846e-06,
+      "loss": 0.0418,
+      "step": 1151
+    },
+    {
+      "epoch": 0.32848588537211293,
+      "grad_norm": 1.3445704598716866,
+      "learning_rate": 7.84207781021514e-06,
+      "loss": 0.0254,
+      "step": 1152
+    },
+    {
+      "epoch": 0.32877102936983177,
+      "grad_norm": 2.8671303035698616,
+      "learning_rate": 7.838276658527847e-06,
+      "loss": 0.0395,
+      "step": 1153
+    },
+    {
+      "epoch": 0.3290561733675506,
+      "grad_norm": 2.130710041966832,
+      "learning_rate": 7.834473085024373e-06,
+      "loss": 0.0316,
+      "step": 1154
+    },
+    {
+      "epoch": 0.32934131736526945,
+      "grad_norm": 2.6888945875232424,
+      "learning_rate": 7.830667092950195e-06,
+      "loss": 0.0586,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3296264613629883,
+      "grad_norm": 1.9325581252131516,
+      "learning_rate": 7.826858685552851e-06,
+      "loss": 0.0405,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3299116053607072,
+      "grad_norm": 2.1097020635632804,
+      "learning_rate": 7.82304786608194e-06,
+      "loss": 0.0237,
+      "step": 1157
+    },
+    {
+      "epoch": 0.330196749358426,
+      "grad_norm": 1.0799679655070387,
+      "learning_rate": 7.819234637789122e-06,
+      "loss": 0.0218,
+      "step": 1158
+    },
+    {
+      "epoch": 0.33048189335614486,
+      "grad_norm": 1.0940971424110126,
+      "learning_rate": 7.815419003928107e-06,
+      "loss": 0.0156,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3307670373538637,
+      "grad_norm": 0.45652875771745444,
+      "learning_rate": 7.811600967754661e-06,
+      "loss": 0.018,
+      "step": 1160
+    },
+    {
+      "epoch": 0.33105218135158254,
+      "grad_norm": 1.5961724407974645,
+      "learning_rate": 7.807780532526604e-06,
+      "loss": 0.0442,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3313373253493014,
+      "grad_norm": 1.5457552919670423,
+      "learning_rate": 7.80395770150379e-06,
+      "loss": 0.035,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3316224693470202,
+      "grad_norm": 1.5335975806195943,
+      "learning_rate": 7.800132477948137e-06,
+      "loss": 0.0428,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3319076133447391,
+      "grad_norm": 1.791656104690741,
+      "learning_rate": 7.796304865123583e-06,
+      "loss": 0.0402,
+      "step": 1164
+    },
+    {
+      "epoch": 0.33219275734245796,
+      "grad_norm": 0.8626638971656029,
+      "learning_rate": 7.79247486629612e-06,
+      "loss": 0.0207,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3324779013401768,
+      "grad_norm": 1.3786813993640084,
+      "learning_rate": 7.788642484733773e-06,
+      "loss": 0.0305,
+      "step": 1166
+    },
+    {
+      "epoch": 0.33276304533789564,
+      "grad_norm": 0.9962974441495369,
+      "learning_rate": 7.784807723706593e-06,
+      "loss": 0.0369,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3330481893356145,
+      "grad_norm": 0.8448490830062997,
+      "learning_rate": 7.780970586486668e-06,
+      "loss": 0.032,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 141.46865931940235,
+      "learning_rate": 7.777131076348115e-06,
+      "loss": 0.5834,
+      "step": 1169
+    },
+    {
+      "epoch": 0.33361847733105215,
+      "grad_norm": 0.6240936201732423,
+      "learning_rate": 7.773289196567066e-06,
+      "loss": 0.0202,
+      "step": 1170
+    },
+    {
+      "epoch": 0.33390362132877105,
+      "grad_norm": 1.6855103583591178,
+      "learning_rate": 7.76944495042169e-06,
+      "loss": 0.031,
+      "step": 1171
+    },
+    {
+      "epoch": 0.3341887653264899,
+      "grad_norm": 1.061334418928402,
+      "learning_rate": 7.76559834119216e-06,
+      "loss": 0.0393,
+      "step": 1172
+    },
+    {
+      "epoch": 0.33447390932420873,
+      "grad_norm": 2.3860632467930434,
+      "learning_rate": 7.761749372160676e-06,
+      "loss": 0.0546,
+      "step": 1173
+    },
+    {
+      "epoch": 0.33475905332192757,
+      "grad_norm": 0.7787717819480605,
+      "learning_rate": 7.757898046611446e-06,
+      "loss": 0.0196,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3350441973196464,
+      "grad_norm": 1.1796680979025957,
+      "learning_rate": 7.754044367830689e-06,
+      "loss": 0.038,
+      "step": 1175
+    },
+    {
+      "epoch": 0.33532934131736525,
+      "grad_norm": 1.3148557580390934,
+      "learning_rate": 7.750188339106635e-06,
+      "loss": 0.0304,
+      "step": 1176
+    },
+    {
+      "epoch": 0.33561448531508414,
+      "grad_norm": 1.2054539450580688,
+      "learning_rate": 7.746329963729517e-06,
+      "loss": 0.0228,
+      "step": 1177
+    },
+    {
+      "epoch": 0.335899629312803,
+      "grad_norm": 2.0449048447789386,
+      "learning_rate": 7.742469244991572e-06,
+      "loss": 0.0661,
+      "step": 1178
+    },
+    {
+      "epoch": 0.3361847733105218,
+      "grad_norm": 0.5740583915566436,
+      "learning_rate": 7.738606186187034e-06,
+      "loss": 0.0179,
+      "step": 1179
+    },
+    {
+      "epoch": 0.33646991730824066,
+      "grad_norm": 0.6910488636543247,
+      "learning_rate": 7.734740790612137e-06,
+      "loss": 0.0277,
+      "step": 1180
+    },
+    {
+      "epoch": 0.3367550613059595,
+      "grad_norm": 1.0694033044260076,
+      "learning_rate": 7.730873061565101e-06,
+      "loss": 0.0224,
+      "step": 1181
+    },
+    {
+      "epoch": 0.33704020530367834,
+      "grad_norm": 0.790204261875235,
+      "learning_rate": 7.72700300234615e-06,
+      "loss": 0.0248,
+      "step": 1182
+    },
+    {
+      "epoch": 0.3373253493013972,
+      "grad_norm": 1.33625424302328,
+      "learning_rate": 7.723130616257485e-06,
+      "loss": 0.0274,
+      "step": 1183
+    },
+    {
+      "epoch": 0.3376104932991161,
+      "grad_norm": 1.0567353013473344,
+      "learning_rate": 7.719255906603298e-06,
+      "loss": 0.0232,
+      "step": 1184
+    },
+    {
+      "epoch": 0.3378956372968349,
+      "grad_norm": 0.30339439930941897,
+      "learning_rate": 7.715378876689763e-06,
+      "loss": 0.0076,
+      "step": 1185
+    },
+    {
+      "epoch": 0.33818078129455376,
+      "grad_norm": 0.6213587414411561,
+      "learning_rate": 7.711499529825032e-06,
+      "loss": 0.0182,
+      "step": 1186
+    },
+    {
+      "epoch": 0.3384659252922726,
+      "grad_norm": 1.6039813029461647,
+      "learning_rate": 7.707617869319235e-06,
+      "loss": 0.041,
+      "step": 1187
+    },
+    {
+      "epoch": 0.33875106928999144,
+      "grad_norm": 0.9910996874197118,
+      "learning_rate": 7.703733898484479e-06,
+      "loss": 0.0186,
+      "step": 1188
+    },
+    {
+      "epoch": 0.3390362132877103,
+      "grad_norm": 0.4274479084393529,
+      "learning_rate": 7.699847620634834e-06,
+      "loss": 0.0073,
+      "step": 1189
+    },
+    {
+      "epoch": 0.3393213572854291,
+      "grad_norm": 1.1140118331424291,
+      "learning_rate": 7.695959039086349e-06,
+      "loss": 0.0169,
+      "step": 1190
+    },
+    {
+      "epoch": 0.339606501283148,
+      "grad_norm": 0.7561760159586366,
+      "learning_rate": 7.692068157157032e-06,
+      "loss": 0.0165,
+      "step": 1191
+    },
+    {
+      "epoch": 0.33989164528086685,
+      "grad_norm": 0.5463013328367404,
+      "learning_rate": 7.688174978166855e-06,
+      "loss": 0.0118,
+      "step": 1192
+    },
+    {
+      "epoch": 0.3401767892785857,
+      "grad_norm": 1.1686541927437784,
+      "learning_rate": 7.684279505437754e-06,
+      "loss": 0.0183,
+      "step": 1193
+    },
+    {
+      "epoch": 0.34046193327630453,
+      "grad_norm": 1.5635211370198485,
+      "learning_rate": 7.680381742293615e-06,
+      "loss": 0.0548,
+      "step": 1194
+    },
+    {
+      "epoch": 0.34074707727402337,
+      "grad_norm": 1.3537452233019234,
+      "learning_rate": 7.676481692060284e-06,
+      "loss": 0.024,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3410322212717422,
+      "grad_norm": 0.7266402871499255,
+      "learning_rate": 7.672579358065554e-06,
+      "loss": 0.0175,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3413173652694611,
+      "grad_norm": 0.5773819066167983,
+      "learning_rate": 7.668674743639173e-06,
+      "loss": 0.0158,
+      "step": 1197
+    },
+    {
+      "epoch": 0.34160250926717994,
+      "grad_norm": 1.1945971562545137,
+      "learning_rate": 7.66476785211283e-06,
+      "loss": 0.0232,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3418876532648988,
+      "grad_norm": 1.7947296038208627,
+      "learning_rate": 7.660858686820157e-06,
+      "loss": 0.0372,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3421727972626176,
+      "grad_norm": 1.2339145196734935,
+      "learning_rate": 7.656947251096729e-06,
+      "loss": 0.0301,
+      "step": 1200
+    },
+    {
+      "epoch": 0.34245794126033646,
+      "grad_norm": 1.4239632584351836,
+      "learning_rate": 7.653033548280056e-06,
+      "loss": 0.0223,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3427430852580553,
+      "grad_norm": 3.5661005357954902,
+      "learning_rate": 7.649117581709581e-06,
+      "loss": 0.0782,
+      "step": 1202
+    },
+    {
+      "epoch": 0.34302822925577414,
+      "grad_norm": 0.4606205342130907,
+      "learning_rate": 7.645199354726681e-06,
+      "loss": 0.008,
+      "step": 1203
+    },
+    {
+      "epoch": 0.34331337325349304,
+      "grad_norm": 1.1808740305202428,
+      "learning_rate": 7.641278870674664e-06,
+      "loss": 0.0138,
+      "step": 1204
+    },
+    {
+      "epoch": 0.3435985172512119,
+      "grad_norm": 1.181312784517819,
+      "learning_rate": 7.637356132898762e-06,
+      "loss": 0.0113,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3438836612489307,
+      "grad_norm": 1.9393291910241433,
+      "learning_rate": 7.633431144746123e-06,
+      "loss": 0.0366,
+      "step": 1206
+    },
+    {
+      "epoch": 0.34416880524664956,
+      "grad_norm": 0.5655309202758987,
+      "learning_rate": 7.629503909565829e-06,
+      "loss": 0.0084,
+      "step": 1207
+    },
+    {
+      "epoch": 0.3444539492443684,
+      "grad_norm": 0.9828332776151091,
+      "learning_rate": 7.625574430708867e-06,
+      "loss": 0.0436,
+      "step": 1208
+    },
+    {
+      "epoch": 0.34473909324208724,
+      "grad_norm": 1.9944966409431035,
+      "learning_rate": 7.621642711528143e-06,
+      "loss": 0.0303,
+      "step": 1209
+    },
+    {
+      "epoch": 0.3450242372398061,
+      "grad_norm": 0.7784849227021717,
+      "learning_rate": 7.617708755378477e-06,
+      "loss": 0.0121,
+      "step": 1210
+    },
+    {
+      "epoch": 0.34530938123752497,
+      "grad_norm": 1.4230792460322752,
+      "learning_rate": 7.613772565616595e-06,
+      "loss": 0.0213,
+      "step": 1211
+    },
+    {
+      "epoch": 0.3455945252352438,
+      "grad_norm": 1.1879362008117293,
+      "learning_rate": 7.609834145601129e-06,
+      "loss": 0.0331,
+      "step": 1212
+    },
+    {
+      "epoch": 0.34587966923296265,
+      "grad_norm": 0.9145028403418679,
+      "learning_rate": 7.605893498692616e-06,
+      "loss": 0.0103,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3461648132306815,
+      "grad_norm": 0.8655139809708533,
+      "learning_rate": 7.601950628253489e-06,
+      "loss": 0.023,
+      "step": 1214
+    },
+    {
+      "epoch": 0.34644995722840033,
+      "grad_norm": 1.7832506022566148,
+      "learning_rate": 7.598005537648082e-06,
+      "loss": 0.0564,
+      "step": 1215
+    },
+    {
+      "epoch": 0.34673510122611917,
+      "grad_norm": 2.1239358299514586,
+      "learning_rate": 7.594058230242623e-06,
+      "loss": 0.0366,
+      "step": 1216
+    },
+    {
+      "epoch": 0.34702024522383806,
+      "grad_norm": 0.878389399490451,
+      "learning_rate": 7.59010870940523e-06,
+      "loss": 0.0119,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3473053892215569,
+      "grad_norm": 2.733265706518621,
+      "learning_rate": 7.58615697850591e-06,
+      "loss": 0.0543,
+      "step": 1218
+    },
+    {
+      "epoch": 0.34759053321927574,
+      "grad_norm": 0.7518484980410628,
+      "learning_rate": 7.582203040916558e-06,
+      "loss": 0.0129,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3478756772169946,
+      "grad_norm": 0.9989613237234266,
+      "learning_rate": 7.578246900010948e-06,
+      "loss": 0.0128,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3481608212147134,
+      "grad_norm": 1.0141029301162927,
+      "learning_rate": 7.57428855916474e-06,
+      "loss": 0.0225,
+      "step": 1221
+    },
+    {
+      "epoch": 0.34844596521243226,
+      "grad_norm": 1.820213625141049,
+      "learning_rate": 7.5703280217554594e-06,
+      "loss": 0.0539,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3487311092101511,
+      "grad_norm": 0.4916304901754792,
+      "learning_rate": 7.566365291162523e-06,
+      "loss": 0.0087,
+      "step": 1223
+    },
+    {
+      "epoch": 0.34901625320787,
+      "grad_norm": 1.5289568499969866,
+      "learning_rate": 7.5624003707672036e-06,
+      "loss": 0.0262,
+      "step": 1224
+    },
+    {
+      "epoch": 0.34930139720558884,
+      "grad_norm": 0.6704716345946136,
+      "learning_rate": 7.558433263952652e-06,
+      "loss": 0.0263,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3495865412033077,
+      "grad_norm": 0.8476756802724102,
+      "learning_rate": 7.554463974103876e-06,
+      "loss": 0.0091,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3498716852010265,
+      "grad_norm": 0.5944495719636592,
+      "learning_rate": 7.5504925046077596e-06,
+      "loss": 0.0154,
+      "step": 1227
+    },
+    {
+      "epoch": 0.35015682919874536,
+      "grad_norm": 1.545818848015318,
+      "learning_rate": 7.546518858853032e-06,
+      "loss": 0.0341,
+      "step": 1228
+    },
+    {
+      "epoch": 0.3504419731964642,
+      "grad_norm": 1.9221079880766434,
+      "learning_rate": 7.542543040230287e-06,
+      "loss": 0.0332,
+      "step": 1229
+    },
+    {
+      "epoch": 0.35072711719418304,
+      "grad_norm": 1.9442964587621157,
+      "learning_rate": 7.538565052131972e-06,
+      "loss": 0.0424,
+      "step": 1230
+    },
+    {
+      "epoch": 0.35101226119190193,
+      "grad_norm": 2.1530989721825544,
+      "learning_rate": 7.534584897952385e-06,
+      "loss": 0.0644,
+      "step": 1231
+    },
+    {
+      "epoch": 0.35129740518962077,
+      "grad_norm": 1.0592053494006788,
+      "learning_rate": 7.530602581087672e-06,
+      "loss": 0.0418,
+      "step": 1232
+    },
+    {
+      "epoch": 0.3515825491873396,
+      "grad_norm": 2.701165505211902,
+      "learning_rate": 7.526618104935824e-06,
+      "loss": 0.0423,
+      "step": 1233
+    },
+    {
+      "epoch": 0.35186769318505845,
+      "grad_norm": 1.9681046663599653,
+      "learning_rate": 7.522631472896671e-06,
+      "loss": 0.0292,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3521528371827773,
+      "grad_norm": 1.0834902051848114,
+      "learning_rate": 7.518642688371893e-06,
+      "loss": 0.0165,
+      "step": 1235
+    },
+    {
+      "epoch": 0.35243798118049613,
+      "grad_norm": 0.729111577118215,
+      "learning_rate": 7.514651754764996e-06,
+      "loss": 0.0113,
+      "step": 1236
+    },
+    {
+      "epoch": 0.352723125178215,
+      "grad_norm": 0.9677835962247436,
+      "learning_rate": 7.510658675481324e-06,
+      "loss": 0.0089,
+      "step": 1237
+    },
+    {
+      "epoch": 0.35300826917593386,
+      "grad_norm": 1.7687113644828218,
+      "learning_rate": 7.5066634539280524e-06,
+      "loss": 0.0368,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3532934131736527,
+      "grad_norm": 0.8467747355722813,
+      "learning_rate": 7.502666093514184e-06,
+      "loss": 0.0212,
+      "step": 1239
+    },
+    {
+      "epoch": 0.35357855717137154,
+      "grad_norm": 1.1715828228303156,
+      "learning_rate": 7.498666597650544e-06,
+      "loss": 0.0316,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3538637011690904,
+      "grad_norm": 0.9081021165247576,
+      "learning_rate": 7.494664969749785e-06,
+      "loss": 0.0227,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3541488451668092,
+      "grad_norm": 1.1369108712808864,
+      "learning_rate": 7.490661213226374e-06,
+      "loss": 0.0359,
+      "step": 1242
+    },
+    {
+      "epoch": 0.35443398916452806,
+      "grad_norm": 2.0527453246581557,
+      "learning_rate": 7.486655331496597e-06,
+      "loss": 0.0355,
+      "step": 1243
+    },
+    {
+      "epoch": 0.35471913316224696,
+      "grad_norm": 1.0630257009017088,
+      "learning_rate": 7.482647327978551e-06,
+      "loss": 0.022,
+      "step": 1244
+    },
+    {
+      "epoch": 0.3550042771599658,
+      "grad_norm": 1.1552022033934928,
+      "learning_rate": 7.478637206092147e-06,
+      "loss": 0.0244,
+      "step": 1245
+    },
+    {
+      "epoch": 0.35528942115768464,
+      "grad_norm": 1.2041123400974558,
+      "learning_rate": 7.474624969259101e-06,
+      "loss": 0.0311,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3555745651554035,
+      "grad_norm": 0.6912486293307928,
+      "learning_rate": 7.470610620902934e-06,
+      "loss": 0.016,
+      "step": 1247
+    },
+    {
+      "epoch": 0.3558597091531223,
+      "grad_norm": 0.7069761757438487,
+      "learning_rate": 7.466594164448967e-06,
+      "loss": 0.0121,
+      "step": 1248
+    },
+    {
+      "epoch": 0.35614485315084116,
+      "grad_norm": 2.406861730589254,
+      "learning_rate": 7.462575603324325e-06,
+      "loss": 0.0573,
+      "step": 1249
+    },
+    {
+      "epoch": 0.35642999714856,
+      "grad_norm": 1.073491482084253,
+      "learning_rate": 7.458554940957922e-06,
+      "loss": 0.0285,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3567151411462789,
+      "grad_norm": 1.4809174105565874,
+      "learning_rate": 7.45453218078047e-06,
+      "loss": 0.0187,
+      "step": 1251
+    },
+    {
+      "epoch": 0.35700028514399773,
+      "grad_norm": 2.5263068063454104,
+      "learning_rate": 7.450507326224469e-06,
+      "loss": 0.0346,
+      "step": 1252
+    },
+    {
+      "epoch": 0.35728542914171657,
+      "grad_norm": 1.0743279609219285,
+      "learning_rate": 7.446480380724208e-06,
+      "loss": 0.015,
+      "step": 1253
+    },
+    {
+      "epoch": 0.3575705731394354,
+      "grad_norm": 1.2267919789411827,
+      "learning_rate": 7.442451347715758e-06,
+      "loss": 0.041,
+      "step": 1254
+    },
+    {
+      "epoch": 0.35785571713715425,
+      "grad_norm": 1.0637373721906973,
+      "learning_rate": 7.438420230636969e-06,
+      "loss": 0.0161,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3581408611348731,
+      "grad_norm": 1.8824790740102795,
+      "learning_rate": 7.434387032927475e-06,
+      "loss": 0.0358,
+      "step": 1256
+    },
+    {
+      "epoch": 0.358426005132592,
+      "grad_norm": 1.2092687912850297,
+      "learning_rate": 7.430351758028682e-06,
+      "loss": 0.0301,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3587111491303108,
+      "grad_norm": 0.6835211094012288,
+      "learning_rate": 7.426314409383768e-06,
+      "loss": 0.0062,
+      "step": 1258
+    },
+    {
+      "epoch": 0.35899629312802966,
+      "grad_norm": 1.3255750122183771,
+      "learning_rate": 7.422274990437682e-06,
+      "loss": 0.0318,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3592814371257485,
+      "grad_norm": 0.9156502566543411,
+      "learning_rate": 7.418233504637138e-06,
+      "loss": 0.0143,
+      "step": 1260
+    },
+    {
+      "epoch": 0.35956658112346734,
+      "grad_norm": 1.1727165754738422,
+      "learning_rate": 7.414189955430615e-06,
+      "loss": 0.0369,
+      "step": 1261
+    },
+    {
+      "epoch": 0.3598517251211862,
+      "grad_norm": 1.4468665217354895,
+      "learning_rate": 7.410144346268351e-06,
+      "loss": 0.0417,
+      "step": 1262
+    },
+    {
+      "epoch": 0.360136869118905,
+      "grad_norm": 1.824457712002065,
+      "learning_rate": 7.4060966806023445e-06,
+      "loss": 0.0353,
+      "step": 1263
+    },
+    {
+      "epoch": 0.3604220131166239,
+      "grad_norm": 0.9767535191086671,
+      "learning_rate": 7.4020469618863455e-06,
+      "loss": 0.021,
+      "step": 1264
+    },
+    {
+      "epoch": 0.36070715711434276,
+      "grad_norm": 1.079593807537144,
+      "learning_rate": 7.3979951935758596e-06,
+      "loss": 0.0178,
+      "step": 1265
+    },
+    {
+      "epoch": 0.3609923011120616,
+      "grad_norm": 1.6139308826860321,
+      "learning_rate": 7.393941379128136e-06,
+      "loss": 0.0293,
+      "step": 1266
+    },
+    {
+      "epoch": 0.36127744510978044,
+      "grad_norm": 1.1527677219155397,
+      "learning_rate": 7.3898855220021734e-06,
+      "loss": 0.032,
+      "step": 1267
+    },
+    {
+      "epoch": 0.3615625891074993,
+      "grad_norm": 1.5856659207811066,
+      "learning_rate": 7.385827625658713e-06,
+      "loss": 0.0261,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3618477331052181,
+      "grad_norm": 1.086362153999573,
+      "learning_rate": 7.3817676935602376e-06,
+      "loss": 0.0128,
+      "step": 1269
+    },
+    {
+      "epoch": 0.36213287710293696,
+      "grad_norm": 1.6988841433717032,
+      "learning_rate": 7.377705729170962e-06,
+      "loss": 0.0371,
+      "step": 1270
+    },
+    {
+      "epoch": 0.36241802110065585,
+      "grad_norm": 1.4325229855389487,
+      "learning_rate": 7.373641735956843e-06,
+      "loss": 0.0227,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3627031650983747,
+      "grad_norm": 0.8830703221340672,
+      "learning_rate": 7.369575717385557e-06,
+      "loss": 0.019,
+      "step": 1272
+    },
+    {
+      "epoch": 0.36298830909609353,
+      "grad_norm": 2.209036557846552,
+      "learning_rate": 7.365507676926523e-06,
+      "loss": 0.0346,
+      "step": 1273
+    },
+    {
+      "epoch": 0.36327345309381237,
+      "grad_norm": 1.5334503346203225,
+      "learning_rate": 7.361437618050873e-06,
+      "loss": 0.0407,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3635585970915312,
+      "grad_norm": 0.8070701947902329,
+      "learning_rate": 7.3573655442314674e-06,
+      "loss": 0.0324,
+      "step": 1275
+    },
+    {
+      "epoch": 0.36384374108925005,
+      "grad_norm": 1.430194693309841,
+      "learning_rate": 7.353291458942884e-06,
+      "loss": 0.0196,
+      "step": 1276
+    },
+    {
+      "epoch": 0.36412888508696895,
+      "grad_norm": 0.3562037606493946,
+      "learning_rate": 7.349215365661417e-06,
+      "loss": 0.0126,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3644140290846878,
+      "grad_norm": 1.1955582337051391,
+      "learning_rate": 7.345137267865075e-06,
+      "loss": 0.0287,
+      "step": 1278
+    },
+    {
+      "epoch": 0.3646991730824066,
+      "grad_norm": 1.5568689189039489,
+      "learning_rate": 7.341057169033576e-06,
+      "loss": 0.0283,
+      "step": 1279
+    },
+    {
+      "epoch": 0.36498431708012546,
+      "grad_norm": 1.3478008946354623,
+      "learning_rate": 7.336975072648346e-06,
+      "loss": 0.0288,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3652694610778443,
+      "grad_norm": 1.0580096876065161,
+      "learning_rate": 7.332890982192514e-06,
+      "loss": 0.0166,
+      "step": 1281
+    },
+    {
+      "epoch": 0.36555460507556314,
+      "grad_norm": 0.5686865684902279,
+      "learning_rate": 7.328804901150914e-06,
+      "loss": 0.0116,
+      "step": 1282
+    },
+    {
+      "epoch": 0.365839749073282,
+      "grad_norm": 1.716366052959552,
+      "learning_rate": 7.324716833010074e-06,
+      "loss": 0.0467,
+      "step": 1283
+    },
+    {
+      "epoch": 0.3661248930710009,
+      "grad_norm": 0.7791990463845944,
+      "learning_rate": 7.32062678125822e-06,
+      "loss": 0.0111,
+      "step": 1284
+    },
+    {
+      "epoch": 0.3664100370687197,
+      "grad_norm": 1.3893700452732172,
+      "learning_rate": 7.316534749385272e-06,
+      "loss": 0.0194,
+      "step": 1285
+    },
+    {
+      "epoch": 0.36669518106643856,
+      "grad_norm": 1.5762832721071067,
+      "learning_rate": 7.312440740882836e-06,
+      "loss": 0.0238,
+      "step": 1286
+    },
+    {
+      "epoch": 0.3669803250641574,
+      "grad_norm": 1.0209692180493717,
+      "learning_rate": 7.308344759244208e-06,
+      "loss": 0.0154,
+      "step": 1287
+    },
+    {
+      "epoch": 0.36726546906187624,
+      "grad_norm": 0.3701931947426716,
+      "learning_rate": 7.304246807964363e-06,
+      "loss": 0.0061,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3675506130595951,
+      "grad_norm": 0.9193845325767704,
+      "learning_rate": 7.300146890539962e-06,
+      "loss": 0.0277,
+      "step": 1289
+    },
+    {
+      "epoch": 0.3678357570573139,
+      "grad_norm": 1.8222398498763657,
+      "learning_rate": 7.2960450104693415e-06,
+      "loss": 0.04,
+      "step": 1290
+    },
+    {
+      "epoch": 0.3681209010550328,
+      "grad_norm": 0.643942314383188,
+      "learning_rate": 7.291941171252512e-06,
+      "loss": 0.0084,
+      "step": 1291
+    },
+    {
+      "epoch": 0.36840604505275165,
+      "grad_norm": 2.2276882078911346,
+      "learning_rate": 7.287835376391157e-06,
+      "loss": 0.0351,
+      "step": 1292
+    },
+    {
+      "epoch": 0.3686911890504705,
+      "grad_norm": 1.0734551871869193,
+      "learning_rate": 7.283727629388628e-06,
+      "loss": 0.0255,
+      "step": 1293
+    },
+    {
+      "epoch": 0.36897633304818933,
+      "grad_norm": 0.9919685813925658,
+      "learning_rate": 7.279617933749937e-06,
+      "loss": 0.0188,
+      "step": 1294
+    },
+    {
+      "epoch": 0.36926147704590817,
+      "grad_norm": 1.6546322046981001,
+      "learning_rate": 7.275506292981771e-06,
+      "loss": 0.032,
+      "step": 1295
+    },
+    {
+      "epoch": 0.369546621043627,
+      "grad_norm": 0.6427488874964754,
+      "learning_rate": 7.271392710592466e-06,
+      "loss": 0.0183,
+      "step": 1296
+    },
+    {
+      "epoch": 0.3698317650413459,
+      "grad_norm": 2.1105811666825605,
+      "learning_rate": 7.2672771900920195e-06,
+      "loss": 0.0342,
+      "step": 1297
+    },
+    {
+      "epoch": 0.37011690903906475,
+      "grad_norm": 1.7984565917708282,
+      "learning_rate": 7.263159734992079e-06,
+      "loss": 0.0358,
+      "step": 1298
+    },
+    {
+      "epoch": 0.3704020530367836,
+      "grad_norm": 2.0892827974495973,
+      "learning_rate": 7.259040348805948e-06,
+      "loss": 0.0637,
+      "step": 1299
+    },
+    {
+      "epoch": 0.3706871970345024,
+      "grad_norm": 1.9245223100243691,
+      "learning_rate": 7.2549190350485734e-06,
+      "loss": 0.0256,
+      "step": 1300
+    },
+    {
+      "epoch": 0.37097234103222126,
+      "grad_norm": 1.332108407780507,
+      "learning_rate": 7.250795797236549e-06,
+      "loss": 0.0288,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3712574850299401,
+      "grad_norm": 2.832303589679555,
+      "learning_rate": 7.246670638888109e-06,
+      "loss": 0.0746,
+      "step": 1302
+    },
+    {
+      "epoch": 0.37154262902765894,
+      "grad_norm": 1.105425246209834,
+      "learning_rate": 7.242543563523128e-06,
+      "loss": 0.0386,
+      "step": 1303
+    },
+    {
+      "epoch": 0.37182777302537784,
+      "grad_norm": 1.228230849256297,
+      "learning_rate": 7.238414574663115e-06,
+      "loss": 0.0233,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3721129170230967,
+      "grad_norm": 1.4222277233624963,
+      "learning_rate": 7.234283675831212e-06,
+      "loss": 0.0163,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3723980610208155,
+      "grad_norm": 1.6737164988568887,
+      "learning_rate": 7.230150870552191e-06,
+      "loss": 0.0436,
+      "step": 1306
+    },
+    {
+      "epoch": 0.37268320501853436,
+      "grad_norm": 0.6672570457070205,
+      "learning_rate": 7.22601616235245e-06,
+      "loss": 0.0154,
+      "step": 1307
+    },
+    {
+      "epoch": 0.3729683490162532,
+      "grad_norm": 0.5059774786468529,
+      "learning_rate": 7.221879554760012e-06,
+      "loss": 0.0119,
+      "step": 1308
+    },
+    {
+      "epoch": 0.37325349301397204,
+      "grad_norm": 2.10429975207149,
+      "learning_rate": 7.217741051304519e-06,
+      "loss": 0.0672,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3735386370116909,
+      "grad_norm": 2.0092255665050374,
+      "learning_rate": 7.213600655517233e-06,
+      "loss": 0.0542,
+      "step": 1310
+    },
+    {
+      "epoch": 0.3738237810094098,
+      "grad_norm": 1.352010356984478,
+      "learning_rate": 7.209458370931029e-06,
+      "loss": 0.0142,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3741089250071286,
+      "grad_norm": 1.3811098769189896,
+      "learning_rate": 7.205314201080392e-06,
+      "loss": 0.0243,
+      "step": 1312
+    },
+    {
+      "epoch": 0.37439406900484745,
+      "grad_norm": 1.1209070881289154,
+      "learning_rate": 7.201168149501421e-06,
+      "loss": 0.0389,
+      "step": 1313
+    },
+    {
+      "epoch": 0.3746792130025663,
+      "grad_norm": 2.065215176648648,
+      "learning_rate": 7.197020219731814e-06,
+      "loss": 0.0266,
+      "step": 1314
+    },
+    {
+      "epoch": 0.37496435700028513,
+      "grad_norm": 2.5480094718136703,
+      "learning_rate": 7.192870415310877e-06,
+      "loss": 0.0328,
+      "step": 1315
+    },
+    {
+      "epoch": 0.37524950099800397,
+      "grad_norm": 0.7872176078118113,
+      "learning_rate": 7.188718739779511e-06,
+      "loss": 0.0139,
+      "step": 1316
+    },
+    {
+      "epoch": 0.37553464499572287,
+      "grad_norm": 0.6808650456171224,
+      "learning_rate": 7.1845651966802184e-06,
+      "loss": 0.0234,
+      "step": 1317
+    },
+    {
+      "epoch": 0.3758197889934417,
+      "grad_norm": 0.49101205287280486,
+      "learning_rate": 7.1804097895570924e-06,
+      "loss": 0.0178,
+      "step": 1318
+    },
+    {
+      "epoch": 0.37610493299116055,
+      "grad_norm": 1.0821122693778222,
+      "learning_rate": 7.176252521955817e-06,
+      "loss": 0.025,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3763900769888794,
+      "grad_norm": 1.6304437741369961,
+      "learning_rate": 7.172093397423663e-06,
+      "loss": 0.0422,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3766752209865982,
+      "grad_norm": 0.7386475855241287,
+      "learning_rate": 7.167932419509485e-06,
+      "loss": 0.0249,
+      "step": 1321
+    },
+    {
+      "epoch": 0.37696036498431706,
+      "grad_norm": 1.2169154892764618,
+      "learning_rate": 7.163769591763723e-06,
+      "loss": 0.045,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3772455089820359,
+      "grad_norm": 1.780773790899091,
+      "learning_rate": 7.159604917738392e-06,
+      "loss": 0.0457,
+      "step": 1323
+    },
+    {
+      "epoch": 0.3775306529797548,
+      "grad_norm": 1.3352312459837943,
+      "learning_rate": 7.155438400987083e-06,
+      "loss": 0.0429,
+      "step": 1324
+    },
+    {
+      "epoch": 0.37781579697747364,
+      "grad_norm": 1.1305902394114138,
+      "learning_rate": 7.151270045064958e-06,
+      "loss": 0.0341,
+      "step": 1325
+    },
+    {
+      "epoch": 0.3781009409751925,
+      "grad_norm": 1.126670135618066,
+      "learning_rate": 7.147099853528753e-06,
+      "loss": 0.0199,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3783860849729113,
+      "grad_norm": 1.137621576993765,
+      "learning_rate": 7.142927829936766e-06,
+      "loss": 0.0518,
+      "step": 1327
+    },
+    {
+      "epoch": 0.37867122897063016,
+      "grad_norm": 1.1773282692234626,
+      "learning_rate": 7.138753977848858e-06,
+      "loss": 0.0279,
+      "step": 1328
+    },
+    {
+      "epoch": 0.378956372968349,
+      "grad_norm": 0.5840674438803083,
+      "learning_rate": 7.134578300826452e-06,
+      "loss": 0.0224,
+      "step": 1329
+    },
+    {
+      "epoch": 0.37924151696606784,
+      "grad_norm": 1.1129186349385658,
+      "learning_rate": 7.130400802432529e-06,
+      "loss": 0.0322,
+      "step": 1330
+    },
+    {
+      "epoch": 0.37952666096378673,
+      "grad_norm": 1.8462980707500372,
+      "learning_rate": 7.12622148623162e-06,
+      "loss": 0.0342,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3798118049615056,
+      "grad_norm": 2.107804657803562,
+      "learning_rate": 7.122040355789815e-06,
+      "loss": 0.0427,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3800969489592244,
+      "grad_norm": 0.7450612627575915,
+      "learning_rate": 7.117857414674741e-06,
+      "loss": 0.0197,
+      "step": 1333
+    },
+    {
+      "epoch": 0.38038209295694325,
+      "grad_norm": 1.1691762047073135,
+      "learning_rate": 7.11367266645558e-06,
+      "loss": 0.0276,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3806672369546621,
+      "grad_norm": 1.0971833274442095,
+      "learning_rate": 7.1094861147030514e-06,
+      "loss": 0.0326,
+      "step": 1335
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 1.5810742872190795,
+      "learning_rate": 7.105297762989413e-06,
+      "loss": 0.038,
+      "step": 1336
+    },
+    {
+      "epoch": 0.3812375249500998,
+      "grad_norm": 1.801470439156759,
+      "learning_rate": 7.101107614888458e-06,
+      "loss": 0.0214,
+      "step": 1337
+    },
+    {
+      "epoch": 0.38152266894781867,
+      "grad_norm": 0.7796639176297265,
+      "learning_rate": 7.096915673975517e-06,
+      "loss": 0.0145,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3818078129455375,
+      "grad_norm": 0.3858558390410885,
+      "learning_rate": 7.092721943827446e-06,
+      "loss": 0.0105,
+      "step": 1339
+    },
+    {
+      "epoch": 0.38209295694325635,
+      "grad_norm": 1.033014186947543,
+      "learning_rate": 7.088526428022628e-06,
+      "loss": 0.0281,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3823781009409752,
+      "grad_norm": 1.9224999850785538,
+      "learning_rate": 7.084329130140972e-06,
+      "loss": 0.0465,
+      "step": 1341
+    },
+    {
+      "epoch": 0.382663244938694,
+      "grad_norm": 1.6708579732466118,
+      "learning_rate": 7.080130053763906e-06,
+      "loss": 0.0379,
+      "step": 1342
+    },
+    {
+      "epoch": 0.38294838893641286,
+      "grad_norm": 0.8987241866315161,
+      "learning_rate": 7.075929202474374e-06,
+      "loss": 0.0251,
+      "step": 1343
+    },
+    {
+      "epoch": 0.38323353293413176,
+      "grad_norm": 1.339937928755385,
+      "learning_rate": 7.071726579856838e-06,
+      "loss": 0.0561,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3835186769318506,
+      "grad_norm": 0.958679031321206,
+      "learning_rate": 7.067522189497269e-06,
+      "loss": 0.041,
+      "step": 1345
+    },
+    {
+      "epoch": 0.38380382092956944,
+      "grad_norm": 1.3336443045217854,
+      "learning_rate": 7.063316034983146e-06,
+      "loss": 0.0176,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3840889649272883,
+      "grad_norm": 0.5339713884516041,
+      "learning_rate": 7.059108119903455e-06,
+      "loss": 0.0127,
+      "step": 1347
+    },
+    {
+      "epoch": 0.3843741089250071,
+      "grad_norm": 0.4258851843158557,
+      "learning_rate": 7.054898447848684e-06,
+      "loss": 0.0169,
+      "step": 1348
+    },
+    {
+      "epoch": 0.38465925292272596,
+      "grad_norm": 1.1816276456155221,
+      "learning_rate": 7.050687022410819e-06,
+      "loss": 0.0324,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3849443969204448,
+      "grad_norm": 0.8946255755547217,
+      "learning_rate": 7.0464738471833436e-06,
+      "loss": 0.0326,
+      "step": 1350
+    },
+    {
+      "epoch": 0.3852295409181637,
+      "grad_norm": 1.1202911450635167,
+      "learning_rate": 7.042258925761233e-06,
+      "loss": 0.0228,
+      "step": 1351
+    },
+    {
+      "epoch": 0.38551468491588253,
+      "grad_norm": 1.6902455592606156,
+      "learning_rate": 7.038042261740952e-06,
+      "loss": 0.0473,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3857998289136014,
+      "grad_norm": 0.5948513631594184,
+      "learning_rate": 7.033823858720454e-06,
+      "loss": 0.02,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3860849729113202,
+      "grad_norm": 1.7951572919018244,
+      "learning_rate": 7.029603720299178e-06,
+      "loss": 0.0324,
+      "step": 1354
+    },
+    {
+      "epoch": 0.38637011690903905,
+      "grad_norm": 2.0429468456038555,
+      "learning_rate": 7.025381850078037e-06,
+      "loss": 0.0345,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3866552609067579,
+      "grad_norm": 0.6318621960506157,
+      "learning_rate": 7.021158251659429e-06,
+      "loss": 0.0163,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3869404049044768,
+      "grad_norm": 1.4457221608558466,
+      "learning_rate": 7.0169329286472235e-06,
+      "loss": 0.0377,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3872255489021956,
+      "grad_norm": 1.01032061892398,
+      "learning_rate": 7.01270588464676e-06,
+      "loss": 0.0219,
+      "step": 1358
+    },
+    {
+      "epoch": 0.38751069289991447,
+      "grad_norm": 0.748496429252588,
+      "learning_rate": 7.008477123264849e-06,
+      "loss": 0.0324,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3877958368976333,
+      "grad_norm": 1.1526332961541026,
+      "learning_rate": 7.004246648109765e-06,
+      "loss": 0.0254,
+      "step": 1360
+    },
+    {
+      "epoch": 0.38808098089535215,
+      "grad_norm": 1.4294101465113094,
+      "learning_rate": 7.000014462791245e-06,
+      "loss": 0.024,
+      "step": 1361
+    },
+    {
+      "epoch": 0.388366124893071,
+      "grad_norm": 0.5519193648295497,
+      "learning_rate": 6.995780570920488e-06,
+      "loss": 0.0244,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3886512688907898,
+      "grad_norm": 1.054024957106729,
+      "learning_rate": 6.991544976110144e-06,
+      "loss": 0.0187,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3889364128885087,
+      "grad_norm": 1.2014758092356506,
+      "learning_rate": 6.98730768197432e-06,
+      "loss": 0.0396,
+      "step": 1364
+    },
+    {
+      "epoch": 0.38922155688622756,
+      "grad_norm": 0.7448263297295125,
+      "learning_rate": 6.9830686921285724e-06,
+      "loss": 0.0155,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3895067008839464,
+      "grad_norm": 0.844577177735389,
+      "learning_rate": 6.978828010189903e-06,
+      "loss": 0.0129,
+      "step": 1366
+    },
+    {
+      "epoch": 0.38979184488166524,
+      "grad_norm": 0.818598787369236,
+      "learning_rate": 6.974585639776757e-06,
+      "loss": 0.0155,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3900769888793841,
+      "grad_norm": 0.6332227156256431,
+      "learning_rate": 6.970341584509025e-06,
+      "loss": 0.0143,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3903621328771029,
+      "grad_norm": 1.3864888311347223,
+      "learning_rate": 6.966095848008028e-06,
+      "loss": 0.0232,
+      "step": 1369
+    },
+    {
+      "epoch": 0.39064727687482176,
+      "grad_norm": 0.807302899981098,
+      "learning_rate": 6.9618484338965274e-06,
+      "loss": 0.0136,
+      "step": 1370
+    },
+    {
+      "epoch": 0.39093242087254065,
+      "grad_norm": 0.7532233628664813,
+      "learning_rate": 6.957599345798714e-06,
+      "loss": 0.0121,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3912175648702595,
+      "grad_norm": 0.8057671659634474,
+      "learning_rate": 6.953348587340205e-06,
+      "loss": 0.0078,
+      "step": 1372
+    },
+    {
+      "epoch": 0.39150270886797833,
+      "grad_norm": 1.6056173303746508,
+      "learning_rate": 6.949096162148048e-06,
+      "loss": 0.0255,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3917878528656972,
+      "grad_norm": 0.921091363738595,
+      "learning_rate": 6.944842073850709e-06,
+      "loss": 0.0411,
+      "step": 1374
+    },
+    {
+      "epoch": 0.392072996863416,
+      "grad_norm": 1.0604866928198136,
+      "learning_rate": 6.94058632607807e-06,
+      "loss": 0.0129,
+      "step": 1375
+    },
+    {
+      "epoch": 0.39235814086113485,
+      "grad_norm": 1.0589721417569484,
+      "learning_rate": 6.9363289224614395e-06,
+      "loss": 0.0358,
+      "step": 1376
+    },
+    {
+      "epoch": 0.39264328485885375,
+      "grad_norm": 0.7464112469258565,
+      "learning_rate": 6.932069866633524e-06,
+      "loss": 0.015,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3929284288565726,
+      "grad_norm": 0.911581766103546,
+      "learning_rate": 6.927809162228456e-06,
+      "loss": 0.0067,
+      "step": 1378
+    },
+    {
+      "epoch": 0.3932135728542914,
+      "grad_norm": 1.9277059427524286,
+      "learning_rate": 6.923546812881759e-06,
+      "loss": 0.0528,
+      "step": 1379
+    },
+    {
+      "epoch": 0.39349871685201027,
+      "grad_norm": 1.1879086450728118,
+      "learning_rate": 6.919282822230372e-06,
+      "loss": 0.0475,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3937838608497291,
+      "grad_norm": 1.0922499341160856,
+      "learning_rate": 6.91501719391263e-06,
+      "loss": 0.0235,
+      "step": 1381
+    },
+    {
+      "epoch": 0.39406900484744795,
+      "grad_norm": 1.0937840797322815,
+      "learning_rate": 6.910749931568265e-06,
+      "loss": 0.0235,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3943541488451668,
+      "grad_norm": 0.7304053983961637,
+      "learning_rate": 6.906481038838401e-06,
+      "loss": 0.0258,
+      "step": 1383
+    },
+    {
+      "epoch": 0.3946392928428857,
+      "grad_norm": 0.937712719820358,
+      "learning_rate": 6.902210519365561e-06,
+      "loss": 0.0246,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3949244368406045,
+      "grad_norm": 1.4809251423169405,
+      "learning_rate": 6.897938376793646e-06,
+      "loss": 0.0474,
+      "step": 1385
+    },
+    {
+      "epoch": 0.39520958083832336,
+      "grad_norm": 0.6276884351320375,
+      "learning_rate": 6.89366461476795e-06,
+      "loss": 0.0167,
+      "step": 1386
+    },
+    {
+      "epoch": 0.3954947248360422,
+      "grad_norm": 1.567668460942998,
+      "learning_rate": 6.889389236935145e-06,
+      "loss": 0.0301,
+      "step": 1387
+    },
+    {
+      "epoch": 0.39577986883376104,
+      "grad_norm": 1.7706303043509528,
+      "learning_rate": 6.885112246943282e-06,
+      "loss": 0.0306,
+      "step": 1388
+    },
+    {
+      "epoch": 0.3960650128314799,
+      "grad_norm": 1.4185752918637853,
+      "learning_rate": 6.880833648441788e-06,
+      "loss": 0.0288,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3963501568291987,
+      "grad_norm": 1.8432413219321235,
+      "learning_rate": 6.876553445081463e-06,
+      "loss": 0.0333,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3966353008269176,
+      "grad_norm": 0.7805065400041529,
+      "learning_rate": 6.872271640514475e-06,
+      "loss": 0.0211,
+      "step": 1391
+    },
+    {
+      "epoch": 0.39692044482463645,
+      "grad_norm": 1.4299986011501835,
+      "learning_rate": 6.867988238394361e-06,
+      "loss": 0.018,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3972055888223553,
+      "grad_norm": 0.8738977526847553,
+      "learning_rate": 6.863703242376016e-06,
+      "loss": 0.0213,
+      "step": 1393
+    },
+    {
+      "epoch": 0.39749073282007413,
+      "grad_norm": 1.883407055792336,
+      "learning_rate": 6.859416656115702e-06,
+      "loss": 0.0263,
+      "step": 1394
+    },
+    {
+      "epoch": 0.397775876817793,
+      "grad_norm": 0.42498552122724303,
+      "learning_rate": 6.855128483271033e-06,
+      "loss": 0.0159,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3980610208155118,
+      "grad_norm": 0.6640145903003858,
+      "learning_rate": 6.850838727500977e-06,
+      "loss": 0.0241,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3983461648132307,
+      "grad_norm": 1.7915114587711756,
+      "learning_rate": 6.846547392465854e-06,
+      "loss": 0.0256,
+      "step": 1397
+    },
+    {
+      "epoch": 0.39863130881094955,
+      "grad_norm": 2.294360606894688,
+      "learning_rate": 6.8422544818273336e-06,
+      "loss": 0.0479,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3989164528086684,
+      "grad_norm": 0.8371789914473499,
+      "learning_rate": 6.837959999248423e-06,
+      "loss": 0.019,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3992015968063872,
+      "grad_norm": 0.7596077582888351,
+      "learning_rate": 6.833663948393479e-06,
+      "loss": 0.0183,
+      "step": 1400
+    },
+    {
+      "epoch": 0.39948674080410607,
+      "grad_norm": 1.3500494545169412,
+      "learning_rate": 6.829366332928191e-06,
+      "loss": 0.0322,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3997718848018249,
+      "grad_norm": 0.5127260480269076,
+      "learning_rate": 6.825067156519584e-06,
+      "loss": 0.0158,
+      "step": 1402
+    },
+    {
+      "epoch": 0.40005702879954375,
+      "grad_norm": 0.6087593590841202,
+      "learning_rate": 6.82076642283602e-06,
+      "loss": 0.01,
+      "step": 1403
+    },
+    {
+      "epoch": 0.40034217279726264,
+      "grad_norm": 1.174996634641959,
+      "learning_rate": 6.816464135547183e-06,
+      "loss": 0.0217,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4006273167949815,
+      "grad_norm": 0.8541388840281969,
+      "learning_rate": 6.812160298324083e-06,
+      "loss": 0.0153,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4009124607927003,
+      "grad_norm": 1.0410711202292575,
+      "learning_rate": 6.807854914839061e-06,
+      "loss": 0.0274,
+      "step": 1406
+    },
+    {
+      "epoch": 0.40119760479041916,
+      "grad_norm": 1.2608003422751877,
+      "learning_rate": 6.803547988765765e-06,
+      "loss": 0.0348,
+      "step": 1407
+    },
+    {
+      "epoch": 0.401482748788138,
+      "grad_norm": 0.8195812427271472,
+      "learning_rate": 6.799239523779171e-06,
+      "loss": 0.0203,
+      "step": 1408
+    },
+    {
+      "epoch": 0.40176789278585684,
+      "grad_norm": 0.6649371018762259,
+      "learning_rate": 6.794929523555559e-06,
+      "loss": 0.0141,
+      "step": 1409
+    },
+    {
+      "epoch": 0.4020530367835757,
+      "grad_norm": 0.38933100823711925,
+      "learning_rate": 6.790617991772525e-06,
+      "loss": 0.0063,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4023381807812946,
+      "grad_norm": 1.8663796814203395,
+      "learning_rate": 6.786304932108967e-06,
+      "loss": 0.0551,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4026233247790134,
+      "grad_norm": 2.198862423064739,
+      "learning_rate": 6.781990348245092e-06,
+      "loss": 0.0528,
+      "step": 1412
+    },
+    {
+      "epoch": 0.40290846877673225,
+      "grad_norm": 1.5778666491396456,
+      "learning_rate": 6.7776742438624026e-06,
+      "loss": 0.0272,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4031936127744511,
+      "grad_norm": 1.082839151667033,
+      "learning_rate": 6.773356622643703e-06,
+      "loss": 0.0173,
+      "step": 1414
+    },
+    {
+      "epoch": 0.40347875677216993,
+      "grad_norm": 1.2816980567733227,
+      "learning_rate": 6.769037488273087e-06,
+      "loss": 0.0382,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4037639007698888,
+      "grad_norm": 1.6873641138117723,
+      "learning_rate": 6.764716844435944e-06,
+      "loss": 0.0343,
+      "step": 1416
+    },
+    {
+      "epoch": 0.40404904476760767,
+      "grad_norm": 1.5688848234236163,
+      "learning_rate": 6.760394694818949e-06,
+      "loss": 0.0323,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4043341887653265,
+      "grad_norm": 0.613312664131555,
+      "learning_rate": 6.756071043110064e-06,
+      "loss": 0.0163,
+      "step": 1418
+    },
+    {
+      "epoch": 0.40461933276304535,
+      "grad_norm": 1.0497026541574568,
+      "learning_rate": 6.751745892998527e-06,
+      "loss": 0.0308,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4049044767607642,
+      "grad_norm": 1.6397322996436774,
+      "learning_rate": 6.747419248174864e-06,
+      "loss": 0.0263,
+      "step": 1420
+    },
+    {
+      "epoch": 0.405189620758483,
+      "grad_norm": 0.718022101787763,
+      "learning_rate": 6.743091112330866e-06,
+      "loss": 0.0115,
+      "step": 1421
+    },
+    {
+      "epoch": 0.40547476475620187,
+      "grad_norm": 0.8207958542231343,
+      "learning_rate": 6.738761489159604e-06,
+      "loss": 0.0319,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4057599087539207,
+      "grad_norm": 1.574000171294008,
+      "learning_rate": 6.734430382355417e-06,
+      "loss": 0.0293,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4060450527516396,
+      "grad_norm": 1.019984530943042,
+      "learning_rate": 6.730097795613903e-06,
+      "loss": 0.0124,
+      "step": 1424
+    },
+    {
+      "epoch": 0.40633019674935844,
+      "grad_norm": 2.417307028940788,
+      "learning_rate": 6.725763732631933e-06,
+      "loss": 0.0384,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4066153407470773,
+      "grad_norm": 0.7247645218622035,
+      "learning_rate": 6.721428197107631e-06,
+      "loss": 0.0126,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4069004847447961,
+      "grad_norm": 1.48458332799653,
+      "learning_rate": 6.717091192740378e-06,
+      "loss": 0.0422,
+      "step": 1427
+    },
+    {
+      "epoch": 0.40718562874251496,
+      "grad_norm": 2.9745577201399938,
+      "learning_rate": 6.712752723230812e-06,
+      "loss": 0.0758,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4074707727402338,
+      "grad_norm": 1.2408092239041282,
+      "learning_rate": 6.708412792280816e-06,
+      "loss": 0.0236,
+      "step": 1429
+    },
+    {
+      "epoch": 0.40775591673795264,
+      "grad_norm": 0.505712627957945,
+      "learning_rate": 6.704071403593524e-06,
+      "loss": 0.0075,
+      "step": 1430
+    },
+    {
+      "epoch": 0.40804106073567153,
+      "grad_norm": 0.8007241478639011,
+      "learning_rate": 6.699728560873313e-06,
+      "loss": 0.012,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4083262047333904,
+      "grad_norm": 1.22383156743963,
+      "learning_rate": 6.695384267825799e-06,
+      "loss": 0.022,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4086113487311092,
+      "grad_norm": 1.643834967167695,
+      "learning_rate": 6.691038528157836e-06,
+      "loss": 0.0297,
+      "step": 1433
+    },
+    {
+      "epoch": 0.40889649272882805,
+      "grad_norm": 0.981130245185286,
+      "learning_rate": 6.686691345577517e-06,
+      "loss": 0.0311,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4091816367265469,
+      "grad_norm": 0.7535527214250041,
+      "learning_rate": 6.682342723794157e-06,
+      "loss": 0.0079,
+      "step": 1435
+    },
+    {
+      "epoch": 0.40946678072426573,
+      "grad_norm": 1.4979534384423328,
+      "learning_rate": 6.67799266651831e-06,
+      "loss": 0.0286,
+      "step": 1436
+    },
+    {
+      "epoch": 0.40975192472198463,
+      "grad_norm": 2.7428696820621137,
+      "learning_rate": 6.673641177461743e-06,
+      "loss": 0.0526,
+      "step": 1437
+    },
+    {
+      "epoch": 0.41003706871970347,
+      "grad_norm": 3.078770079875373,
+      "learning_rate": 6.669288260337455e-06,
+      "loss": 0.0906,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4103222127174223,
+      "grad_norm": 1.6189746261305773,
+      "learning_rate": 6.66493391885966e-06,
+      "loss": 0.0318,
+      "step": 1439
+    },
+    {
+      "epoch": 0.41060735671514115,
+      "grad_norm": 1.2290982178336693,
+      "learning_rate": 6.660578156743782e-06,
+      "loss": 0.0205,
+      "step": 1440
+    },
+    {
+      "epoch": 0.41089250071286,
+      "grad_norm": 0.9641872975294026,
+      "learning_rate": 6.656220977706465e-06,
+      "loss": 0.0189,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4111776447105788,
+      "grad_norm": 1.1312515292408134,
+      "learning_rate": 6.6518623854655615e-06,
+      "loss": 0.0288,
+      "step": 1442
+    },
+    {
+      "epoch": 0.41146278870829767,
+      "grad_norm": 0.8164142521434059,
+      "learning_rate": 6.647502383740123e-06,
+      "loss": 0.019,
+      "step": 1443
+    },
+    {
+      "epoch": 0.41174793270601656,
+      "grad_norm": 2.1749643294622114,
+      "learning_rate": 6.64314097625041e-06,
+      "loss": 0.0326,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4120330767037354,
+      "grad_norm": 0.6688149744527794,
+      "learning_rate": 6.638778166717879e-06,
+      "loss": 0.0117,
+      "step": 1445
+    },
+    {
+      "epoch": 0.41231822070145424,
+      "grad_norm": 0.9144749242956802,
+      "learning_rate": 6.634413958865187e-06,
+      "loss": 0.0292,
+      "step": 1446
+    },
+    {
+      "epoch": 0.4126033646991731,
+      "grad_norm": 2.0650177025758447,
+      "learning_rate": 6.6300483564161775e-06,
+      "loss": 0.0408,
+      "step": 1447
+    },
+    {
+      "epoch": 0.4128885086968919,
+      "grad_norm": 1.5032365198506417,
+      "learning_rate": 6.625681363095892e-06,
+      "loss": 0.0305,
+      "step": 1448
+    },
+    {
+      "epoch": 0.41317365269461076,
+      "grad_norm": 1.2503228614033306,
+      "learning_rate": 6.621312982630551e-06,
+      "loss": 0.0214,
+      "step": 1449
+    },
+    {
+      "epoch": 0.4134587966923296,
+      "grad_norm": 1.5579491899758215,
+      "learning_rate": 6.616943218747566e-06,
+      "loss": 0.0208,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4137439406900485,
+      "grad_norm": 0.5046986151483627,
+      "learning_rate": 6.612572075175521e-06,
+      "loss": 0.0077,
+      "step": 1451
+    },
+    {
+      "epoch": 0.41402908468776733,
+      "grad_norm": 1.3996170926428428,
+      "learning_rate": 6.608199555644186e-06,
+      "loss": 0.0335,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4143142286854862,
+      "grad_norm": 1.8024425394477386,
+      "learning_rate": 6.603825663884497e-06,
+      "loss": 0.0391,
+      "step": 1453
+    },
+    {
+      "epoch": 0.414599372683205,
+      "grad_norm": 1.4905431440579864,
+      "learning_rate": 6.599450403628564e-06,
+      "loss": 0.0417,
+      "step": 1454
+    },
+    {
+      "epoch": 0.41488451668092385,
+      "grad_norm": 1.874227242654859,
+      "learning_rate": 6.59507377860967e-06,
+      "loss": 0.06,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4151696606786427,
+      "grad_norm": 0.6546908200537064,
+      "learning_rate": 6.590695792562251e-06,
+      "loss": 0.0129,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4154548046763616,
+      "grad_norm": 1.2440886365948998,
+      "learning_rate": 6.5863164492219165e-06,
+      "loss": 0.0125,
+      "step": 1457
+    },
+    {
+      "epoch": 0.41573994867408043,
+      "grad_norm": 0.9649143457410724,
+      "learning_rate": 6.5819357523254255e-06,
+      "loss": 0.0215,
+      "step": 1458
+    },
+    {
+      "epoch": 0.41602509267179927,
+      "grad_norm": 0.9042777652710828,
+      "learning_rate": 6.577553705610695e-06,
+      "loss": 0.0156,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4163102366695181,
+      "grad_norm": 1.5328371548614659,
+      "learning_rate": 6.573170312816797e-06,
+      "loss": 0.0389,
+      "step": 1460
+    },
+    {
+      "epoch": 0.41659538066723695,
+      "grad_norm": 1.7516837901407996,
+      "learning_rate": 6.568785577683945e-06,
+      "loss": 0.0282,
+      "step": 1461
+    },
+    {
+      "epoch": 0.4168805246649558,
+      "grad_norm": 0.7279340731763488,
+      "learning_rate": 6.564399503953502e-06,
+      "loss": 0.0082,
+      "step": 1462
+    },
+    {
+      "epoch": 0.4171656686626746,
+      "grad_norm": 1.0423109263190555,
+      "learning_rate": 6.560012095367976e-06,
+      "loss": 0.0256,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4174508126603935,
+      "grad_norm": 0.8846001166727955,
+      "learning_rate": 6.555623355671008e-06,
+      "loss": 0.0184,
+      "step": 1464
+    },
+    {
+      "epoch": 0.41773595665811236,
+      "grad_norm": 0.8616631964720626,
+      "learning_rate": 6.551233288607378e-06,
+      "loss": 0.0269,
+      "step": 1465
+    },
+    {
+      "epoch": 0.4180211006558312,
+      "grad_norm": 0.6610304293410116,
+      "learning_rate": 6.5468418979229995e-06,
+      "loss": 0.011,
+      "step": 1466
+    },
+    {
+      "epoch": 0.41830624465355004,
+      "grad_norm": 2.0231953278181263,
+      "learning_rate": 6.542449187364913e-06,
+      "loss": 0.0481,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4185913886512689,
+      "grad_norm": 0.7272833402001642,
+      "learning_rate": 6.538055160681288e-06,
+      "loss": 0.0129,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4188765326489877,
+      "grad_norm": 1.4865399155591759,
+      "learning_rate": 6.533659821621414e-06,
+      "loss": 0.0188,
+      "step": 1469
+    },
+    {
+      "epoch": 0.41916167664670656,
+      "grad_norm": 0.9858783676809323,
+      "learning_rate": 6.529263173935703e-06,
+      "loss": 0.0218,
+      "step": 1470
+    },
+    {
+      "epoch": 0.41944682064442546,
+      "grad_norm": 0.9797679454187829,
+      "learning_rate": 6.524865221375681e-06,
+      "loss": 0.0236,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4197319646421443,
+      "grad_norm": 0.6172684716843654,
+      "learning_rate": 6.52046596769399e-06,
+      "loss": 0.009,
+      "step": 1472
+    },
+    {
+      "epoch": 0.42001710863986313,
+      "grad_norm": 1.7614776588873011,
+      "learning_rate": 6.51606541664438e-06,
+      "loss": 0.0498,
+      "step": 1473
+    },
+    {
+      "epoch": 0.420302252637582,
+      "grad_norm": 0.5970946705806073,
+      "learning_rate": 6.511663571981708e-06,
+      "loss": 0.0085,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4205873966353008,
+      "grad_norm": 0.8599355447218311,
+      "learning_rate": 6.507260437461939e-06,
+      "loss": 0.0214,
+      "step": 1475
+    },
+    {
+      "epoch": 0.42087254063301965,
+      "grad_norm": 1.0987421015708239,
+      "learning_rate": 6.502856016842132e-06,
+      "loss": 0.0185,
+      "step": 1476
+    },
+    {
+      "epoch": 0.42115768463073855,
+      "grad_norm": 0.44275052835685647,
+      "learning_rate": 6.498450313880449e-06,
+      "loss": 0.0062,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4214428286284574,
+      "grad_norm": 1.813940453890275,
+      "learning_rate": 6.4940433323361425e-06,
+      "loss": 0.0287,
+      "step": 1478
+    },
+    {
+      "epoch": 0.42172797262617623,
+      "grad_norm": 1.140392945104892,
+      "learning_rate": 6.489635075969558e-06,
+      "loss": 0.0305,
+      "step": 1479
+    },
+    {
+      "epoch": 0.42201311662389507,
+      "grad_norm": 1.8157436853386422,
+      "learning_rate": 6.485225548542129e-06,
+      "loss": 0.0561,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4222982606216139,
+      "grad_norm": 2.1226592153844464,
+      "learning_rate": 6.4808147538163715e-06,
+      "loss": 0.0298,
+      "step": 1481
+    },
+    {
+      "epoch": 0.42258340461933275,
+      "grad_norm": 0.20001428446574743,
+      "learning_rate": 6.476402695555884e-06,
+      "loss": 0.005,
+      "step": 1482
+    },
+    {
+      "epoch": 0.4228685486170516,
+      "grad_norm": 2.5663295718601873,
+      "learning_rate": 6.471989377525344e-06,
+      "loss": 0.0587,
+      "step": 1483
+    },
+    {
+      "epoch": 0.4231536926147705,
+      "grad_norm": 1.0554595923300076,
+      "learning_rate": 6.467574803490504e-06,
+      "loss": 0.0356,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4234388366124893,
+      "grad_norm": 1.0690511762758468,
+      "learning_rate": 6.463158977218185e-06,
+      "loss": 0.0199,
+      "step": 1485
+    },
+    {
+      "epoch": 0.42372398061020816,
+      "grad_norm": 1.1410017307638216,
+      "learning_rate": 6.458741902476281e-06,
+      "loss": 0.0301,
+      "step": 1486
+    },
+    {
+      "epoch": 0.424009124607927,
+      "grad_norm": 2.2153038532557616,
+      "learning_rate": 6.454323583033748e-06,
+      "loss": 0.036,
+      "step": 1487
+    },
+    {
+      "epoch": 0.42429426860564584,
+      "grad_norm": 1.1095267953641579,
+      "learning_rate": 6.449904022660604e-06,
+      "loss": 0.0212,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4245794126033647,
+      "grad_norm": 0.8268000886152754,
+      "learning_rate": 6.445483225127931e-06,
+      "loss": 0.029,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4248645566010835,
+      "grad_norm": 0.5393177997173784,
+      "learning_rate": 6.441061194207858e-06,
+      "loss": 0.0102,
+      "step": 1490
+    },
+    {
+      "epoch": 0.4251497005988024,
+      "grad_norm": 1.8543252777041668,
+      "learning_rate": 6.436637933673575e-06,
+      "loss": 0.0242,
+      "step": 1491
+    },
+    {
+      "epoch": 0.42543484459652126,
+      "grad_norm": 1.63129603581456,
+      "learning_rate": 6.4322134472993145e-06,
+      "loss": 0.0428,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4257199885942401,
+      "grad_norm": 1.6335340263445486,
+      "learning_rate": 6.42778773886036e-06,
+      "loss": 0.0353,
+      "step": 1493
+    },
+    {
+      "epoch": 0.42600513259195893,
+      "grad_norm": 1.5007279654260655,
+      "learning_rate": 6.423360812133034e-06,
+      "loss": 0.0449,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4262902765896778,
+      "grad_norm": 0.34770534993641233,
+      "learning_rate": 6.4189326708946995e-06,
+      "loss": 0.0123,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4265754205873966,
+      "grad_norm": 0.5863596956438394,
+      "learning_rate": 6.414503318923757e-06,
+      "loss": 0.0157,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4268605645851155,
+      "grad_norm": 0.476093826497124,
+      "learning_rate": 6.410072759999643e-06,
+      "loss": 0.0125,
+      "step": 1497
+    },
+    {
+      "epoch": 0.42714570858283435,
+      "grad_norm": 1.7901802911684266,
+      "learning_rate": 6.405640997902813e-06,
+      "loss": 0.0329,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4274308525805532,
+      "grad_norm": 0.5017338212552557,
+      "learning_rate": 6.401208036414762e-06,
+      "loss": 0.0056,
+      "step": 1499
+    },
+    {
+      "epoch": 0.42771599657827203,
+      "grad_norm": 2.0718273104929588,
+      "learning_rate": 6.396773879318001e-06,
+      "loss": 0.0372,
+      "step": 1500
+    },
+    {
+      "epoch": 0.42800114057599087,
+      "grad_norm": 1.55254709417258,
+      "learning_rate": 6.392338530396065e-06,
+      "loss": 0.0098,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4282862845737097,
+      "grad_norm": 0.8803879294604957,
+      "learning_rate": 6.387901993433501e-06,
+      "loss": 0.0137,
+      "step": 1502
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.9428880424761917,
+      "learning_rate": 6.383464272215874e-06,
+      "loss": 0.0186,
+      "step": 1503
+    },
+    {
+      "epoch": 0.42885657256914744,
+      "grad_norm": 1.2850260719004754,
+      "learning_rate": 6.379025370529755e-06,
+      "loss": 0.0149,
+      "step": 1504
+    },
+    {
+      "epoch": 0.4291417165668663,
+      "grad_norm": 0.4867048785280101,
+      "learning_rate": 6.374585292162732e-06,
+      "loss": 0.0168,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4294268605645851,
+      "grad_norm": 0.6247035768828255,
+      "learning_rate": 6.370144040903385e-06,
+      "loss": 0.0145,
+      "step": 1506
+    },
+    {
+      "epoch": 0.42971200456230396,
+      "grad_norm": 0.3953757300162384,
+      "learning_rate": 6.365701620541304e-06,
+      "loss": 0.0056,
+      "step": 1507
+    },
+    {
+      "epoch": 0.4299971485600228,
+      "grad_norm": 0.6458629274018028,
+      "learning_rate": 6.361258034867071e-06,
+      "loss": 0.0204,
+      "step": 1508
+    },
+    {
+      "epoch": 0.43028229255774164,
+      "grad_norm": 1.7431227583997453,
+      "learning_rate": 6.356813287672262e-06,
+      "loss": 0.0472,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4305674365554605,
+      "grad_norm": 1.3899155935216172,
+      "learning_rate": 6.352367382749448e-06,
+      "loss": 0.0242,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4308525805531794,
+      "grad_norm": 1.2294226877678534,
+      "learning_rate": 6.347920323892189e-06,
+      "loss": 0.0272,
+      "step": 1511
+    },
+    {
+      "epoch": 0.4311377245508982,
+      "grad_norm": 1.1560872748885018,
+      "learning_rate": 6.343472114895022e-06,
+      "loss": 0.0314,
+      "step": 1512
+    },
+    {
+      "epoch": 0.43142286854861706,
+      "grad_norm": 0.2978850832967228,
+      "learning_rate": 6.339022759553474e-06,
+      "loss": 0.0098,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4317080125463359,
+      "grad_norm": 1.1481971058772948,
+      "learning_rate": 6.334572261664041e-06,
+      "loss": 0.0289,
+      "step": 1514
+    },
+    {
+      "epoch": 0.43199315654405473,
+      "grad_norm": 2.2882732185225914,
+      "learning_rate": 6.330120625024204e-06,
+      "loss": 0.0508,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4322783005417736,
+      "grad_norm": 0.80064981962937,
+      "learning_rate": 6.32566785343241e-06,
+      "loss": 0.0124,
+      "step": 1516
+    },
+    {
+      "epoch": 0.43256344453949247,
+      "grad_norm": 1.5961426275986392,
+      "learning_rate": 6.321213950688073e-06,
+      "loss": 0.0304,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4328485885372113,
+      "grad_norm": 1.2598701107379529,
+      "learning_rate": 6.316758920591576e-06,
+      "loss": 0.0249,
+      "step": 1518
+    },
+    {
+      "epoch": 0.43313373253493015,
+      "grad_norm": 1.8114852964203785,
+      "learning_rate": 6.312302766944263e-06,
+      "loss": 0.0433,
+      "step": 1519
+    },
+    {
+      "epoch": 0.433418876532649,
+      "grad_norm": 2.1634382274709756,
+      "learning_rate": 6.307845493548433e-06,
+      "loss": 0.0384,
+      "step": 1520
+    },
+    {
+      "epoch": 0.43370402053036783,
+      "grad_norm": 1.6679111444495003,
+      "learning_rate": 6.303387104207347e-06,
+      "loss": 0.0361,
+      "step": 1521
+    },
+    {
+      "epoch": 0.43398916452808667,
+      "grad_norm": 1.0012504241438478,
+      "learning_rate": 6.2989276027252134e-06,
+      "loss": 0.0218,
+      "step": 1522
+    },
+    {
+      "epoch": 0.4342743085258055,
+      "grad_norm": 1.2371276572473315,
+      "learning_rate": 6.2944669929071885e-06,
+      "loss": 0.0268,
+      "step": 1523
+    },
+    {
+      "epoch": 0.4345594525235244,
+      "grad_norm": 0.5710678689888669,
+      "learning_rate": 6.29000527855938e-06,
+      "loss": 0.0115,
+      "step": 1524
+    },
+    {
+      "epoch": 0.43484459652124324,
+      "grad_norm": 1.0062812882252696,
+      "learning_rate": 6.285542463488834e-06,
+      "loss": 0.0141,
+      "step": 1525
+    },
+    {
+      "epoch": 0.4351297405189621,
+      "grad_norm": 1.2851888569327077,
+      "learning_rate": 6.2810785515035345e-06,
+      "loss": 0.027,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4354148845166809,
+      "grad_norm": 1.4671361717954792,
+      "learning_rate": 6.276613546412406e-06,
+      "loss": 0.0416,
+      "step": 1527
+    },
+    {
+      "epoch": 0.43570002851439976,
+      "grad_norm": 1.224206846786883,
+      "learning_rate": 6.2721474520253e-06,
+      "loss": 0.0195,
+      "step": 1528
+    },
+    {
+      "epoch": 0.4359851725121186,
+      "grad_norm": 1.9344425177803088,
+      "learning_rate": 6.2676802721530035e-06,
+      "loss": 0.0291,
+      "step": 1529
+    },
+    {
+      "epoch": 0.43627031650983744,
+      "grad_norm": 0.34256835211154624,
+      "learning_rate": 6.263212010607226e-06,
+      "loss": 0.0074,
+      "step": 1530
+    },
+    {
+      "epoch": 0.43655546050755634,
+      "grad_norm": 1.4291677082822674,
+      "learning_rate": 6.2587426712006005e-06,
+      "loss": 0.0313,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4368406045052752,
+      "grad_norm": 1.1108722988310464,
+      "learning_rate": 6.254272257746678e-06,
+      "loss": 0.0235,
+      "step": 1532
+    },
+    {
+      "epoch": 0.437125748502994,
+      "grad_norm": 0.876472008383419,
+      "learning_rate": 6.249800774059934e-06,
+      "loss": 0.0121,
+      "step": 1533
+    },
+    {
+      "epoch": 0.43741089250071286,
+      "grad_norm": 1.5645424818809175,
+      "learning_rate": 6.245328223955744e-06,
+      "loss": 0.0377,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4376960364984317,
+      "grad_norm": 2.1968349193676375,
+      "learning_rate": 6.240854611250406e-06,
+      "loss": 0.0447,
+      "step": 1535
+    },
+    {
+      "epoch": 0.43798118049615054,
+      "grad_norm": 1.6968143052217195,
+      "learning_rate": 6.236379939761117e-06,
+      "loss": 0.0398,
+      "step": 1536
+    },
+    {
+      "epoch": 0.43826632449386943,
+      "grad_norm": 1.7686078679002104,
+      "learning_rate": 6.231904213305979e-06,
+      "loss": 0.0277,
+      "step": 1537
+    },
+    {
+      "epoch": 0.43855146849158827,
+      "grad_norm": 2.1556753958526764,
+      "learning_rate": 6.227427435703997e-06,
+      "loss": 0.0453,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4388366124893071,
+      "grad_norm": 1.5287737335788831,
+      "learning_rate": 6.2229496107750685e-06,
+      "loss": 0.0547,
+      "step": 1539
+    },
+    {
+      "epoch": 0.43912175648702595,
+      "grad_norm": 1.9309123246484545,
+      "learning_rate": 6.218470742339987e-06,
+      "loss": 0.0245,
+      "step": 1540
+    },
+    {
+      "epoch": 0.4394069004847448,
+      "grad_norm": 1.1195972239606753,
+      "learning_rate": 6.21399083422044e-06,
+      "loss": 0.0373,
+      "step": 1541
+    },
+    {
+      "epoch": 0.43969204448246363,
+      "grad_norm": 0.8200635359851965,
+      "learning_rate": 6.2095098902389926e-06,
+      "loss": 0.0108,
+      "step": 1542
+    },
+    {
+      "epoch": 0.43997718848018247,
+      "grad_norm": 1.633335602846522,
+      "learning_rate": 6.205027914219105e-06,
+      "loss": 0.0248,
+      "step": 1543
+    },
+    {
+      "epoch": 0.44026233247790136,
+      "grad_norm": 0.9579796630214975,
+      "learning_rate": 6.20054490998511e-06,
+      "loss": 0.0135,
+      "step": 1544
+    },
+    {
+      "epoch": 0.4405474764756202,
+      "grad_norm": 0.9207521721967611,
+      "learning_rate": 6.1960608813622215e-06,
+      "loss": 0.0276,
+      "step": 1545
+    },
+    {
+      "epoch": 0.44083262047333904,
+      "grad_norm": 1.5832680552524554,
+      "learning_rate": 6.191575832176524e-06,
+      "loss": 0.0256,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4411177644710579,
+      "grad_norm": 1.3157435085679845,
+      "learning_rate": 6.187089766254979e-06,
+      "loss": 0.0224,
+      "step": 1547
+    },
+    {
+      "epoch": 0.4414029084687767,
+      "grad_norm": 1.026787695779842,
+      "learning_rate": 6.182602687425407e-06,
+      "loss": 0.0372,
+      "step": 1548
+    },
+    {
+      "epoch": 0.44168805246649556,
+      "grad_norm": 0.948727292811553,
+      "learning_rate": 6.178114599516504e-06,
+      "loss": 0.0287,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4419731964642144,
+      "grad_norm": 0.6305066109325781,
+      "learning_rate": 6.173625506357814e-06,
+      "loss": 0.0223,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4422583404619333,
+      "grad_norm": 0.7164994011017372,
+      "learning_rate": 6.169135411779749e-06,
+      "loss": 0.0119,
+      "step": 1551
+    },
+    {
+      "epoch": 0.44254348445965214,
+      "grad_norm": 0.8216759106611643,
+      "learning_rate": 6.164644319613571e-06,
+      "loss": 0.0271,
+      "step": 1552
+    },
+    {
+      "epoch": 0.442828628457371,
+      "grad_norm": 0.8327268854217976,
+      "learning_rate": 6.160152233691393e-06,
+      "loss": 0.022,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4431137724550898,
+      "grad_norm": 1.1286241666524117,
+      "learning_rate": 6.155659157846178e-06,
+      "loss": 0.0303,
+      "step": 1554
+    },
+    {
+      "epoch": 0.44339891645280866,
+      "grad_norm": 1.5466332840928882,
+      "learning_rate": 6.151165095911733e-06,
+      "loss": 0.0355,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4436840604505275,
+      "grad_norm": 1.164291430707095,
+      "learning_rate": 6.1466700517227044e-06,
+      "loss": 0.0292,
+      "step": 1556
+    },
+    {
+      "epoch": 0.4439692044482464,
+      "grad_norm": 0.7237425608379466,
+      "learning_rate": 6.142174029114579e-06,
+      "loss": 0.0204,
+      "step": 1557
+    },
+    {
+      "epoch": 0.44425434844596523,
+      "grad_norm": 1.525607471024849,
+      "learning_rate": 6.137677031923679e-06,
+      "loss": 0.0319,
+      "step": 1558
+    },
+    {
+      "epoch": 0.44453949244368407,
+      "grad_norm": 1.5076148523093191,
+      "learning_rate": 6.133179063987156e-06,
+      "loss": 0.0336,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4448246364414029,
+      "grad_norm": 2.470324848338266,
+      "learning_rate": 6.128680129142991e-06,
+      "loss": 0.0173,
+      "step": 1560
+    },
+    {
+      "epoch": 0.44510978043912175,
+      "grad_norm": 3.040160807782214,
+      "learning_rate": 6.1241802312299895e-06,
+      "loss": 0.0709,
+      "step": 1561
+    },
+    {
+      "epoch": 0.4453949244368406,
+      "grad_norm": 1.0375626585244984,
+      "learning_rate": 6.119679374087778e-06,
+      "loss": 0.027,
+      "step": 1562
+    },
+    {
+      "epoch": 0.44568006843455943,
+      "grad_norm": 1.3814735291614966,
+      "learning_rate": 6.115177561556806e-06,
+      "loss": 0.0252,
+      "step": 1563
+    },
+    {
+      "epoch": 0.4459652124322783,
+      "grad_norm": 1.2628722116063182,
+      "learning_rate": 6.110674797478332e-06,
+      "loss": 0.0215,
+      "step": 1564
+    },
+    {
+      "epoch": 0.44625035642999716,
+      "grad_norm": 1.0677934713873056,
+      "learning_rate": 6.10617108569443e-06,
+      "loss": 0.0293,
+      "step": 1565
+    },
+    {
+      "epoch": 0.446535500427716,
+      "grad_norm": 1.3607904858833422,
+      "learning_rate": 6.1016664300479835e-06,
+      "loss": 0.0382,
+      "step": 1566
+    },
+    {
+      "epoch": 0.44682064442543484,
+      "grad_norm": 2.4973158350192355,
+      "learning_rate": 6.097160834382678e-06,
+      "loss": 0.0577,
+      "step": 1567
+    },
+    {
+      "epoch": 0.4471057884231537,
+      "grad_norm": 1.4543142327687228,
+      "learning_rate": 6.092654302543002e-06,
+      "loss": 0.0569,
+      "step": 1568
+    },
+    {
+      "epoch": 0.4473909324208725,
+      "grad_norm": 1.547257628130523,
+      "learning_rate": 6.088146838374247e-06,
+      "loss": 0.0202,
+      "step": 1569
+    },
+    {
+      "epoch": 0.44767607641859136,
+      "grad_norm": 1.1158703535313215,
+      "learning_rate": 6.083638445722493e-06,
+      "loss": 0.0192,
+      "step": 1570
+    },
+    {
+      "epoch": 0.44796122041631026,
+      "grad_norm": 0.9709682394236485,
+      "learning_rate": 6.079129128434619e-06,
+      "loss": 0.031,
+      "step": 1571
+    },
+    {
+      "epoch": 0.4482463644140291,
+      "grad_norm": 1.6318763828200638,
+      "learning_rate": 6.074618890358287e-06,
+      "loss": 0.0184,
+      "step": 1572
+    },
+    {
+      "epoch": 0.44853150841174794,
+      "grad_norm": 1.8459498966727417,
+      "learning_rate": 6.07010773534195e-06,
+      "loss": 0.039,
+      "step": 1573
+    },
+    {
+      "epoch": 0.4488166524094668,
+      "grad_norm": 1.2404446802074922,
+      "learning_rate": 6.06559566723484e-06,
+      "loss": 0.0182,
+      "step": 1574
+    },
+    {
+      "epoch": 0.4491017964071856,
+      "grad_norm": 1.4442916601321651,
+      "learning_rate": 6.06108268988697e-06,
+      "loss": 0.0401,
+      "step": 1575
+    },
+    {
+      "epoch": 0.44938694040490446,
+      "grad_norm": 0.9320753790810795,
+      "learning_rate": 6.056568807149127e-06,
+      "loss": 0.0239,
+      "step": 1576
+    },
+    {
+      "epoch": 0.44967208440262335,
+      "grad_norm": 1.4017667599463124,
+      "learning_rate": 6.052054022872873e-06,
+      "loss": 0.023,
+      "step": 1577
+    },
+    {
+      "epoch": 0.4499572284003422,
+      "grad_norm": 0.6833624466019657,
+      "learning_rate": 6.047538340910534e-06,
+      "loss": 0.0144,
+      "step": 1578
+    },
+    {
+      "epoch": 0.45024237239806103,
+      "grad_norm": 0.6630610636763746,
+      "learning_rate": 6.0430217651152105e-06,
+      "loss": 0.02,
+      "step": 1579
+    },
+    {
+      "epoch": 0.45052751639577987,
+      "grad_norm": 0.6627483075350021,
+      "learning_rate": 6.0385042993407574e-06,
+      "loss": 0.0152,
+      "step": 1580
+    },
+    {
+      "epoch": 0.4508126603934987,
+      "grad_norm": 0.7801825576978031,
+      "learning_rate": 6.033985947441795e-06,
+      "loss": 0.0186,
+      "step": 1581
+    },
+    {
+      "epoch": 0.45109780439121755,
+      "grad_norm": 0.7431161777834322,
+      "learning_rate": 6.029466713273695e-06,
+      "loss": 0.0115,
+      "step": 1582
+    },
+    {
+      "epoch": 0.4513829483889364,
+      "grad_norm": 0.4049070036061823,
+      "learning_rate": 6.024946600692585e-06,
+      "loss": 0.0074,
+      "step": 1583
+    },
+    {
+      "epoch": 0.4516680923866553,
+      "grad_norm": 1.0478163608067583,
+      "learning_rate": 6.020425613555341e-06,
+      "loss": 0.0199,
+      "step": 1584
+    },
+    {
+      "epoch": 0.4519532363843741,
+      "grad_norm": 1.5885560292025787,
+      "learning_rate": 6.015903755719588e-06,
+      "loss": 0.0454,
+      "step": 1585
+    },
+    {
+      "epoch": 0.45223838038209296,
+      "grad_norm": 1.0854421779536836,
+      "learning_rate": 6.011381031043686e-06,
+      "loss": 0.0294,
+      "step": 1586
+    },
+    {
+      "epoch": 0.4525235243798118,
+      "grad_norm": 0.4134891652172671,
+      "learning_rate": 6.006857443386746e-06,
+      "loss": 0.012,
+      "step": 1587
+    },
+    {
+      "epoch": 0.45280866837753064,
+      "grad_norm": 1.7921075489007599,
+      "learning_rate": 6.002332996608605e-06,
+      "loss": 0.0277,
+      "step": 1588
+    },
+    {
+      "epoch": 0.4530938123752495,
+      "grad_norm": 0.9673035410671176,
+      "learning_rate": 5.9978076945698395e-06,
+      "loss": 0.017,
+      "step": 1589
+    },
+    {
+      "epoch": 0.4533789563729683,
+      "grad_norm": 0.9724566217783264,
+      "learning_rate": 5.993281541131753e-06,
+      "loss": 0.0319,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4536641003706872,
+      "grad_norm": 0.7175181508761327,
+      "learning_rate": 5.9887545401563775e-06,
+      "loss": 0.017,
+      "step": 1591
+    },
+    {
+      "epoch": 0.45394924436840606,
+      "grad_norm": 1.1816527100772103,
+      "learning_rate": 5.984226695506464e-06,
+      "loss": 0.0279,
+      "step": 1592
+    },
+    {
+      "epoch": 0.4542343883661249,
+      "grad_norm": 1.9390267938168535,
+      "learning_rate": 5.979698011045492e-06,
+      "loss": 0.038,
+      "step": 1593
+    },
+    {
+      "epoch": 0.45451953236384374,
+      "grad_norm": 0.9155009626135624,
+      "learning_rate": 5.975168490637644e-06,
+      "loss": 0.052,
+      "step": 1594
+    },
+    {
+      "epoch": 0.4548046763615626,
+      "grad_norm": 0.7977284818854188,
+      "learning_rate": 5.970638138147829e-06,
+      "loss": 0.0107,
+      "step": 1595
+    },
+    {
+      "epoch": 0.4550898203592814,
+      "grad_norm": 0.9196711480479474,
+      "learning_rate": 5.966106957441661e-06,
+      "loss": 0.0176,
+      "step": 1596
+    },
+    {
+      "epoch": 0.4553749643570003,
+      "grad_norm": 4.36651610023969,
+      "learning_rate": 5.961574952385457e-06,
+      "loss": 0.0302,
+      "step": 1597
+    },
+    {
+      "epoch": 0.45566010835471915,
+      "grad_norm": 2.9956145689115043,
+      "learning_rate": 5.957042126846243e-06,
+      "loss": 0.0496,
+      "step": 1598
+    },
+    {
+      "epoch": 0.455945252352438,
+      "grad_norm": 1.318241867584188,
+      "learning_rate": 5.952508484691742e-06,
+      "loss": 0.0248,
+      "step": 1599
+    },
+    {
+      "epoch": 0.45623039635015683,
+      "grad_norm": 1.4632248370825405,
+      "learning_rate": 5.947974029790375e-06,
+      "loss": 0.0288,
+      "step": 1600
+    },
+    {
+      "epoch": 0.45651554034787567,
+      "grad_norm": 1.5614361952777818,
+      "learning_rate": 5.943438766011256e-06,
+      "loss": 0.0505,
+      "step": 1601
+    },
+    {
+      "epoch": 0.4568006843455945,
+      "grad_norm": 0.5404823438930301,
+      "learning_rate": 5.938902697224189e-06,
+      "loss": 0.0065,
+      "step": 1602
+    },
+    {
+      "epoch": 0.45708582834331335,
+      "grad_norm": 1.0138237952237483,
+      "learning_rate": 5.934365827299666e-06,
+      "loss": 0.0248,
+      "step": 1603
+    },
+    {
+      "epoch": 0.45737097234103224,
+      "grad_norm": 1.3912521886461697,
+      "learning_rate": 5.9298281601088616e-06,
+      "loss": 0.0226,
+      "step": 1604
+    },
+    {
+      "epoch": 0.4576561163387511,
+      "grad_norm": 1.346109081340996,
+      "learning_rate": 5.92528969952363e-06,
+      "loss": 0.0431,
+      "step": 1605
+    },
+    {
+      "epoch": 0.4579412603364699,
+      "grad_norm": 0.6722377723333736,
+      "learning_rate": 5.9207504494165025e-06,
+      "loss": 0.0057,
+      "step": 1606
+    },
+    {
+      "epoch": 0.45822640433418876,
+      "grad_norm": 1.5756505347726604,
+      "learning_rate": 5.916210413660687e-06,
+      "loss": 0.0323,
+      "step": 1607
+    },
+    {
+      "epoch": 0.4585115483319076,
+      "grad_norm": 1.372238468851667,
+      "learning_rate": 5.9116695961300584e-06,
+      "loss": 0.0216,
+      "step": 1608
+    },
+    {
+      "epoch": 0.45879669232962644,
+      "grad_norm": 0.7560326833267913,
+      "learning_rate": 5.907128000699159e-06,
+      "loss": 0.0141,
+      "step": 1609
+    },
+    {
+      "epoch": 0.4590818363273453,
+      "grad_norm": 1.8272989751745645,
+      "learning_rate": 5.9025856312431985e-06,
+      "loss": 0.031,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4593669803250642,
+      "grad_norm": 1.515596989644,
+      "learning_rate": 5.898042491638042e-06,
+      "loss": 0.0302,
+      "step": 1611
+    },
+    {
+      "epoch": 0.459652124322783,
+      "grad_norm": 1.3089273048621382,
+      "learning_rate": 5.8934985857602144e-06,
+      "loss": 0.0314,
+      "step": 1612
+    },
+    {
+      "epoch": 0.45993726832050186,
+      "grad_norm": 1.0921868001336128,
+      "learning_rate": 5.8889539174868935e-06,
+      "loss": 0.0199,
+      "step": 1613
+    },
+    {
+      "epoch": 0.4602224123182207,
+      "grad_norm": 0.9944417014298922,
+      "learning_rate": 5.88440849069591e-06,
+      "loss": 0.0139,
+      "step": 1614
+    },
+    {
+      "epoch": 0.46050755631593954,
+      "grad_norm": 0.9965377028835146,
+      "learning_rate": 5.879862309265738e-06,
+      "loss": 0.0235,
+      "step": 1615
+    },
+    {
+      "epoch": 0.4607927003136584,
+      "grad_norm": 1.4517072630287036,
+      "learning_rate": 5.875315377075497e-06,
+      "loss": 0.0417,
+      "step": 1616
+    },
+    {
+      "epoch": 0.46107784431137727,
+      "grad_norm": 0.4827625546781854,
+      "learning_rate": 5.87076769800495e-06,
+      "loss": 0.0184,
+      "step": 1617
+    },
+    {
+      "epoch": 0.4613629883090961,
+      "grad_norm": 1.9595291045089964,
+      "learning_rate": 5.866219275934494e-06,
+      "loss": 0.0388,
+      "step": 1618
+    },
+    {
+      "epoch": 0.46164813230681495,
+      "grad_norm": 0.9604209914453493,
+      "learning_rate": 5.86167011474516e-06,
+      "loss": 0.0438,
+      "step": 1619
+    },
+    {
+      "epoch": 0.4619332763045338,
+      "grad_norm": 1.5066448543730424,
+      "learning_rate": 5.857120218318612e-06,
+      "loss": 0.0418,
+      "step": 1620
+    },
+    {
+      "epoch": 0.46221842030225263,
+      "grad_norm": 1.5026872619543066,
+      "learning_rate": 5.852569590537138e-06,
+      "loss": 0.0284,
+      "step": 1621
+    },
+    {
+      "epoch": 0.46250356429997147,
+      "grad_norm": 1.0384555882506465,
+      "learning_rate": 5.848018235283654e-06,
+      "loss": 0.0217,
+      "step": 1622
+    },
+    {
+      "epoch": 0.4627887082976903,
+      "grad_norm": 1.2663294731314405,
+      "learning_rate": 5.843466156441693e-06,
+      "loss": 0.0215,
+      "step": 1623
+    },
+    {
+      "epoch": 0.4630738522954092,
+      "grad_norm": 1.4544791021792642,
+      "learning_rate": 5.838913357895408e-06,
+      "loss": 0.043,
+      "step": 1624
+    },
+    {
+      "epoch": 0.46335899629312804,
+      "grad_norm": 1.9146030561749325,
+      "learning_rate": 5.834359843529565e-06,
+      "loss": 0.0429,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4636441402908469,
+      "grad_norm": 0.5614062355847471,
+      "learning_rate": 5.82980561722954e-06,
+      "loss": 0.0084,
+      "step": 1626
+    },
+    {
+      "epoch": 0.4639292842885657,
+      "grad_norm": 1.0668880452470082,
+      "learning_rate": 5.8252506828813195e-06,
+      "loss": 0.0259,
+      "step": 1627
+    },
+    {
+      "epoch": 0.46421442828628456,
+      "grad_norm": 0.8362296876846605,
+      "learning_rate": 5.82069504437149e-06,
+      "loss": 0.0219,
+      "step": 1628
+    },
+    {
+      "epoch": 0.4644995722840034,
+      "grad_norm": 1.865874275154151,
+      "learning_rate": 5.816138705587242e-06,
+      "loss": 0.0339,
+      "step": 1629
+    },
+    {
+      "epoch": 0.46478471628172224,
+      "grad_norm": 0.7831355484928149,
+      "learning_rate": 5.811581670416363e-06,
+      "loss": 0.011,
+      "step": 1630
+    },
+    {
+      "epoch": 0.46506986027944114,
+      "grad_norm": 0.9660875848485134,
+      "learning_rate": 5.807023942747232e-06,
+      "loss": 0.0145,
+      "step": 1631
+    },
+    {
+      "epoch": 0.46535500427716,
+      "grad_norm": 0.6626505988518036,
+      "learning_rate": 5.802465526468825e-06,
+      "loss": 0.0171,
+      "step": 1632
+    },
+    {
+      "epoch": 0.4656401482748788,
+      "grad_norm": 1.8732921332420605,
+      "learning_rate": 5.797906425470696e-06,
+      "loss": 0.0412,
+      "step": 1633
+    },
+    {
+      "epoch": 0.46592529227259766,
+      "grad_norm": 0.9110735893383872,
+      "learning_rate": 5.7933466436429945e-06,
+      "loss": 0.0196,
+      "step": 1634
+    },
+    {
+      "epoch": 0.4662104362703165,
+      "grad_norm": 0.9187090445005635,
+      "learning_rate": 5.788786184876443e-06,
+      "loss": 0.0086,
+      "step": 1635
+    },
+    {
+      "epoch": 0.46649558026803534,
+      "grad_norm": 0.4577963271928218,
+      "learning_rate": 5.784225053062342e-06,
+      "loss": 0.0172,
+      "step": 1636
+    },
+    {
+      "epoch": 0.46678072426575423,
+      "grad_norm": 0.6514140579973696,
+      "learning_rate": 5.77966325209257e-06,
+      "loss": 0.0185,
+      "step": 1637
+    },
+    {
+      "epoch": 0.46706586826347307,
+      "grad_norm": 1.5558388744790788,
+      "learning_rate": 5.775100785859576e-06,
+      "loss": 0.0379,
+      "step": 1638
+    },
+    {
+      "epoch": 0.4673510122611919,
+      "grad_norm": 1.3399688016409907,
+      "learning_rate": 5.770537658256373e-06,
+      "loss": 0.0327,
+      "step": 1639
+    },
+    {
+      "epoch": 0.46763615625891075,
+      "grad_norm": 1.1869080276664048,
+      "learning_rate": 5.765973873176544e-06,
+      "loss": 0.0194,
+      "step": 1640
+    },
+    {
+      "epoch": 0.4679213002566296,
+      "grad_norm": 1.2920295998953273,
+      "learning_rate": 5.761409434514223e-06,
+      "loss": 0.0158,
+      "step": 1641
+    },
+    {
+      "epoch": 0.46820644425434843,
+      "grad_norm": 0.5899684738930414,
+      "learning_rate": 5.756844346164115e-06,
+      "loss": 0.0119,
+      "step": 1642
+    },
+    {
+      "epoch": 0.46849158825206727,
+      "grad_norm": 0.7664070156217964,
+      "learning_rate": 5.752278612021471e-06,
+      "loss": 0.0178,
+      "step": 1643
+    },
+    {
+      "epoch": 0.46877673224978617,
+      "grad_norm": 1.0535022309389133,
+      "learning_rate": 5.747712235982094e-06,
+      "loss": 0.0217,
+      "step": 1644
+    },
+    {
+      "epoch": 0.469061876247505,
+      "grad_norm": 0.5765641658694458,
+      "learning_rate": 5.743145221942333e-06,
+      "loss": 0.0088,
+      "step": 1645
+    },
+    {
+      "epoch": 0.46934702024522384,
+      "grad_norm": 1.220409106870033,
+      "learning_rate": 5.73857757379909e-06,
+      "loss": 0.0312,
+      "step": 1646
+    },
+    {
+      "epoch": 0.4696321642429427,
+      "grad_norm": 1.1369091506463587,
+      "learning_rate": 5.7340092954497965e-06,
+      "loss": 0.0241,
+      "step": 1647
+    },
+    {
+      "epoch": 0.4699173082406615,
+      "grad_norm": 1.0456981783807946,
+      "learning_rate": 5.729440390792429e-06,
+      "loss": 0.0241,
+      "step": 1648
+    },
+    {
+      "epoch": 0.47020245223838036,
+      "grad_norm": 2.3070988825991146,
+      "learning_rate": 5.724870863725497e-06,
+      "loss": 0.0403,
+      "step": 1649
+    },
+    {
+      "epoch": 0.4704875962360992,
+      "grad_norm": 1.3376794149494875,
+      "learning_rate": 5.72030071814804e-06,
+      "loss": 0.0313,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4707727402338181,
+      "grad_norm": 0.9046821280854266,
+      "learning_rate": 5.715729957959625e-06,
+      "loss": 0.0101,
+      "step": 1651
+    },
+    {
+      "epoch": 0.47105788423153694,
+      "grad_norm": 1.0675955765788645,
+      "learning_rate": 5.7111585870603455e-06,
+      "loss": 0.0191,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4713430282292558,
+      "grad_norm": 0.5649793938230241,
+      "learning_rate": 5.7065866093508145e-06,
+      "loss": 0.0112,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4716281722269746,
+      "grad_norm": 1.3714092185991984,
+      "learning_rate": 5.7020140287321645e-06,
+      "loss": 0.0291,
+      "step": 1654
+    },
+    {
+      "epoch": 0.47191331622469346,
+      "grad_norm": 0.5776238056475193,
+      "learning_rate": 5.697440849106039e-06,
+      "loss": 0.0129,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4721984602224123,
+      "grad_norm": 0.9710900939411276,
+      "learning_rate": 5.692867074374596e-06,
+      "loss": 0.0429,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4724836042201312,
+      "grad_norm": 0.6235588357191956,
+      "learning_rate": 5.688292708440501e-06,
+      "loss": 0.0092,
+      "step": 1657
+    },
+    {
+      "epoch": 0.47276874821785003,
+      "grad_norm": 1.3140976352998177,
+      "learning_rate": 5.6837177552069204e-06,
+      "loss": 0.0238,
+      "step": 1658
+    },
+    {
+      "epoch": 0.47305389221556887,
+      "grad_norm": 1.4703932405577713,
+      "learning_rate": 5.679142218577525e-06,
+      "loss": 0.0233,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4733390362132877,
+      "grad_norm": 3.059192393508655,
+      "learning_rate": 5.6745661024564854e-06,
+      "loss": 0.0941,
+      "step": 1660
+    },
+    {
+      "epoch": 0.47362418021100655,
+      "grad_norm": 1.2480858491288185,
+      "learning_rate": 5.66998941074846e-06,
+      "loss": 0.0254,
+      "step": 1661
+    },
+    {
+      "epoch": 0.4739093242087254,
+      "grad_norm": 1.3391566425560755,
+      "learning_rate": 5.665412147358607e-06,
+      "loss": 0.0192,
+      "step": 1662
+    },
+    {
+      "epoch": 0.47419446820644423,
+      "grad_norm": 1.9530592193510963,
+      "learning_rate": 5.6608343161925616e-06,
+      "loss": 0.0348,
+      "step": 1663
+    },
+    {
+      "epoch": 0.4744796122041631,
+      "grad_norm": 0.5882117331069833,
+      "learning_rate": 5.656255921156455e-06,
+      "loss": 0.0078,
+      "step": 1664
+    },
+    {
+      "epoch": 0.47476475620188197,
+      "grad_norm": 0.4806565752566173,
+      "learning_rate": 5.651676966156892e-06,
+      "loss": 0.0166,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4750499001996008,
+      "grad_norm": 2.5545952290218805,
+      "learning_rate": 5.6470974551009565e-06,
+      "loss": 0.0401,
+      "step": 1666
+    },
+    {
+      "epoch": 0.47533504419731964,
+      "grad_norm": 3.454081035186067,
+      "learning_rate": 5.642517391896208e-06,
+      "loss": 0.0461,
+      "step": 1667
+    },
+    {
+      "epoch": 0.4756201881950385,
+      "grad_norm": 1.2109065735867222,
+      "learning_rate": 5.637936780450679e-06,
+      "loss": 0.0287,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4759053321927573,
+      "grad_norm": 1.1002493916885174,
+      "learning_rate": 5.633355624672865e-06,
+      "loss": 0.0123,
+      "step": 1669
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.7138212776082841,
+      "learning_rate": 5.628773928471729e-06,
+      "loss": 0.008,
+      "step": 1670
+    },
+    {
+      "epoch": 0.47647562018819506,
+      "grad_norm": 1.273915884803688,
+      "learning_rate": 5.6241916957566955e-06,
+      "loss": 0.014,
+      "step": 1671
+    },
+    {
+      "epoch": 0.4767607641859139,
+      "grad_norm": 1.0112889834715977,
+      "learning_rate": 5.619608930437647e-06,
+      "loss": 0.0296,
+      "step": 1672
+    },
+    {
+      "epoch": 0.47704590818363274,
+      "grad_norm": 2.625237325555971,
+      "learning_rate": 5.6150256364249165e-06,
+      "loss": 0.0315,
+      "step": 1673
+    },
+    {
+      "epoch": 0.4773310521813516,
+      "grad_norm": 0.9222000915216457,
+      "learning_rate": 5.610441817629293e-06,
+      "loss": 0.0214,
+      "step": 1674
+    },
+    {
+      "epoch": 0.4776161961790704,
+      "grad_norm": 1.2738057180224518,
+      "learning_rate": 5.605857477962009e-06,
+      "loss": 0.06,
+      "step": 1675
+    },
+    {
+      "epoch": 0.47790134017678926,
+      "grad_norm": 1.7750846497598578,
+      "learning_rate": 5.601272621334749e-06,
+      "loss": 0.0754,
+      "step": 1676
+    },
+    {
+      "epoch": 0.47818648417450815,
+      "grad_norm": 1.7959815648598039,
+      "learning_rate": 5.5966872516596275e-06,
+      "loss": 0.0617,
+      "step": 1677
+    },
+    {
+      "epoch": 0.478471628172227,
+      "grad_norm": 1.1371874135314433,
+      "learning_rate": 5.592101372849207e-06,
+      "loss": 0.0196,
+      "step": 1678
+    },
+    {
+      "epoch": 0.47875677216994583,
+      "grad_norm": 0.9732582128190234,
+      "learning_rate": 5.587514988816477e-06,
+      "loss": 0.0151,
+      "step": 1679
+    },
+    {
+      "epoch": 0.47904191616766467,
+      "grad_norm": 1.3674954779188182,
+      "learning_rate": 5.582928103474863e-06,
+      "loss": 0.0298,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4793270601653835,
+      "grad_norm": 0.9010991025266389,
+      "learning_rate": 5.5783407207382146e-06,
+      "loss": 0.0208,
+      "step": 1681
+    },
+    {
+      "epoch": 0.47961220416310235,
+      "grad_norm": 0.7798121219399466,
+      "learning_rate": 5.573752844520812e-06,
+      "loss": 0.0186,
+      "step": 1682
+    },
+    {
+      "epoch": 0.4798973481608212,
+      "grad_norm": 0.8532491665099611,
+      "learning_rate": 5.569164478737346e-06,
+      "loss": 0.0285,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4801824921585401,
+      "grad_norm": 1.2652300108229633,
+      "learning_rate": 5.564575627302936e-06,
+      "loss": 0.0146,
+      "step": 1684
+    },
+    {
+      "epoch": 0.4804676361562589,
+      "grad_norm": 1.0842547917810184,
+      "learning_rate": 5.5599862941331084e-06,
+      "loss": 0.0236,
+      "step": 1685
+    },
+    {
+      "epoch": 0.48075278015397777,
+      "grad_norm": 0.7692465444203684,
+      "learning_rate": 5.5553964831438044e-06,
+      "loss": 0.0098,
+      "step": 1686
+    },
+    {
+      "epoch": 0.4810379241516966,
+      "grad_norm": 0.6516988532185854,
+      "learning_rate": 5.5508061982513715e-06,
+      "loss": 0.0095,
+      "step": 1687
+    },
+    {
+      "epoch": 0.48132306814941545,
+      "grad_norm": 0.86894709268537,
+      "learning_rate": 5.546215443372562e-06,
+      "loss": 0.0207,
+      "step": 1688
+    },
+    {
+      "epoch": 0.4816082121471343,
+      "grad_norm": 1.1589813428803937,
+      "learning_rate": 5.541624222424529e-06,
+      "loss": 0.0182,
+      "step": 1689
+    },
+    {
+      "epoch": 0.4818933561448531,
+      "grad_norm": 0.4753465975864277,
+      "learning_rate": 5.537032539324824e-06,
+      "loss": 0.0143,
+      "step": 1690
+    },
+    {
+      "epoch": 0.482178500142572,
+      "grad_norm": 0.3429031884155126,
+      "learning_rate": 5.53244039799139e-06,
+      "loss": 0.0089,
+      "step": 1691
+    },
+    {
+      "epoch": 0.48246364414029086,
+      "grad_norm": 1.9674672537859532,
+      "learning_rate": 5.527847802342567e-06,
+      "loss": 0.054,
+      "step": 1692
+    },
+    {
+      "epoch": 0.4827487881380097,
+      "grad_norm": 0.9183324853750284,
+      "learning_rate": 5.523254756297074e-06,
+      "loss": 0.0332,
+      "step": 1693
+    },
+    {
+      "epoch": 0.48303393213572854,
+      "grad_norm": 1.5405624693680082,
+      "learning_rate": 5.518661263774023e-06,
+      "loss": 0.0625,
+      "step": 1694
+    },
+    {
+      "epoch": 0.4833190761334474,
+      "grad_norm": 1.4478394054980839,
+      "learning_rate": 5.514067328692902e-06,
+      "loss": 0.0234,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4836042201311662,
+      "grad_norm": 1.198470695313575,
+      "learning_rate": 5.509472954973578e-06,
+      "loss": 0.0339,
+      "step": 1696
+    },
+    {
+      "epoch": 0.4838893641288851,
+      "grad_norm": 1.4872008148949296,
+      "learning_rate": 5.504878146536291e-06,
+      "loss": 0.0152,
+      "step": 1697
+    },
+    {
+      "epoch": 0.48417450812660395,
+      "grad_norm": 0.9089989321086841,
+      "learning_rate": 5.500282907301655e-06,
+      "loss": 0.0227,
+      "step": 1698
+    },
+    {
+      "epoch": 0.4844596521243228,
+      "grad_norm": 0.9108428863142347,
+      "learning_rate": 5.495687241190646e-06,
+      "loss": 0.0081,
+      "step": 1699
+    },
+    {
+      "epoch": 0.48474479612204163,
+      "grad_norm": 1.7700228667742393,
+      "learning_rate": 5.491091152124611e-06,
+      "loss": 0.0397,
+      "step": 1700
+    },
+    {
+      "epoch": 0.48502994011976047,
+      "grad_norm": 1.086368323032613,
+      "learning_rate": 5.4864946440252555e-06,
+      "loss": 0.0295,
+      "step": 1701
+    },
+    {
+      "epoch": 0.4853150841174793,
+      "grad_norm": 0.8881192538969089,
+      "learning_rate": 5.4818977208146375e-06,
+      "loss": 0.0315,
+      "step": 1702
+    },
+    {
+      "epoch": 0.48560022811519815,
+      "grad_norm": 0.8299716634196358,
+      "learning_rate": 5.477300386415176e-06,
+      "loss": 0.0125,
+      "step": 1703
+    },
+    {
+      "epoch": 0.48588537211291705,
+      "grad_norm": 0.8778183424306557,
+      "learning_rate": 5.47270264474964e-06,
+      "loss": 0.0225,
+      "step": 1704
+    },
+    {
+      "epoch": 0.4861705161106359,
+      "grad_norm": 0.3759935217970214,
+      "learning_rate": 5.46810449974114e-06,
+      "loss": 0.0067,
+      "step": 1705
+    },
+    {
+      "epoch": 0.4864556601083547,
+      "grad_norm": 1.012504942599993,
+      "learning_rate": 5.46350595531314e-06,
+      "loss": 0.0342,
+      "step": 1706
+    },
+    {
+      "epoch": 0.48674080410607357,
+      "grad_norm": 0.3155753567046455,
+      "learning_rate": 5.458907015389435e-06,
+      "loss": 0.0068,
+      "step": 1707
+    },
+    {
+      "epoch": 0.4870259481037924,
+      "grad_norm": 0.8772082266297694,
+      "learning_rate": 5.454307683894166e-06,
+      "loss": 0.0137,
+      "step": 1708
+    },
+    {
+      "epoch": 0.48731109210151125,
+      "grad_norm": 0.8540371209025869,
+      "learning_rate": 5.449707964751801e-06,
+      "loss": 0.011,
+      "step": 1709
+    },
+    {
+      "epoch": 0.4875962360992301,
+      "grad_norm": 0.9996853289237507,
+      "learning_rate": 5.445107861887144e-06,
+      "loss": 0.0156,
+      "step": 1710
+    },
+    {
+      "epoch": 0.487881380096949,
+      "grad_norm": 1.4433308389528134,
+      "learning_rate": 5.440507379225325e-06,
+      "loss": 0.0347,
+      "step": 1711
+    },
+    {
+      "epoch": 0.4881665240946678,
+      "grad_norm": 0.6514104791176355,
+      "learning_rate": 5.435906520691798e-06,
+      "loss": 0.0279,
+      "step": 1712
+    },
+    {
+      "epoch": 0.48845166809238666,
+      "grad_norm": 1.1622561862286143,
+      "learning_rate": 5.431305290212335e-06,
+      "loss": 0.0531,
+      "step": 1713
+    },
+    {
+      "epoch": 0.4887368120901055,
+      "grad_norm": 1.0205243505665194,
+      "learning_rate": 5.426703691713029e-06,
+      "loss": 0.034,
+      "step": 1714
+    },
+    {
+      "epoch": 0.48902195608782434,
+      "grad_norm": 1.5178878114736611,
+      "learning_rate": 5.422101729120287e-06,
+      "loss": 0.0326,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4893071000855432,
+      "grad_norm": 1.5210315919144914,
+      "learning_rate": 5.417499406360823e-06,
+      "loss": 0.0244,
+      "step": 1716
+    },
+    {
+      "epoch": 0.4895922440832621,
+      "grad_norm": 0.7365109722429185,
+      "learning_rate": 5.412896727361663e-06,
+      "loss": 0.0109,
+      "step": 1717
+    },
+    {
+      "epoch": 0.4898773880809809,
+      "grad_norm": 1.5587515861114862,
+      "learning_rate": 5.408293696050132e-06,
+      "loss": 0.0252,
+      "step": 1718
+    },
+    {
+      "epoch": 0.49016253207869975,
+      "grad_norm": 0.8960330096591171,
+      "learning_rate": 5.40369031635386e-06,
+      "loss": 0.0364,
+      "step": 1719
+    },
+    {
+      "epoch": 0.4904476760764186,
+      "grad_norm": 0.6398440868605254,
+      "learning_rate": 5.399086592200774e-06,
+      "loss": 0.0134,
+      "step": 1720
+    },
+    {
+      "epoch": 0.49073282007413743,
+      "grad_norm": 1.2962219418993994,
+      "learning_rate": 5.394482527519089e-06,
+      "loss": 0.0417,
+      "step": 1721
+    },
+    {
+      "epoch": 0.49101796407185627,
+      "grad_norm": 1.1342913204571548,
+      "learning_rate": 5.38987812623732e-06,
+      "loss": 0.0349,
+      "step": 1722
+    },
+    {
+      "epoch": 0.4913031080695751,
+      "grad_norm": 1.415562479211778,
+      "learning_rate": 5.385273392284262e-06,
+      "loss": 0.0341,
+      "step": 1723
+    },
+    {
+      "epoch": 0.491588252067294,
+      "grad_norm": 0.570721980159015,
+      "learning_rate": 5.380668329588996e-06,
+      "loss": 0.0092,
+      "step": 1724
+    },
+    {
+      "epoch": 0.49187339606501285,
+      "grad_norm": 1.2011080438952686,
+      "learning_rate": 5.376062942080883e-06,
+      "loss": 0.0199,
+      "step": 1725
+    },
+    {
+      "epoch": 0.4921585400627317,
+      "grad_norm": 0.8641679648671838,
+      "learning_rate": 5.371457233689563e-06,
+      "loss": 0.0344,
+      "step": 1726
+    },
+    {
+      "epoch": 0.4924436840604505,
+      "grad_norm": 1.3972786914451918,
+      "learning_rate": 5.366851208344948e-06,
+      "loss": 0.0267,
+      "step": 1727
+    },
+    {
+      "epoch": 0.49272882805816937,
+      "grad_norm": 0.9723899245443296,
+      "learning_rate": 5.362244869977224e-06,
+      "loss": 0.0241,
+      "step": 1728
+    },
+    {
+      "epoch": 0.4930139720558882,
+      "grad_norm": 1.2779418929535273,
+      "learning_rate": 5.357638222516833e-06,
+      "loss": 0.0228,
+      "step": 1729
+    },
+    {
+      "epoch": 0.49329911605360705,
+      "grad_norm": 1.596318162921975,
+      "learning_rate": 5.353031269894497e-06,
+      "loss": 0.0324,
+      "step": 1730
+    },
+    {
+      "epoch": 0.49358426005132594,
+      "grad_norm": 1.066749627601742,
+      "learning_rate": 5.348424016041188e-06,
+      "loss": 0.02,
+      "step": 1731
+    },
+    {
+      "epoch": 0.4938694040490448,
+      "grad_norm": 2.035347543339062,
+      "learning_rate": 5.343816464888136e-06,
+      "loss": 0.0404,
+      "step": 1732
+    },
+    {
+      "epoch": 0.4941545480467636,
+      "grad_norm": 3.3356286353565414,
+      "learning_rate": 5.339208620366827e-06,
+      "loss": 0.0574,
+      "step": 1733
+    },
+    {
+      "epoch": 0.49443969204448246,
+      "grad_norm": 1.5158864916198709,
+      "learning_rate": 5.334600486408995e-06,
+      "loss": 0.0407,
+      "step": 1734
+    },
+    {
+      "epoch": 0.4947248360422013,
+      "grad_norm": 1.0504079623189557,
+      "learning_rate": 5.3299920669466224e-06,
+      "loss": 0.0199,
+      "step": 1735
+    },
+    {
+      "epoch": 0.49500998003992014,
+      "grad_norm": 1.2772606321263538,
+      "learning_rate": 5.325383365911937e-06,
+      "loss": 0.0203,
+      "step": 1736
+    },
+    {
+      "epoch": 0.49529512403763903,
+      "grad_norm": 0.7819050740242356,
+      "learning_rate": 5.320774387237403e-06,
+      "loss": 0.0141,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4955802680353579,
+      "grad_norm": 1.0278612287001418,
+      "learning_rate": 5.316165134855724e-06,
+      "loss": 0.0157,
+      "step": 1738
+    },
+    {
+      "epoch": 0.4958654120330767,
+      "grad_norm": 1.4731255811320176,
+      "learning_rate": 5.311555612699837e-06,
+      "loss": 0.0289,
+      "step": 1739
+    },
+    {
+      "epoch": 0.49615055603079555,
+      "grad_norm": 0.7990479306692042,
+      "learning_rate": 5.306945824702908e-06,
+      "loss": 0.0091,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4964357000285144,
+      "grad_norm": 1.3645150027377915,
+      "learning_rate": 5.302335774798331e-06,
+      "loss": 0.0417,
+      "step": 1741
+    },
+    {
+      "epoch": 0.49672084402623323,
+      "grad_norm": 0.5097086067391874,
+      "learning_rate": 5.297725466919722e-06,
+      "loss": 0.0067,
+      "step": 1742
+    },
+    {
+      "epoch": 0.49700598802395207,
+      "grad_norm": 1.220765771700666,
+      "learning_rate": 5.29311490500092e-06,
+      "loss": 0.0422,
+      "step": 1743
+    },
+    {
+      "epoch": 0.49729113202167097,
+      "grad_norm": 0.6971180492227302,
+      "learning_rate": 5.288504092975975e-06,
+      "loss": 0.0206,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4975762760193898,
+      "grad_norm": 1.0855891533617872,
+      "learning_rate": 5.2838930347791605e-06,
+      "loss": 0.0377,
+      "step": 1745
+    },
+    {
+      "epoch": 0.49786142001710865,
+      "grad_norm": 0.5153649468444501,
+      "learning_rate": 5.279281734344951e-06,
+      "loss": 0.0115,
+      "step": 1746
+    },
+    {
+      "epoch": 0.4981465640148275,
+      "grad_norm": 0.8506337273843959,
+      "learning_rate": 5.274670195608029e-06,
+      "loss": 0.0127,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4984317080125463,
+      "grad_norm": 1.1073160831408677,
+      "learning_rate": 5.270058422503284e-06,
+      "loss": 0.0191,
+      "step": 1748
+    },
+    {
+      "epoch": 0.49871685201026517,
+      "grad_norm": 1.7959372472589628,
+      "learning_rate": 5.265446418965801e-06,
+      "loss": 0.0349,
+      "step": 1749
+    },
+    {
+      "epoch": 0.499001996007984,
+      "grad_norm": 1.2999018982531467,
+      "learning_rate": 5.260834188930867e-06,
+      "loss": 0.0144,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4992871400057029,
+      "grad_norm": 1.1152004670396407,
+      "learning_rate": 5.256221736333959e-06,
+      "loss": 0.0389,
+      "step": 1751
+    },
+    {
+      "epoch": 0.49957228400342174,
+      "grad_norm": 1.194164124119914,
+      "learning_rate": 5.25160906511074e-06,
+      "loss": 0.0264,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4998574280011406,
+      "grad_norm": 1.2213677453155716,
+      "learning_rate": 5.246996179197068e-06,
+      "loss": 0.0245,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5001425719988595,
+      "grad_norm": 0.5547879059243418,
+      "learning_rate": 5.24238308252898e-06,
+      "loss": 0.017,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5004277159965783,
+      "grad_norm": 0.7060694139102462,
+      "learning_rate": 5.23776977904269e-06,
+      "loss": 0.0196,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5007128599942972,
+      "grad_norm": 0.8256575425352183,
+      "learning_rate": 5.233156272674595e-06,
+      "loss": 0.0109,
+      "step": 1756
+    },
+    {
+      "epoch": 0.500998003992016,
+      "grad_norm": 1.2395878805797904,
+      "learning_rate": 5.228542567361258e-06,
+      "loss": 0.0264,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5012831479897348,
+      "grad_norm": 0.5321859041552769,
+      "learning_rate": 5.223928667039414e-06,
+      "loss": 0.0097,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5015682919874537,
+      "grad_norm": 0.6929216929421017,
+      "learning_rate": 5.219314575645972e-06,
+      "loss": 0.03,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5018534359851725,
+      "grad_norm": 1.3885854817205727,
+      "learning_rate": 5.214700297117989e-06,
+      "loss": 0.0214,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5021385799828914,
+      "grad_norm": 1.0482514572374837,
+      "learning_rate": 5.210085835392696e-06,
+      "loss": 0.0229,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5024237239806102,
+      "grad_norm": 1.180111096741413,
+      "learning_rate": 5.205471194407474e-06,
+      "loss": 0.0476,
+      "step": 1762
+    },
+    {
+      "epoch": 0.502708867978329,
+      "grad_norm": 1.8163067516658773,
+      "learning_rate": 5.2008563780998554e-06,
+      "loss": 0.0289,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5029940119760479,
+      "grad_norm": 0.7027677059471549,
+      "learning_rate": 5.196241390407526e-06,
+      "loss": 0.023,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5032791559737667,
+      "grad_norm": 1.8922631201234337,
+      "learning_rate": 5.1916262352683154e-06,
+      "loss": 0.0283,
+      "step": 1765
+    },
+    {
+      "epoch": 0.5035642999714856,
+      "grad_norm": 1.5571212020941443,
+      "learning_rate": 5.187010916620196e-06,
+      "loss": 0.0361,
+      "step": 1766
+    },
+    {
+      "epoch": 0.5038494439692045,
+      "grad_norm": 0.8174284451105421,
+      "learning_rate": 5.182395438401282e-06,
+      "loss": 0.0224,
+      "step": 1767
+    },
+    {
+      "epoch": 0.5041345879669233,
+      "grad_norm": 1.3080120864075888,
+      "learning_rate": 5.177779804549821e-06,
+      "loss": 0.0142,
+      "step": 1768
+    },
+    {
+      "epoch": 0.5044197319646422,
+      "grad_norm": 1.8386928220755963,
+      "learning_rate": 5.173164019004195e-06,
+      "loss": 0.0305,
+      "step": 1769
+    },
+    {
+      "epoch": 0.504704875962361,
+      "grad_norm": 1.1936198989153632,
+      "learning_rate": 5.168548085702916e-06,
+      "loss": 0.0387,
+      "step": 1770
+    },
+    {
+      "epoch": 0.5049900199600799,
+      "grad_norm": 1.6312386336058349,
+      "learning_rate": 5.16393200858462e-06,
+      "loss": 0.0306,
+      "step": 1771
+    },
+    {
+      "epoch": 0.5052751639577987,
+      "grad_norm": 0.9145988600574612,
+      "learning_rate": 5.159315791588068e-06,
+      "loss": 0.011,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5055603079555175,
+      "grad_norm": 0.902930547135958,
+      "learning_rate": 5.154699438652139e-06,
+      "loss": 0.0148,
+      "step": 1773
+    },
+    {
+      "epoch": 0.5058454519532364,
+      "grad_norm": 0.8416064963586531,
+      "learning_rate": 5.150082953715826e-06,
+      "loss": 0.0142,
+      "step": 1774
+    },
+    {
+      "epoch": 0.5061305959509552,
+      "grad_norm": 0.8900714074477286,
+      "learning_rate": 5.145466340718241e-06,
+      "loss": 0.0219,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5064157399486741,
+      "grad_norm": 1.66826695692096,
+      "learning_rate": 5.140849603598598e-06,
+      "loss": 0.042,
+      "step": 1776
+    },
+    {
+      "epoch": 0.5067008839463929,
+      "grad_norm": 2.656687757346725,
+      "learning_rate": 5.13623274629622e-06,
+      "loss": 0.0651,
+      "step": 1777
+    },
+    {
+      "epoch": 0.5069860279441117,
+      "grad_norm": 0.5318371444639051,
+      "learning_rate": 5.131615772750534e-06,
+      "loss": 0.0124,
+      "step": 1778
+    },
+    {
+      "epoch": 0.5072711719418306,
+      "grad_norm": 1.870474093394001,
+      "learning_rate": 5.126998686901066e-06,
+      "loss": 0.0444,
+      "step": 1779
+    },
+    {
+      "epoch": 0.5075563159395494,
+      "grad_norm": 0.27375968758014385,
+      "learning_rate": 5.122381492687431e-06,
+      "loss": 0.0082,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5078414599372684,
+      "grad_norm": 0.5572083483107301,
+      "learning_rate": 5.1177641940493505e-06,
+      "loss": 0.0198,
+      "step": 1781
+    },
+    {
+      "epoch": 0.5081266039349872,
+      "grad_norm": 1.8211201480961516,
+      "learning_rate": 5.113146794926619e-06,
+      "loss": 0.0605,
+      "step": 1782
+    },
+    {
+      "epoch": 0.508411747932706,
+      "grad_norm": 1.629287587730149,
+      "learning_rate": 5.108529299259129e-06,
+      "loss": 0.0485,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5086968919304249,
+      "grad_norm": 1.2077436875656857,
+      "learning_rate": 5.10391171098685e-06,
+      "loss": 0.0262,
+      "step": 1784
+    },
+    {
+      "epoch": 0.5089820359281437,
+      "grad_norm": 1.016814636419437,
+      "learning_rate": 5.099294034049829e-06,
+      "loss": 0.0424,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5092671799258626,
+      "grad_norm": 0.7224054605047718,
+      "learning_rate": 5.0946762723881935e-06,
+      "loss": 0.0101,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5095523239235814,
+      "grad_norm": 1.081931745901931,
+      "learning_rate": 5.09005842994214e-06,
+      "loss": 0.0277,
+      "step": 1787
+    },
+    {
+      "epoch": 0.5098374679213002,
+      "grad_norm": 0.5980162131093523,
+      "learning_rate": 5.0854405106519336e-06,
+      "loss": 0.0144,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5101226119190191,
+      "grad_norm": 0.32617215799420385,
+      "learning_rate": 5.080822518457909e-06,
+      "loss": 0.0091,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5104077559167379,
+      "grad_norm": 1.637195558047833,
+      "learning_rate": 5.0762044573004555e-06,
+      "loss": 0.0346,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5106928999144568,
+      "grad_norm": 1.9376813290083361,
+      "learning_rate": 5.071586331120031e-06,
+      "loss": 0.0268,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5109780439121756,
+      "grad_norm": 0.8777109804793124,
+      "learning_rate": 5.066968143857141e-06,
+      "loss": 0.0099,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5112631879098944,
+      "grad_norm": 1.1639274694760067,
+      "learning_rate": 5.062349899452345e-06,
+      "loss": 0.0325,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5115483319076134,
+      "grad_norm": 0.9469853264768066,
+      "learning_rate": 5.057731601846253e-06,
+      "loss": 0.0388,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5118334759053322,
+      "grad_norm": 1.55965099823004,
+      "learning_rate": 5.053113254979522e-06,
+      "loss": 0.0317,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5121186199030511,
+      "grad_norm": 1.0572192459413616,
+      "learning_rate": 5.048494862792845e-06,
+      "loss": 0.0279,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5124037639007699,
+      "grad_norm": 0.6540128291643972,
+      "learning_rate": 5.043876429226962e-06,
+      "loss": 0.0082,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5126889078984888,
+      "grad_norm": 1.4677331238705937,
+      "learning_rate": 5.039257958222638e-06,
+      "loss": 0.052,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5129740518962076,
+      "grad_norm": 0.6886544253965207,
+      "learning_rate": 5.0346394537206776e-06,
+      "loss": 0.0224,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5132591958939264,
+      "grad_norm": 1.674371036871377,
+      "learning_rate": 5.030020919661913e-06,
+      "loss": 0.0487,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5135443398916453,
+      "grad_norm": 1.374316064323489,
+      "learning_rate": 5.025402359987201e-06,
+      "loss": 0.0288,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5138294838893641,
+      "grad_norm": 0.8573307882459512,
+      "learning_rate": 5.020783778637415e-06,
+      "loss": 0.0275,
+      "step": 1802
+    },
+    {
+      "epoch": 0.514114627887083,
+      "grad_norm": 1.2301193990464756,
+      "learning_rate": 5.016165179553459e-06,
+      "loss": 0.0263,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5143997718848018,
+      "grad_norm": 1.2182295130590761,
+      "learning_rate": 5.011546566676239e-06,
+      "loss": 0.0359,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5146849158825206,
+      "grad_norm": 1.133488244714276,
+      "learning_rate": 5.006927943946681e-06,
+      "loss": 0.0167,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5149700598802395,
+      "grad_norm": 0.9176237886469092,
+      "learning_rate": 5.0023093153057174e-06,
+      "loss": 0.02,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5152552038779584,
+      "grad_norm": 1.0261123312743583,
+      "learning_rate": 4.997690684694285e-06,
+      "loss": 0.0333,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5155403478756773,
+      "grad_norm": 1.5110718674130426,
+      "learning_rate": 4.99307205605332e-06,
+      "loss": 0.0294,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5158254918733961,
+      "grad_norm": 0.6864276343112098,
+      "learning_rate": 4.988453433323763e-06,
+      "loss": 0.0347,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5161106358711149,
+      "grad_norm": 1.1248317448171876,
+      "learning_rate": 4.983834820446542e-06,
+      "loss": 0.0347,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5163957798688338,
+      "grad_norm": 0.9943083426147765,
+      "learning_rate": 4.979216221362585e-06,
+      "loss": 0.0255,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5166809238665526,
+      "grad_norm": 0.9053821929583953,
+      "learning_rate": 4.9745976400128e-06,
+      "loss": 0.0165,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5169660678642715,
+      "grad_norm": 1.3710164561870049,
+      "learning_rate": 4.96997908033809e-06,
+      "loss": 0.0246,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5172512118619903,
+      "grad_norm": 0.879598522359968,
+      "learning_rate": 4.965360546279324e-06,
+      "loss": 0.0239,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5175363558597091,
+      "grad_norm": 0.6133796107685641,
+      "learning_rate": 4.960742041777365e-06,
+      "loss": 0.016,
+      "step": 1815
+    },
+    {
+      "epoch": 0.517821499857428,
+      "grad_norm": 0.7668338592272866,
+      "learning_rate": 4.95612357077304e-06,
+      "loss": 0.0233,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5181066438551468,
+      "grad_norm": 1.8711063892336894,
+      "learning_rate": 4.951505137207155e-06,
+      "loss": 0.0433,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5183917878528657,
+      "grad_norm": 1.4693009707813043,
+      "learning_rate": 4.94688674502048e-06,
+      "loss": 0.0192,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5186769318505845,
+      "grad_norm": 1.0043129018742638,
+      "learning_rate": 4.942268398153748e-06,
+      "loss": 0.032,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5189620758483033,
+      "grad_norm": 0.9700760923720372,
+      "learning_rate": 4.937650100547656e-06,
+      "loss": 0.03,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5192472198460223,
+      "grad_norm": 1.6874603824297643,
+      "learning_rate": 4.933031856142862e-06,
+      "loss": 0.0434,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5195323638437411,
+      "grad_norm": 1.6134747702626453,
+      "learning_rate": 4.92841366887997e-06,
+      "loss": 0.0249,
+      "step": 1822
+    },
+    {
+      "epoch": 0.51981750784146,
+      "grad_norm": 0.7329551510874709,
+      "learning_rate": 4.9237955426995445e-06,
+      "loss": 0.0282,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5201026518391788,
+      "grad_norm": 1.502060656484669,
+      "learning_rate": 4.919177481542094e-06,
+      "loss": 0.0248,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5203877958368976,
+      "grad_norm": 1.0356299700200353,
+      "learning_rate": 4.914559489348068e-06,
+      "loss": 0.0133,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5206729398346165,
+      "grad_norm": 1.3304571261943245,
+      "learning_rate": 4.909941570057861e-06,
+      "loss": 0.0259,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5209580838323353,
+      "grad_norm": 0.8800083351648127,
+      "learning_rate": 4.905323727611807e-06,
+      "loss": 0.0253,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5212432278300542,
+      "grad_norm": 0.3567933218332538,
+      "learning_rate": 4.900705965950172e-06,
+      "loss": 0.012,
+      "step": 1828
+    },
+    {
+      "epoch": 0.521528371827773,
+      "grad_norm": 0.8401451215832865,
+      "learning_rate": 4.896088289013153e-06,
+      "loss": 0.0191,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5218135158254918,
+      "grad_norm": 1.1818507943017555,
+      "learning_rate": 4.891470700740872e-06,
+      "loss": 0.0253,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5220986598232107,
+      "grad_norm": 0.9347321602247102,
+      "learning_rate": 4.886853205073382e-06,
+      "loss": 0.0204,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5223838038209295,
+      "grad_norm": 1.2285692447409555,
+      "learning_rate": 4.882235805950651e-06,
+      "loss": 0.0204,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5226689478186484,
+      "grad_norm": 0.43955725006957547,
+      "learning_rate": 4.877618507312568e-06,
+      "loss": 0.0174,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5229540918163673,
+      "grad_norm": 0.6565153634012186,
+      "learning_rate": 4.873001313098937e-06,
+      "loss": 0.0135,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5232392358140862,
+      "grad_norm": 1.113652633722711,
+      "learning_rate": 4.868384227249468e-06,
+      "loss": 0.0239,
+      "step": 1835
+    },
+    {
+      "epoch": 0.523524379811805,
+      "grad_norm": 1.2861091962055604,
+      "learning_rate": 4.8637672537037815e-06,
+      "loss": 0.0519,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.8292555437480343,
+      "learning_rate": 4.859150396401404e-06,
+      "loss": 0.0145,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5240946678072427,
+      "grad_norm": 2.2543232111596896,
+      "learning_rate": 4.85453365928176e-06,
+      "loss": 0.0325,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5243798118049615,
+      "grad_norm": 0.7480667232985087,
+      "learning_rate": 4.849917046284175e-06,
+      "loss": 0.0139,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5246649558026804,
+      "grad_norm": 0.4608883072551659,
+      "learning_rate": 4.8453005613478635e-06,
+      "loss": 0.0082,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5249500998003992,
+      "grad_norm": 0.9653852346298967,
+      "learning_rate": 4.8406842084119345e-06,
+      "loss": 0.0378,
+      "step": 1841
+    },
+    {
+      "epoch": 0.525235243798118,
+      "grad_norm": 0.6538953169532171,
+      "learning_rate": 4.8360679914153805e-06,
+      "loss": 0.0236,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5255203877958369,
+      "grad_norm": 2.7130229909581063,
+      "learning_rate": 4.831451914297086e-06,
+      "loss": 0.0351,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5258055317935557,
+      "grad_norm": 0.9988544260022847,
+      "learning_rate": 4.826835980995806e-06,
+      "loss": 0.0313,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5260906757912746,
+      "grad_norm": 0.4820387580248128,
+      "learning_rate": 4.82222019545018e-06,
+      "loss": 0.0094,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5263758197889934,
+      "grad_norm": 0.6020909214467295,
+      "learning_rate": 4.81760456159872e-06,
+      "loss": 0.0207,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5266609637867123,
+      "grad_norm": 0.34297314496197695,
+      "learning_rate": 4.812989083379806e-06,
+      "loss": 0.0091,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5269461077844312,
+      "grad_norm": 0.9821597455371975,
+      "learning_rate": 4.808373764731686e-06,
+      "loss": 0.026,
+      "step": 1848
+    },
+    {
+      "epoch": 0.52723125178215,
+      "grad_norm": 1.427547834808575,
+      "learning_rate": 4.8037586095924756e-06,
+      "loss": 0.04,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5275163957798689,
+      "grad_norm": 0.9056240346340622,
+      "learning_rate": 4.799143621900146e-06,
+      "loss": 0.0355,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5278015397775877,
+      "grad_norm": 1.603722643513683,
+      "learning_rate": 4.794528805592529e-06,
+      "loss": 0.0342,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5280866837753065,
+      "grad_norm": 0.5955654585757876,
+      "learning_rate": 4.789914164607305e-06,
+      "loss": 0.0158,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5283718277730254,
+      "grad_norm": 1.1955850884312524,
+      "learning_rate": 4.785299702882012e-06,
+      "loss": 0.0153,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5286569717707442,
+      "grad_norm": 0.8638149177647332,
+      "learning_rate": 4.7806854243540304e-06,
+      "loss": 0.0283,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5289421157684631,
+      "grad_norm": 0.9683679448303014,
+      "learning_rate": 4.776071332960586e-06,
+      "loss": 0.0224,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5292272597661819,
+      "grad_norm": 1.0544906834955496,
+      "learning_rate": 4.771457432638744e-06,
+      "loss": 0.018,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5295124037639007,
+      "grad_norm": 0.5133813970021956,
+      "learning_rate": 4.766843727325408e-06,
+      "loss": 0.0095,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5297975477616196,
+      "grad_norm": 1.0106462976613688,
+      "learning_rate": 4.7622302209573105e-06,
+      "loss": 0.0304,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5300826917593384,
+      "grad_norm": 1.4067320553021943,
+      "learning_rate": 4.757616917471021e-06,
+      "loss": 0.0326,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5303678357570573,
+      "grad_norm": 0.9000475309193106,
+      "learning_rate": 4.753003820802933e-06,
+      "loss": 0.0154,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5306529797547762,
+      "grad_norm": 0.7559939464533812,
+      "learning_rate": 4.748390934889261e-06,
+      "loss": 0.0252,
+      "step": 1861
+    },
+    {
+      "epoch": 0.530938123752495,
+      "grad_norm": 0.7567111911478112,
+      "learning_rate": 4.743778263666045e-06,
+      "loss": 0.0098,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5312232677502139,
+      "grad_norm": 2.6084997004648582,
+      "learning_rate": 4.739165811069135e-06,
+      "loss": 0.0385,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5315084117479327,
+      "grad_norm": 0.7239400307328033,
+      "learning_rate": 4.734553581034199e-06,
+      "loss": 0.0275,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5317935557456516,
+      "grad_norm": 0.7103481052891866,
+      "learning_rate": 4.729941577496718e-06,
+      "loss": 0.0091,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5320786997433704,
+      "grad_norm": 1.4097950329346804,
+      "learning_rate": 4.7253298043919725e-06,
+      "loss": 0.0234,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5323638437410892,
+      "grad_norm": 1.7077434751311544,
+      "learning_rate": 4.72071826565505e-06,
+      "loss": 0.0292,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5326489877388081,
+      "grad_norm": 0.5457004148411975,
+      "learning_rate": 4.71610696522084e-06,
+      "loss": 0.009,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5329341317365269,
+      "grad_norm": 0.5406920476646435,
+      "learning_rate": 4.7114959070240254e-06,
+      "loss": 0.0079,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5332192757342458,
+      "grad_norm": 1.8334680965862749,
+      "learning_rate": 4.706885094999082e-06,
+      "loss": 0.0344,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5335044197319646,
+      "grad_norm": 1.011692063206279,
+      "learning_rate": 4.70227453308028e-06,
+      "loss": 0.0244,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5337895637296834,
+      "grad_norm": 0.8936841628052329,
+      "learning_rate": 4.697664225201671e-06,
+      "loss": 0.0127,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5340747077274023,
+      "grad_norm": 0.8113649868154692,
+      "learning_rate": 4.693054175297095e-06,
+      "loss": 0.0192,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5343598517251212,
+      "grad_norm": 1.1137611632560478,
+      "learning_rate": 4.688444387300165e-06,
+      "loss": 0.0284,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5346449957228401,
+      "grad_norm": 0.9679934453395643,
+      "learning_rate": 4.683834865144277e-06,
+      "loss": 0.0222,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5349301397205589,
+      "grad_norm": 2.406256084642815,
+      "learning_rate": 4.679225612762598e-06,
+      "loss": 0.0469,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5352152837182778,
+      "grad_norm": 1.9285676628551522,
+      "learning_rate": 4.6746166340880636e-06,
+      "loss": 0.0238,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5355004277159966,
+      "grad_norm": 1.5021513223725813,
+      "learning_rate": 4.6700079330533775e-06,
+      "loss": 0.0436,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5357855717137154,
+      "grad_norm": 1.0941741453384195,
+      "learning_rate": 4.665399513591008e-06,
+      "loss": 0.0329,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5360707157114343,
+      "grad_norm": 0.39730754365810067,
+      "learning_rate": 4.660791379633175e-06,
+      "loss": 0.01,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5363558597091531,
+      "grad_norm": 1.7227973412048632,
+      "learning_rate": 4.656183535111865e-06,
+      "loss": 0.0294,
+      "step": 1881
+    },
+    {
+      "epoch": 0.536641003706872,
+      "grad_norm": 0.4622647741414944,
+      "learning_rate": 4.651575983958813e-06,
+      "loss": 0.0105,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5369261477045908,
+      "grad_norm": 0.4489009206942475,
+      "learning_rate": 4.646968730105503e-06,
+      "loss": 0.0092,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5372112917023096,
+      "grad_norm": 0.6693708589940758,
+      "learning_rate": 4.6423617774831675e-06,
+      "loss": 0.0131,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5374964357000285,
+      "grad_norm": 0.8512303652055226,
+      "learning_rate": 4.63775513002278e-06,
+      "loss": 0.0085,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5377815796977473,
+      "grad_norm": 1.6322548013574654,
+      "learning_rate": 4.633148791655053e-06,
+      "loss": 0.0318,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5380667236954663,
+      "grad_norm": 1.2405260613430487,
+      "learning_rate": 4.628542766310438e-06,
+      "loss": 0.0147,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5383518676931851,
+      "grad_norm": 0.4807074519508227,
+      "learning_rate": 4.623937057919118e-06,
+      "loss": 0.0111,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5386370116909039,
+      "grad_norm": 1.3954446874800244,
+      "learning_rate": 4.619331670411005e-06,
+      "loss": 0.0262,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5389221556886228,
+      "grad_norm": 1.116115723360862,
+      "learning_rate": 4.61472660771574e-06,
+      "loss": 0.0225,
+      "step": 1890
+    },
+    {
+      "epoch": 0.5392072996863416,
+      "grad_norm": 0.5958119317926713,
+      "learning_rate": 4.610121873762681e-06,
+      "loss": 0.0177,
+      "step": 1891
+    },
+    {
+      "epoch": 0.5394924436840605,
+      "grad_norm": 1.299507428802239,
+      "learning_rate": 4.605517472480912e-06,
+      "loss": 0.0259,
+      "step": 1892
+    },
+    {
+      "epoch": 0.5397775876817793,
+      "grad_norm": 0.8638980441903176,
+      "learning_rate": 4.6009134077992276e-06,
+      "loss": 0.0165,
+      "step": 1893
+    },
+    {
+      "epoch": 0.5400627316794981,
+      "grad_norm": 1.549781320683181,
+      "learning_rate": 4.5963096836461405e-06,
+      "loss": 0.0241,
+      "step": 1894
+    },
+    {
+      "epoch": 0.540347875677217,
+      "grad_norm": 0.8766341228007742,
+      "learning_rate": 4.5917063039498705e-06,
+      "loss": 0.0163,
+      "step": 1895
+    },
+    {
+      "epoch": 0.5406330196749358,
+      "grad_norm": 0.7287071167783781,
+      "learning_rate": 4.587103272638339e-06,
+      "loss": 0.0062,
+      "step": 1896
+    },
+    {
+      "epoch": 0.5409181636726547,
+      "grad_norm": 0.8942627426261667,
+      "learning_rate": 4.582500593639178e-06,
+      "loss": 0.0124,
+      "step": 1897
+    },
+    {
+      "epoch": 0.5412033076703735,
+      "grad_norm": 0.9701042144339818,
+      "learning_rate": 4.5778982708797146e-06,
+      "loss": 0.0094,
+      "step": 1898
+    },
+    {
+      "epoch": 0.5414884516680923,
+      "grad_norm": 0.755399450943285,
+      "learning_rate": 4.573296308286971e-06,
+      "loss": 0.0141,
+      "step": 1899
+    },
+    {
+      "epoch": 0.5417735956658112,
+      "grad_norm": 0.7506799238469901,
+      "learning_rate": 4.568694709787665e-06,
+      "loss": 0.0257,
+      "step": 1900
+    },
+    {
+      "epoch": 0.5420587396635301,
+      "grad_norm": 1.2250538112210514,
+      "learning_rate": 4.564093479308205e-06,
+      "loss": 0.0179,
+      "step": 1901
+    },
+    {
+      "epoch": 0.542343883661249,
+      "grad_norm": 1.6755192524244813,
+      "learning_rate": 4.559492620774676e-06,
+      "loss": 0.0249,
+      "step": 1902
+    },
+    {
+      "epoch": 0.5426290276589678,
+      "grad_norm": 0.8966395601407366,
+      "learning_rate": 4.554892138112856e-06,
+      "loss": 0.0127,
+      "step": 1903
+    },
+    {
+      "epoch": 0.5429141716566867,
+      "grad_norm": 1.5035988745836084,
+      "learning_rate": 4.5502920352482004e-06,
+      "loss": 0.0439,
+      "step": 1904
+    },
+    {
+      "epoch": 0.5431993156544055,
+      "grad_norm": 1.7840819508041883,
+      "learning_rate": 4.545692316105835e-06,
+      "loss": 0.0239,
+      "step": 1905
+    },
+    {
+      "epoch": 0.5434844596521243,
+      "grad_norm": 0.31678794815398437,
+      "learning_rate": 4.541092984610566e-06,
+      "loss": 0.0033,
+      "step": 1906
+    },
+    {
+      "epoch": 0.5437696036498432,
+      "grad_norm": 1.0022488627050703,
+      "learning_rate": 4.536494044686863e-06,
+      "loss": 0.0187,
+      "step": 1907
+    },
+    {
+      "epoch": 0.544054747647562,
+      "grad_norm": 2.297509688971944,
+      "learning_rate": 4.531895500258861e-06,
+      "loss": 0.0299,
+      "step": 1908
+    },
+    {
+      "epoch": 0.5443398916452808,
+      "grad_norm": 1.5802445456531595,
+      "learning_rate": 4.5272973552503614e-06,
+      "loss": 0.0303,
+      "step": 1909
+    },
+    {
+      "epoch": 0.5446250356429997,
+      "grad_norm": 0.7996338326856399,
+      "learning_rate": 4.522699613584825e-06,
+      "loss": 0.0123,
+      "step": 1910
+    },
+    {
+      "epoch": 0.5449101796407185,
+      "grad_norm": 1.5678553036220793,
+      "learning_rate": 4.5181022791853625e-06,
+      "loss": 0.0278,
+      "step": 1911
+    },
+    {
+      "epoch": 0.5451953236384374,
+      "grad_norm": 1.2576223644296152,
+      "learning_rate": 4.513505355974747e-06,
+      "loss": 0.0215,
+      "step": 1912
+    },
+    {
+      "epoch": 0.5454804676361562,
+      "grad_norm": 2.7271023649179162,
+      "learning_rate": 4.50890884787539e-06,
+      "loss": 0.0357,
+      "step": 1913
+    },
+    {
+      "epoch": 0.5457656116338752,
+      "grad_norm": 0.5599793478699585,
+      "learning_rate": 4.5043127588093545e-06,
+      "loss": 0.0136,
+      "step": 1914
+    },
+    {
+      "epoch": 0.546050755631594,
+      "grad_norm": 1.134277118458878,
+      "learning_rate": 4.4997170926983465e-06,
+      "loss": 0.0126,
+      "step": 1915
+    },
+    {
+      "epoch": 0.5463358996293128,
+      "grad_norm": 1.7917881712046206,
+      "learning_rate": 4.495121853463709e-06,
+      "loss": 0.0273,
+      "step": 1916
+    },
+    {
+      "epoch": 0.5466210436270317,
+      "grad_norm": 2.0607418747445347,
+      "learning_rate": 4.490527045026423e-06,
+      "loss": 0.0351,
+      "step": 1917
+    },
+    {
+      "epoch": 0.5469061876247505,
+      "grad_norm": 0.4922356489606496,
+      "learning_rate": 4.4859326713070996e-06,
+      "loss": 0.008,
+      "step": 1918
+    },
+    {
+      "epoch": 0.5471913316224694,
+      "grad_norm": 0.9315432349848304,
+      "learning_rate": 4.481338736225978e-06,
+      "loss": 0.0252,
+      "step": 1919
+    },
+    {
+      "epoch": 0.5474764756201882,
+      "grad_norm": 1.2682141717774775,
+      "learning_rate": 4.476745243702927e-06,
+      "loss": 0.0212,
+      "step": 1920
+    },
+    {
+      "epoch": 0.547761619617907,
+      "grad_norm": 0.850819395547047,
+      "learning_rate": 4.472152197657434e-06,
+      "loss": 0.0206,
+      "step": 1921
+    },
+    {
+      "epoch": 0.5480467636156259,
+      "grad_norm": 0.8083997376498508,
+      "learning_rate": 4.46755960200861e-06,
+      "loss": 0.0197,
+      "step": 1922
+    },
+    {
+      "epoch": 0.5483319076133447,
+      "grad_norm": 1.4339268897454358,
+      "learning_rate": 4.462967460675178e-06,
+      "loss": 0.0295,
+      "step": 1923
+    },
+    {
+      "epoch": 0.5486170516110636,
+      "grad_norm": 2.1314141503209276,
+      "learning_rate": 4.458375777575473e-06,
+      "loss": 0.056,
+      "step": 1924
+    },
+    {
+      "epoch": 0.5489021956087824,
+      "grad_norm": 1.7031785767526983,
+      "learning_rate": 4.453784556627439e-06,
+      "loss": 0.0476,
+      "step": 1925
+    },
+    {
+      "epoch": 0.5491873396065012,
+      "grad_norm": 1.2263205466527023,
+      "learning_rate": 4.449193801748629e-06,
+      "loss": 0.016,
+      "step": 1926
+    },
+    {
+      "epoch": 0.5494724836042202,
+      "grad_norm": 0.7947498299549632,
+      "learning_rate": 4.4446035168561955e-06,
+      "loss": 0.0155,
+      "step": 1927
+    },
+    {
+      "epoch": 0.549757627601939,
+      "grad_norm": 0.4966588434774226,
+      "learning_rate": 4.440013705866895e-06,
+      "loss": 0.0085,
+      "step": 1928
+    },
+    {
+      "epoch": 0.5500427715996579,
+      "grad_norm": 1.6419794017402862,
+      "learning_rate": 4.435424372697066e-06,
+      "loss": 0.0344,
+      "step": 1929
+    },
+    {
+      "epoch": 0.5503279155973767,
+      "grad_norm": 0.8164655165099096,
+      "learning_rate": 4.430835521262656e-06,
+      "loss": 0.0304,
+      "step": 1930
+    },
+    {
+      "epoch": 0.5506130595950955,
+      "grad_norm": 1.6944102182051353,
+      "learning_rate": 4.4262471554791895e-06,
+      "loss": 0.0395,
+      "step": 1931
+    },
+    {
+      "epoch": 0.5508982035928144,
+      "grad_norm": 1.7309887958840326,
+      "learning_rate": 4.421659279261785e-06,
+      "loss": 0.0425,
+      "step": 1932
+    },
+    {
+      "epoch": 0.5511833475905332,
+      "grad_norm": 2.5273881172214496,
+      "learning_rate": 4.417071896525138e-06,
+      "loss": 0.0521,
+      "step": 1933
+    },
+    {
+      "epoch": 0.5514684915882521,
+      "grad_norm": 3.4173362033274346,
+      "learning_rate": 4.412485011183525e-06,
+      "loss": 0.0638,
+      "step": 1934
+    },
+    {
+      "epoch": 0.5517536355859709,
+      "grad_norm": 0.9743219257548228,
+      "learning_rate": 4.407898627150795e-06,
+      "loss": 0.0186,
+      "step": 1935
+    },
+    {
+      "epoch": 0.5520387795836897,
+      "grad_norm": 2.349422676591217,
+      "learning_rate": 4.403312748340373e-06,
+      "loss": 0.0569,
+      "step": 1936
+    },
+    {
+      "epoch": 0.5523239235814086,
+      "grad_norm": 0.53449757299917,
+      "learning_rate": 4.398727378665252e-06,
+      "loss": 0.0097,
+      "step": 1937
+    },
+    {
+      "epoch": 0.5526090675791274,
+      "grad_norm": 1.0831620617986117,
+      "learning_rate": 4.394142522037991e-06,
+      "loss": 0.0263,
+      "step": 1938
+    },
+    {
+      "epoch": 0.5528942115768463,
+      "grad_norm": 1.383221056170318,
+      "learning_rate": 4.38955818237071e-06,
+      "loss": 0.0337,
+      "step": 1939
+    },
+    {
+      "epoch": 0.5531793555745651,
+      "grad_norm": 1.6813526071732885,
+      "learning_rate": 4.384974363575086e-06,
+      "loss": 0.0291,
+      "step": 1940
+    },
+    {
+      "epoch": 0.553464499572284,
+      "grad_norm": 0.41544282220199685,
+      "learning_rate": 4.3803910695623555e-06,
+      "loss": 0.017,
+      "step": 1941
+    },
+    {
+      "epoch": 0.5537496435700029,
+      "grad_norm": 2.098566976792298,
+      "learning_rate": 4.375808304243305e-06,
+      "loss": 0.0207,
+      "step": 1942
+    },
+    {
+      "epoch": 0.5540347875677217,
+      "grad_norm": 0.5784708780835224,
+      "learning_rate": 4.3712260715282716e-06,
+      "loss": 0.007,
+      "step": 1943
+    },
+    {
+      "epoch": 0.5543199315654406,
+      "grad_norm": 1.2225541332263947,
+      "learning_rate": 4.3666443753271355e-06,
+      "loss": 0.0319,
+      "step": 1944
+    },
+    {
+      "epoch": 0.5546050755631594,
+      "grad_norm": 1.08491250220113,
+      "learning_rate": 4.362063219549323e-06,
+      "loss": 0.0229,
+      "step": 1945
+    },
+    {
+      "epoch": 0.5548902195608783,
+      "grad_norm": 0.6817244520114021,
+      "learning_rate": 4.3574826081037935e-06,
+      "loss": 0.0214,
+      "step": 1946
+    },
+    {
+      "epoch": 0.5551753635585971,
+      "grad_norm": 1.1664569501875723,
+      "learning_rate": 4.352902544899044e-06,
+      "loss": 0.0229,
+      "step": 1947
+    },
+    {
+      "epoch": 0.5554605075563159,
+      "grad_norm": 1.293836452879822,
+      "learning_rate": 4.34832303384311e-06,
+      "loss": 0.0156,
+      "step": 1948
+    },
+    {
+      "epoch": 0.5557456515540348,
+      "grad_norm": 0.8022586547523411,
+      "learning_rate": 4.343744078843545e-06,
+      "loss": 0.0338,
+      "step": 1949
+    },
+    {
+      "epoch": 0.5560307955517536,
+      "grad_norm": 0.6083879507772137,
+      "learning_rate": 4.33916568380744e-06,
+      "loss": 0.0184,
+      "step": 1950
+    },
+    {
+      "epoch": 0.5563159395494724,
+      "grad_norm": 0.7357572185407785,
+      "learning_rate": 4.334587852641396e-06,
+      "loss": 0.0127,
+      "step": 1951
+    },
+    {
+      "epoch": 0.5566010835471913,
+      "grad_norm": 0.7103951655209286,
+      "learning_rate": 4.330010589251541e-06,
+      "loss": 0.0137,
+      "step": 1952
+    },
+    {
+      "epoch": 0.5568862275449101,
+      "grad_norm": 2.0436097540893847,
+      "learning_rate": 4.325433897543516e-06,
+      "loss": 0.0439,
+      "step": 1953
+    },
+    {
+      "epoch": 0.5571713715426291,
+      "grad_norm": 1.0347939995255582,
+      "learning_rate": 4.320857781422476e-06,
+      "loss": 0.0144,
+      "step": 1954
+    },
+    {
+      "epoch": 0.5574565155403479,
+      "grad_norm": 1.2205245382671843,
+      "learning_rate": 4.31628224479308e-06,
+      "loss": 0.0343,
+      "step": 1955
+    },
+    {
+      "epoch": 0.5577416595380668,
+      "grad_norm": 0.48096511602458747,
+      "learning_rate": 4.311707291559502e-06,
+      "loss": 0.0151,
+      "step": 1956
+    },
+    {
+      "epoch": 0.5580268035357856,
+      "grad_norm": 1.858218874925223,
+      "learning_rate": 4.307132925625405e-06,
+      "loss": 0.024,
+      "step": 1957
+    },
+    {
+      "epoch": 0.5583119475335044,
+      "grad_norm": 0.41558860469055026,
+      "learning_rate": 4.302559150893962e-06,
+      "loss": 0.0081,
+      "step": 1958
+    },
+    {
+      "epoch": 0.5585970915312233,
+      "grad_norm": 0.6185726461872646,
+      "learning_rate": 4.297985971267836e-06,
+      "loss": 0.0148,
+      "step": 1959
+    },
+    {
+      "epoch": 0.5588822355289421,
+      "grad_norm": 1.629041987647595,
+      "learning_rate": 4.293413390649186e-06,
+      "loss": 0.0382,
+      "step": 1960
+    },
+    {
+      "epoch": 0.559167379526661,
+      "grad_norm": 1.1013738682804133,
+      "learning_rate": 4.288841412939656e-06,
+      "loss": 0.0123,
+      "step": 1961
+    },
+    {
+      "epoch": 0.5594525235243798,
+      "grad_norm": 0.7207246603680382,
+      "learning_rate": 4.284270042040377e-06,
+      "loss": 0.0092,
+      "step": 1962
+    },
+    {
+      "epoch": 0.5597376675220986,
+      "grad_norm": 1.3146322345300236,
+      "learning_rate": 4.279699281851962e-06,
+      "loss": 0.0497,
+      "step": 1963
+    },
+    {
+      "epoch": 0.5600228115198175,
+      "grad_norm": 1.3720103682911193,
+      "learning_rate": 4.275129136274505e-06,
+      "loss": 0.02,
+      "step": 1964
+    },
+    {
+      "epoch": 0.5603079555175363,
+      "grad_norm": 0.4413728823645133,
+      "learning_rate": 4.270559609207572e-06,
+      "loss": 0.0068,
+      "step": 1965
+    },
+    {
+      "epoch": 0.5605930995152552,
+      "grad_norm": 1.042243710858581,
+      "learning_rate": 4.265990704550204e-06,
+      "loss": 0.0202,
+      "step": 1966
+    },
+    {
+      "epoch": 0.5608782435129741,
+      "grad_norm": 1.6738885287470393,
+      "learning_rate": 4.261422426200912e-06,
+      "loss": 0.05,
+      "step": 1967
+    },
+    {
+      "epoch": 0.561163387510693,
+      "grad_norm": 0.6925766021391758,
+      "learning_rate": 4.256854778057668e-06,
+      "loss": 0.0084,
+      "step": 1968
+    },
+    {
+      "epoch": 0.5614485315084118,
+      "grad_norm": 1.3643708719882088,
+      "learning_rate": 4.252287764017908e-06,
+      "loss": 0.0386,
+      "step": 1969
+    },
+    {
+      "epoch": 0.5617336755061306,
+      "grad_norm": 0.5506155217176552,
+      "learning_rate": 4.24772138797853e-06,
+      "loss": 0.0084,
+      "step": 1970
+    },
+    {
+      "epoch": 0.5620188195038495,
+      "grad_norm": 0.555170139364492,
+      "learning_rate": 4.243155653835886e-06,
+      "loss": 0.0118,
+      "step": 1971
+    },
+    {
+      "epoch": 0.5623039635015683,
+      "grad_norm": 0.9763249846495575,
+      "learning_rate": 4.238590565485779e-06,
+      "loss": 0.0139,
+      "step": 1972
+    },
+    {
+      "epoch": 0.5625891074992871,
+      "grad_norm": 0.7772741042704194,
+      "learning_rate": 4.23402612682346e-06,
+      "loss": 0.0125,
+      "step": 1973
+    },
+    {
+      "epoch": 0.562874251497006,
+      "grad_norm": 0.40993352316801507,
+      "learning_rate": 4.2294623417436284e-06,
+      "loss": 0.0057,
+      "step": 1974
+    },
+    {
+      "epoch": 0.5631593954947248,
+      "grad_norm": 2.595191318759739,
+      "learning_rate": 4.224899214140425e-06,
+      "loss": 0.0483,
+      "step": 1975
+    },
+    {
+      "epoch": 0.5634445394924437,
+      "grad_norm": 1.0047435454745783,
+      "learning_rate": 4.220336747907431e-06,
+      "loss": 0.0387,
+      "step": 1976
+    },
+    {
+      "epoch": 0.5637296834901625,
+      "grad_norm": 0.8252660492184049,
+      "learning_rate": 4.2157749469376594e-06,
+      "loss": 0.0179,
+      "step": 1977
+    },
+    {
+      "epoch": 0.5640148274878813,
+      "grad_norm": 1.7631265123768307,
+      "learning_rate": 4.21121381512356e-06,
+      "loss": 0.0259,
+      "step": 1978
+    },
+    {
+      "epoch": 0.5642999714856002,
+      "grad_norm": 1.1505145111500783,
+      "learning_rate": 4.206653356357007e-06,
+      "loss": 0.0294,
+      "step": 1979
+    },
+    {
+      "epoch": 0.564585115483319,
+      "grad_norm": 1.4455244317990026,
+      "learning_rate": 4.202093574529305e-06,
+      "loss": 0.0312,
+      "step": 1980
+    },
+    {
+      "epoch": 0.564870259481038,
+      "grad_norm": 0.8626915090427404,
+      "learning_rate": 4.197534473531177e-06,
+      "loss": 0.0276,
+      "step": 1981
+    },
+    {
+      "epoch": 0.5651554034787568,
+      "grad_norm": 1.5858817877250049,
+      "learning_rate": 4.192976057252768e-06,
+      "loss": 0.0192,
+      "step": 1982
+    },
+    {
+      "epoch": 0.5654405474764757,
+      "grad_norm": 1.410131591822338,
+      "learning_rate": 4.18841832958364e-06,
+      "loss": 0.0247,
+      "step": 1983
+    },
+    {
+      "epoch": 0.5657256914741945,
+      "grad_norm": 0.7445170622774078,
+      "learning_rate": 4.18386129441276e-06,
+      "loss": 0.0121,
+      "step": 1984
+    },
+    {
+      "epoch": 0.5660108354719133,
+      "grad_norm": 0.8809924184189645,
+      "learning_rate": 4.179304955628511e-06,
+      "loss": 0.019,
+      "step": 1985
+    },
+    {
+      "epoch": 0.5662959794696322,
+      "grad_norm": 1.5083664563446835,
+      "learning_rate": 4.174749317118683e-06,
+      "loss": 0.0261,
+      "step": 1986
+    },
+    {
+      "epoch": 0.566581123467351,
+      "grad_norm": 2.2901675873569656,
+      "learning_rate": 4.170194382770462e-06,
+      "loss": 0.0279,
+      "step": 1987
+    },
+    {
+      "epoch": 0.5668662674650699,
+      "grad_norm": 1.5826055506344388,
+      "learning_rate": 4.165640156470436e-06,
+      "loss": 0.0337,
+      "step": 1988
+    },
+    {
+      "epoch": 0.5671514114627887,
+      "grad_norm": 1.486333132761605,
+      "learning_rate": 4.161086642104594e-06,
+      "loss": 0.0147,
+      "step": 1989
+    },
+    {
+      "epoch": 0.5674365554605075,
+      "grad_norm": 1.2557151112171123,
+      "learning_rate": 4.156533843558309e-06,
+      "loss": 0.0224,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5677216994582264,
+      "grad_norm": 0.682005313725667,
+      "learning_rate": 4.151981764716347e-06,
+      "loss": 0.0079,
+      "step": 1991
+    },
+    {
+      "epoch": 0.5680068434559452,
+      "grad_norm": 2.0989444982377665,
+      "learning_rate": 4.147430409462863e-06,
+      "loss": 0.0654,
+      "step": 1992
+    },
+    {
+      "epoch": 0.568291987453664,
+      "grad_norm": 1.3172125697333474,
+      "learning_rate": 4.142879781681389e-06,
+      "loss": 0.0382,
+      "step": 1993
+    },
+    {
+      "epoch": 0.568577131451383,
+      "grad_norm": 1.0200075411284906,
+      "learning_rate": 4.138329885254842e-06,
+      "loss": 0.0155,
+      "step": 1994
+    },
+    {
+      "epoch": 0.5688622754491018,
+      "grad_norm": 1.6810817501742972,
+      "learning_rate": 4.133780724065508e-06,
+      "loss": 0.0284,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5691474194468207,
+      "grad_norm": 0.3058438071030809,
+      "learning_rate": 4.1292323019950515e-06,
+      "loss": 0.0057,
+      "step": 1996
+    },
+    {
+      "epoch": 0.5694325634445395,
+      "grad_norm": 3.2345016887299174,
+      "learning_rate": 4.124684622924503e-06,
+      "loss": 0.0603,
+      "step": 1997
+    },
+    {
+      "epoch": 0.5697177074422584,
+      "grad_norm": 0.25562993510263476,
+      "learning_rate": 4.120137690734264e-06,
+      "loss": 0.0058,
+      "step": 1998
+    },
+    {
+      "epoch": 0.5700028514399772,
+      "grad_norm": 0.9607045079360945,
+      "learning_rate": 4.115591509304092e-06,
+      "loss": 0.019,
+      "step": 1999
+    },
+    {
+      "epoch": 0.570287995437696,
+      "grad_norm": 0.6007173797861988,
+      "learning_rate": 4.111046082513109e-06,
+      "loss": 0.0082,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5705731394354149,
+      "grad_norm": 0.8863429154698693,
+      "learning_rate": 4.106501414239787e-06,
+      "loss": 0.0118,
+      "step": 2001
+    },
+    {
+      "epoch": 0.5708582834331337,
+      "grad_norm": 1.5083003179196242,
+      "learning_rate": 4.1019575083619595e-06,
+      "loss": 0.0158,
+      "step": 2002
+    },
+    {
+      "epoch": 0.5711434274308526,
+      "grad_norm": 0.7391504005003044,
+      "learning_rate": 4.097414368756803e-06,
+      "loss": 0.0088,
+      "step": 2003
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.6832287705746387,
+      "learning_rate": 4.092871999300841e-06,
+      "loss": 0.0094,
+      "step": 2004
+    },
+    {
+      "epoch": 0.5717137154262902,
+      "grad_norm": 1.6772859676317076,
+      "learning_rate": 4.088330403869943e-06,
+      "loss": 0.0478,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5719988594240091,
+      "grad_norm": 1.0306291075921608,
+      "learning_rate": 4.083789586339315e-06,
+      "loss": 0.0237,
+      "step": 2006
+    },
+    {
+      "epoch": 0.572284003421728,
+      "grad_norm": 1.184619272802862,
+      "learning_rate": 4.079249550583498e-06,
+      "loss": 0.0283,
+      "step": 2007
+    },
+    {
+      "epoch": 0.5725691474194469,
+      "grad_norm": 1.4096930747734462,
+      "learning_rate": 4.074710300476372e-06,
+      "loss": 0.014,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5728542914171657,
+      "grad_norm": 0.783881045789122,
+      "learning_rate": 4.07017183989114e-06,
+      "loss": 0.0115,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5731394354148845,
+      "grad_norm": 0.30172775611238883,
+      "learning_rate": 4.065634172700335e-06,
+      "loss": 0.0045,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5734245794126034,
+      "grad_norm": 0.6976220498401472,
+      "learning_rate": 4.0610973027758124e-06,
+      "loss": 0.0272,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5737097234103222,
+      "grad_norm": 1.1052303028320305,
+      "learning_rate": 4.0565612339887455e-06,
+      "loss": 0.017,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5739948674080411,
+      "grad_norm": 1.7404973231076455,
+      "learning_rate": 4.052025970209626e-06,
+      "loss": 0.0392,
+      "step": 2013
+    },
+    {
+      "epoch": 0.5742800114057599,
+      "grad_norm": 0.674052952485932,
+      "learning_rate": 4.047491515308259e-06,
+      "loss": 0.0082,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5745651554034787,
+      "grad_norm": 0.7117202921993683,
+      "learning_rate": 4.042957873153758e-06,
+      "loss": 0.0178,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5748502994011976,
+      "grad_norm": 1.1146848695110856,
+      "learning_rate": 4.038425047614545e-06,
+      "loss": 0.0295,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5751354433989164,
+      "grad_norm": 0.4209229851025855,
+      "learning_rate": 4.033893042558341e-06,
+      "loss": 0.0121,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5754205873966353,
+      "grad_norm": 0.48277471651577564,
+      "learning_rate": 4.029361861852171e-06,
+      "loss": 0.0069,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5757057313943541,
+      "grad_norm": 1.7222798865166078,
+      "learning_rate": 4.024831509362357e-06,
+      "loss": 0.0298,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5759908753920729,
+      "grad_norm": 1.0337635489366164,
+      "learning_rate": 4.020301988954511e-06,
+      "loss": 0.0159,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5762760193897919,
+      "grad_norm": 1.3775357400508321,
+      "learning_rate": 4.015773304493536e-06,
+      "loss": 0.0158,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5765611633875107,
+      "grad_norm": 0.5083277988878048,
+      "learning_rate": 4.011245459843625e-06,
+      "loss": 0.0067,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5768463073852296,
+      "grad_norm": 0.5286132139270061,
+      "learning_rate": 4.006718458868248e-06,
+      "loss": 0.0062,
+      "step": 2023
+    },
+    {
+      "epoch": 0.5771314513829484,
+      "grad_norm": 1.1038055454886817,
+      "learning_rate": 4.002192305430162e-06,
+      "loss": 0.0129,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5774165953806673,
+      "grad_norm": 1.1714514779633876,
+      "learning_rate": 3.997667003391397e-06,
+      "loss": 0.031,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5777017393783861,
+      "grad_norm": 1.7218432058203235,
+      "learning_rate": 3.993142556613255e-06,
+      "loss": 0.0407,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5779868833761049,
+      "grad_norm": 1.3615661214766888,
+      "learning_rate": 3.988618968956315e-06,
+      "loss": 0.0442,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5782720273738238,
+      "grad_norm": 1.9624857942185958,
+      "learning_rate": 3.9840962442804145e-06,
+      "loss": 0.0478,
+      "step": 2028
+    },
+    {
+      "epoch": 0.5785571713715426,
+      "grad_norm": 0.4209016778937654,
+      "learning_rate": 3.97957438644466e-06,
+      "loss": 0.0074,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5788423153692615,
+      "grad_norm": 2.0516788767101795,
+      "learning_rate": 3.975053399307416e-06,
+      "loss": 0.0409,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5791274593669803,
+      "grad_norm": 0.8532747901837053,
+      "learning_rate": 3.970533286726306e-06,
+      "loss": 0.0157,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5794126033646991,
+      "grad_norm": 0.8877673247289528,
+      "learning_rate": 3.966014052558206e-06,
+      "loss": 0.0167,
+      "step": 2032
+    },
+    {
+      "epoch": 0.579697747362418,
+      "grad_norm": 1.0452094981847408,
+      "learning_rate": 3.961495700659243e-06,
+      "loss": 0.0177,
+      "step": 2033
+    },
+    {
+      "epoch": 0.5799828913601369,
+      "grad_norm": 2.336128746094747,
+      "learning_rate": 3.956978234884791e-06,
+      "loss": 0.0511,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5802680353578558,
+      "grad_norm": 1.3293832741262577,
+      "learning_rate": 3.952461659089467e-06,
+      "loss": 0.0206,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5805531793555746,
+      "grad_norm": 3.2429992646007246,
+      "learning_rate": 3.947945977127129e-06,
+      "loss": 0.0435,
+      "step": 2036
+    },
+    {
+      "epoch": 0.5808383233532934,
+      "grad_norm": 1.4889692622010537,
+      "learning_rate": 3.943431192850874e-06,
+      "loss": 0.0168,
+      "step": 2037
+    },
+    {
+      "epoch": 0.5811234673510123,
+      "grad_norm": 0.7779287362969187,
+      "learning_rate": 3.93891731011303e-06,
+      "loss": 0.0081,
+      "step": 2038
+    },
+    {
+      "epoch": 0.5814086113487311,
+      "grad_norm": 1.1967880354641325,
+      "learning_rate": 3.934404332765161e-06,
+      "loss": 0.0154,
+      "step": 2039
+    },
+    {
+      "epoch": 0.58169375534645,
+      "grad_norm": 1.1772392533629603,
+      "learning_rate": 3.929892264658052e-06,
+      "loss": 0.0269,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5819788993441688,
+      "grad_norm": 1.8826797992290683,
+      "learning_rate": 3.925381109641715e-06,
+      "loss": 0.0459,
+      "step": 2041
+    },
+    {
+      "epoch": 0.5822640433418876,
+      "grad_norm": 0.8638860916601324,
+      "learning_rate": 3.9208708715653825e-06,
+      "loss": 0.0174,
+      "step": 2042
+    },
+    {
+      "epoch": 0.5825491873396065,
+      "grad_norm": 1.0986393666105592,
+      "learning_rate": 3.916361554277508e-06,
+      "loss": 0.019,
+      "step": 2043
+    },
+    {
+      "epoch": 0.5828343313373253,
+      "grad_norm": 0.9183125572551752,
+      "learning_rate": 3.911853161625756e-06,
+      "loss": 0.0183,
+      "step": 2044
+    },
+    {
+      "epoch": 0.5831194753350442,
+      "grad_norm": 1.9954417238979274,
+      "learning_rate": 3.9073456974570004e-06,
+      "loss": 0.0218,
+      "step": 2045
+    },
+    {
+      "epoch": 0.583404619332763,
+      "grad_norm": 0.6191794329765098,
+      "learning_rate": 3.902839165617323e-06,
+      "loss": 0.014,
+      "step": 2046
+    },
+    {
+      "epoch": 0.583689763330482,
+      "grad_norm": 1.4231403686313335,
+      "learning_rate": 3.898333569952018e-06,
+      "loss": 0.0296,
+      "step": 2047
+    },
+    {
+      "epoch": 0.5839749073282008,
+      "grad_norm": 0.8776537238404536,
+      "learning_rate": 3.89382891430557e-06,
+      "loss": 0.0083,
+      "step": 2048
+    },
+    {
+      "epoch": 0.5842600513259196,
+      "grad_norm": 0.8708883202320266,
+      "learning_rate": 3.889325202521668e-06,
+      "loss": 0.0202,
+      "step": 2049
+    },
+    {
+      "epoch": 0.5845451953236385,
+      "grad_norm": 0.8831775072235944,
+      "learning_rate": 3.884822438443196e-06,
+      "loss": 0.024,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5848303393213573,
+      "grad_norm": 2.2216278743868196,
+      "learning_rate": 3.880320625912224e-06,
+      "loss": 0.0326,
+      "step": 2051
+    },
+    {
+      "epoch": 0.5851154833190761,
+      "grad_norm": 0.6026223557249291,
+      "learning_rate": 3.875819768770012e-06,
+      "loss": 0.0197,
+      "step": 2052
+    },
+    {
+      "epoch": 0.585400627316795,
+      "grad_norm": 1.4057182511451267,
+      "learning_rate": 3.87131987085701e-06,
+      "loss": 0.028,
+      "step": 2053
+    },
+    {
+      "epoch": 0.5856857713145138,
+      "grad_norm": 1.487605532299638,
+      "learning_rate": 3.866820936012844e-06,
+      "loss": 0.0113,
+      "step": 2054
+    },
+    {
+      "epoch": 0.5859709153122327,
+      "grad_norm": 1.4068652241281197,
+      "learning_rate": 3.862322968076322e-06,
+      "loss": 0.0553,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5862560593099515,
+      "grad_norm": 0.6717247126069004,
+      "learning_rate": 3.857825970885422e-06,
+      "loss": 0.0162,
+      "step": 2056
+    },
+    {
+      "epoch": 0.5865412033076703,
+      "grad_norm": 0.2160215600154683,
+      "learning_rate": 3.853329948277297e-06,
+      "loss": 0.0048,
+      "step": 2057
+    },
+    {
+      "epoch": 0.5868263473053892,
+      "grad_norm": 0.7554944197765426,
+      "learning_rate": 3.848834904088268e-06,
+      "loss": 0.0157,
+      "step": 2058
+    },
+    {
+      "epoch": 0.587111491303108,
+      "grad_norm": 0.893784839410386,
+      "learning_rate": 3.844340842153823e-06,
+      "loss": 0.0186,
+      "step": 2059
+    },
+    {
+      "epoch": 0.5873966353008269,
+      "grad_norm": 0.742558705028012,
+      "learning_rate": 3.839847766308607e-06,
+      "loss": 0.0158,
+      "step": 2060
+    },
+    {
+      "epoch": 0.5876817792985458,
+      "grad_norm": 1.8275272220434577,
+      "learning_rate": 3.8353556803864324e-06,
+      "loss": 0.0421,
+      "step": 2061
+    },
+    {
+      "epoch": 0.5879669232962647,
+      "grad_norm": 0.4661961963077854,
+      "learning_rate": 3.830864588220253e-06,
+      "loss": 0.0086,
+      "step": 2062
+    },
+    {
+      "epoch": 0.5882520672939835,
+      "grad_norm": 0.4600167908756378,
+      "learning_rate": 3.826374493642187e-06,
+      "loss": 0.0143,
+      "step": 2063
+    },
+    {
+      "epoch": 0.5885372112917023,
+      "grad_norm": 0.9455308154957766,
+      "learning_rate": 3.821885400483497e-06,
+      "loss": 0.021,
+      "step": 2064
+    },
+    {
+      "epoch": 0.5888223552894212,
+      "grad_norm": 0.7579095990471869,
+      "learning_rate": 3.817397312574592e-06,
+      "loss": 0.0143,
+      "step": 2065
+    },
+    {
+      "epoch": 0.58910749928714,
+      "grad_norm": 1.4968058315352246,
+      "learning_rate": 3.8129102337450236e-06,
+      "loss": 0.0333,
+      "step": 2066
+    },
+    {
+      "epoch": 0.5893926432848589,
+      "grad_norm": 1.1681439442987638,
+      "learning_rate": 3.8084241678234777e-06,
+      "loss": 0.0183,
+      "step": 2067
+    },
+    {
+      "epoch": 0.5896777872825777,
+      "grad_norm": 1.0976297564753572,
+      "learning_rate": 3.8039391186377806e-06,
+      "loss": 0.0132,
+      "step": 2068
+    },
+    {
+      "epoch": 0.5899629312802965,
+      "grad_norm": 1.8400386955643093,
+      "learning_rate": 3.7994550900148914e-06,
+      "loss": 0.0233,
+      "step": 2069
+    },
+    {
+      "epoch": 0.5902480752780154,
+      "grad_norm": 0.8935945500146714,
+      "learning_rate": 3.7949720857808953e-06,
+      "loss": 0.0119,
+      "step": 2070
+    },
+    {
+      "epoch": 0.5905332192757342,
+      "grad_norm": 1.0682714756646516,
+      "learning_rate": 3.7904901097610074e-06,
+      "loss": 0.0108,
+      "step": 2071
+    },
+    {
+      "epoch": 0.590818363273453,
+      "grad_norm": 0.9744085122592351,
+      "learning_rate": 3.786009165779563e-06,
+      "loss": 0.0099,
+      "step": 2072
+    },
+    {
+      "epoch": 0.5911035072711719,
+      "grad_norm": 1.3356235518474884,
+      "learning_rate": 3.7815292576600143e-06,
+      "loss": 0.0374,
+      "step": 2073
+    },
+    {
+      "epoch": 0.5913886512688908,
+      "grad_norm": 2.052387547798954,
+      "learning_rate": 3.777050389224933e-06,
+      "loss": 0.0582,
+      "step": 2074
+    },
+    {
+      "epoch": 0.5916737952666097,
+      "grad_norm": 1.021714477749963,
+      "learning_rate": 3.7725725642960047e-06,
+      "loss": 0.0175,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5919589392643285,
+      "grad_norm": 0.6765650754941559,
+      "learning_rate": 3.768095786694021e-06,
+      "loss": 0.0131,
+      "step": 2076
+    },
+    {
+      "epoch": 0.5922440832620474,
+      "grad_norm": 0.5186344994748767,
+      "learning_rate": 3.7636200602388855e-06,
+      "loss": 0.0061,
+      "step": 2077
+    },
+    {
+      "epoch": 0.5925292272597662,
+      "grad_norm": 0.6496702410485596,
+      "learning_rate": 3.759145388749595e-06,
+      "loss": 0.008,
+      "step": 2078
+    },
+    {
+      "epoch": 0.592814371257485,
+      "grad_norm": 1.003400177949479,
+      "learning_rate": 3.7546717760442565e-06,
+      "loss": 0.0215,
+      "step": 2079
+    },
+    {
+      "epoch": 0.5930995152552039,
+      "grad_norm": 1.4201821861215709,
+      "learning_rate": 3.7501992259400665e-06,
+      "loss": 0.0193,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5933846592529227,
+      "grad_norm": 1.7584435403151972,
+      "learning_rate": 3.7457277422533213e-06,
+      "loss": 0.03,
+      "step": 2081
+    },
+    {
+      "epoch": 0.5936698032506416,
+      "grad_norm": 1.1121101189301186,
+      "learning_rate": 3.7412573287994e-06,
+      "loss": 0.0168,
+      "step": 2082
+    },
+    {
+      "epoch": 0.5939549472483604,
+      "grad_norm": 0.9801459726894103,
+      "learning_rate": 3.7367879893927763e-06,
+      "loss": 0.0236,
+      "step": 2083
+    },
+    {
+      "epoch": 0.5942400912460792,
+      "grad_norm": 0.9674593323408074,
+      "learning_rate": 3.732319727846998e-06,
+      "loss": 0.0179,
+      "step": 2084
+    },
+    {
+      "epoch": 0.5945252352437981,
+      "grad_norm": 1.222026333353283,
+      "learning_rate": 3.7278525479747014e-06,
+      "loss": 0.0365,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5948103792415169,
+      "grad_norm": 2.025476959542898,
+      "learning_rate": 3.7233864535875953e-06,
+      "loss": 0.0373,
+      "step": 2086
+    },
+    {
+      "epoch": 0.5950955232392359,
+      "grad_norm": 1.0776536635630782,
+      "learning_rate": 3.7189214484964663e-06,
+      "loss": 0.0215,
+      "step": 2087
+    },
+    {
+      "epoch": 0.5953806672369547,
+      "grad_norm": 1.1705278143136957,
+      "learning_rate": 3.7144575365111677e-06,
+      "loss": 0.0114,
+      "step": 2088
+    },
+    {
+      "epoch": 0.5956658112346735,
+      "grad_norm": 2.401216532684097,
+      "learning_rate": 3.7099947214406217e-06,
+      "loss": 0.032,
+      "step": 2089
+    },
+    {
+      "epoch": 0.5959509552323924,
+      "grad_norm": 1.6998711218014348,
+      "learning_rate": 3.705533007092812e-06,
+      "loss": 0.0462,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5962360992301112,
+      "grad_norm": 1.5985776456990062,
+      "learning_rate": 3.7010723972747887e-06,
+      "loss": 0.045,
+      "step": 2091
+    },
+    {
+      "epoch": 0.5965212432278301,
+      "grad_norm": 0.22163354454740036,
+      "learning_rate": 3.6966128957926528e-06,
+      "loss": 0.0035,
+      "step": 2092
+    },
+    {
+      "epoch": 0.5968063872255489,
+      "grad_norm": 0.860485640046081,
+      "learning_rate": 3.6921545064515667e-06,
+      "loss": 0.0232,
+      "step": 2093
+    },
+    {
+      "epoch": 0.5970915312232677,
+      "grad_norm": 0.19848014052903284,
+      "learning_rate": 3.6876972330557393e-06,
+      "loss": 0.0023,
+      "step": 2094
+    },
+    {
+      "epoch": 0.5973766752209866,
+      "grad_norm": 1.194915635682168,
+      "learning_rate": 3.6832410794084255e-06,
+      "loss": 0.0232,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5976618192187054,
+      "grad_norm": 1.6969899400898916,
+      "learning_rate": 3.6787860493119274e-06,
+      "loss": 0.0479,
+      "step": 2096
+    },
+    {
+      "epoch": 0.5979469632164243,
+      "grad_norm": 0.6961043811511708,
+      "learning_rate": 3.6743321465675918e-06,
+      "loss": 0.0059,
+      "step": 2097
+    },
+    {
+      "epoch": 0.5982321072141431,
+      "grad_norm": 1.5504097165767623,
+      "learning_rate": 3.669879374975796e-06,
+      "loss": 0.0533,
+      "step": 2098
+    },
+    {
+      "epoch": 0.598517251211862,
+      "grad_norm": 0.6225399533823082,
+      "learning_rate": 3.6654277383359613e-06,
+      "loss": 0.0241,
+      "step": 2099
+    },
+    {
+      "epoch": 0.5988023952095808,
+      "grad_norm": 0.8199738615224776,
+      "learning_rate": 3.6609772404465293e-06,
+      "loss": 0.0153,
+      "step": 2100
+    },
+    {
+      "epoch": 0.5990875392072997,
+      "grad_norm": 2.6367146702980135,
+      "learning_rate": 3.6565278851049803e-06,
+      "loss": 0.0232,
+      "step": 2101
+    },
+    {
+      "epoch": 0.5993726832050186,
+      "grad_norm": 2.218376029082567,
+      "learning_rate": 3.6520796761078126e-06,
+      "loss": 0.0264,
+      "step": 2102
+    },
+    {
+      "epoch": 0.5996578272027374,
+      "grad_norm": 0.42063773687844663,
+      "learning_rate": 3.6476326172505516e-06,
+      "loss": 0.0077,
+      "step": 2103
+    },
+    {
+      "epoch": 0.5999429712004563,
+      "grad_norm": 1.4846079891375457,
+      "learning_rate": 3.6431867123277393e-06,
+      "loss": 0.0274,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6002281151981751,
+      "grad_norm": 2.0672737709470628,
+      "learning_rate": 3.6387419651329326e-06,
+      "loss": 0.0252,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6005132591958939,
+      "grad_norm": 1.3500653562887743,
+      "learning_rate": 3.6342983794586974e-06,
+      "loss": 0.0152,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6007984031936128,
+      "grad_norm": 1.7611157268739988,
+      "learning_rate": 3.6298559590966153e-06,
+      "loss": 0.0286,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6010835471913316,
+      "grad_norm": 1.2728775668088093,
+      "learning_rate": 3.625414707837268e-06,
+      "loss": 0.0179,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6013686911890505,
+      "grad_norm": 0.9960972880229151,
+      "learning_rate": 3.6209746294702442e-06,
+      "loss": 0.0232,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6016538351867693,
+      "grad_norm": 0.6157173200508199,
+      "learning_rate": 3.6165357277841294e-06,
+      "loss": 0.0134,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6019389791844881,
+      "grad_norm": 0.8400067311463636,
+      "learning_rate": 3.6120980065665023e-06,
+      "loss": 0.0112,
+      "step": 2111
+    },
+    {
+      "epoch": 0.602224123182207,
+      "grad_norm": 1.2339712983273081,
+      "learning_rate": 3.607661469603937e-06,
+      "loss": 0.0239,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6025092671799258,
+      "grad_norm": 1.5580285145203188,
+      "learning_rate": 3.6032261206819995e-06,
+      "loss": 0.0412,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6027944111776448,
+      "grad_norm": 1.1722662736059852,
+      "learning_rate": 3.598791963585239e-06,
+      "loss": 0.0152,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6030795551753636,
+      "grad_norm": 1.2912381233776336,
+      "learning_rate": 3.5943590020971873e-06,
+      "loss": 0.0272,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6033646991730824,
+      "grad_norm": 2.1137858949289448,
+      "learning_rate": 3.5899272400003603e-06,
+      "loss": 0.0288,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6036498431708013,
+      "grad_norm": 0.40725160937380805,
+      "learning_rate": 3.585496681076244e-06,
+      "loss": 0.005,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6039349871685201,
+      "grad_norm": 1.069151927238274,
+      "learning_rate": 3.581067329105301e-06,
+      "loss": 0.0286,
+      "step": 2118
+    },
+    {
+      "epoch": 0.604220131166239,
+      "grad_norm": 1.6223267876797507,
+      "learning_rate": 3.5766391878669676e-06,
+      "loss": 0.028,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6045052751639578,
+      "grad_norm": 0.9274658583181319,
+      "learning_rate": 3.5722122611396416e-06,
+      "loss": 0.0124,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6047904191616766,
+      "grad_norm": 1.280180345583819,
+      "learning_rate": 3.5677865527006876e-06,
+      "loss": 0.0377,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6050755631593955,
+      "grad_norm": 1.4904942621790036,
+      "learning_rate": 3.563362066326427e-06,
+      "loss": 0.0263,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6053607071571143,
+      "grad_norm": 1.2524362229042842,
+      "learning_rate": 3.5589388057921435e-06,
+      "loss": 0.0267,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6056458511548332,
+      "grad_norm": 4.061187413865505,
+      "learning_rate": 3.5545167748720705e-06,
+      "loss": 0.0713,
+      "step": 2124
+    },
+    {
+      "epoch": 0.605930995152552,
+      "grad_norm": 0.5362589492288503,
+      "learning_rate": 3.550095977339396e-06,
+      "loss": 0.0094,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6062161391502708,
+      "grad_norm": 0.8432265034751789,
+      "learning_rate": 3.545676416966254e-06,
+      "loss": 0.0184,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6065012831479898,
+      "grad_norm": 1.3167325005559367,
+      "learning_rate": 3.541258097523722e-06,
+      "loss": 0.0289,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6067864271457086,
+      "grad_norm": 1.0810373994278544,
+      "learning_rate": 3.536841022781816e-06,
+      "loss": 0.019,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6070715711434275,
+      "grad_norm": 1.2837062613842711,
+      "learning_rate": 3.532425196509498e-06,
+      "loss": 0.0268,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6073567151411463,
+      "grad_norm": 1.1490746860147354,
+      "learning_rate": 3.5280106224746575e-06,
+      "loss": 0.0251,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6076418591388651,
+      "grad_norm": 1.0489665093288734,
+      "learning_rate": 3.5235973044441163e-06,
+      "loss": 0.018,
+      "step": 2131
+    },
+    {
+      "epoch": 0.607927003136584,
+      "grad_norm": 1.726137132954225,
+      "learning_rate": 3.5191852461836306e-06,
+      "loss": 0.0241,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6082121471343028,
+      "grad_norm": 0.7090919389506171,
+      "learning_rate": 3.514774451457873e-06,
+      "loss": 0.0186,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6084972911320217,
+      "grad_norm": 0.4590722376042221,
+      "learning_rate": 3.510364924030443e-06,
+      "loss": 0.0097,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6087824351297405,
+      "grad_norm": 1.1018085553098087,
+      "learning_rate": 3.505956667663859e-06,
+      "loss": 0.015,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6090675791274593,
+      "grad_norm": 0.6254529574769264,
+      "learning_rate": 3.5015496861195526e-06,
+      "loss": 0.0135,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6093527231251782,
+      "grad_norm": 1.6428276895681884,
+      "learning_rate": 3.497143983157868e-06,
+      "loss": 0.0266,
+      "step": 2137
+    },
+    {
+      "epoch": 0.609637867122897,
+      "grad_norm": 1.1906175686284322,
+      "learning_rate": 3.4927395625380626e-06,
+      "loss": 0.0403,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6099230111206159,
+      "grad_norm": 0.7327641862389733,
+      "learning_rate": 3.488336428018293e-06,
+      "loss": 0.0166,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6102081551183347,
+      "grad_norm": 1.8597660390527846,
+      "learning_rate": 3.4839345833556217e-06,
+      "loss": 0.0324,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6104932991160537,
+      "grad_norm": 3.452549046142695,
+      "learning_rate": 3.479534032306011e-06,
+      "loss": 0.0538,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6107784431137725,
+      "grad_norm": 1.2114221044219349,
+      "learning_rate": 3.4751347786243193e-06,
+      "loss": 0.0152,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6110635871114913,
+      "grad_norm": 1.2721385499066669,
+      "learning_rate": 3.470736826064299e-06,
+      "loss": 0.0175,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6113487311092102,
+      "grad_norm": 1.6293019383768632,
+      "learning_rate": 3.4663401783785865e-06,
+      "loss": 0.0483,
+      "step": 2144
+    },
+    {
+      "epoch": 0.611633875106929,
+      "grad_norm": 1.445639177748986,
+      "learning_rate": 3.4619448393187126e-06,
+      "loss": 0.0155,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6119190191046479,
+      "grad_norm": 1.0315630583680009,
+      "learning_rate": 3.4575508126350875e-06,
+      "loss": 0.0126,
+      "step": 2146
+    },
+    {
+      "epoch": 0.6122041631023667,
+      "grad_norm": 0.7398009312525474,
+      "learning_rate": 3.453158102077001e-06,
+      "loss": 0.0229,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6124893071000855,
+      "grad_norm": 0.3557371653456857,
+      "learning_rate": 3.4487667113926226e-06,
+      "loss": 0.0115,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6127744510978044,
+      "grad_norm": 1.2762208256712575,
+      "learning_rate": 3.4443766443289948e-06,
+      "loss": 0.0246,
+      "step": 2149
+    },
+    {
+      "epoch": 0.6130595950955232,
+      "grad_norm": 1.3646703557027133,
+      "learning_rate": 3.439987904632026e-06,
+      "loss": 0.0311,
+      "step": 2150
+    },
+    {
+      "epoch": 0.613344739093242,
+      "grad_norm": 1.2229542617252571,
+      "learning_rate": 3.4356004960464994e-06,
+      "loss": 0.0227,
+      "step": 2151
+    },
+    {
+      "epoch": 0.6136298830909609,
+      "grad_norm": 2.3224264173489786,
+      "learning_rate": 3.431214422316057e-06,
+      "loss": 0.0362,
+      "step": 2152
+    },
+    {
+      "epoch": 0.6139150270886797,
+      "grad_norm": 0.709410518708107,
+      "learning_rate": 3.426829687183204e-06,
+      "loss": 0.0114,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6142001710863987,
+      "grad_norm": 1.0413949214912819,
+      "learning_rate": 3.4224462943893057e-06,
+      "loss": 0.0236,
+      "step": 2154
+    },
+    {
+      "epoch": 0.6144853150841175,
+      "grad_norm": 0.7296747342861363,
+      "learning_rate": 3.418064247674576e-06,
+      "loss": 0.0112,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6147704590818364,
+      "grad_norm": 1.2435332019082928,
+      "learning_rate": 3.413683550778084e-06,
+      "loss": 0.0158,
+      "step": 2156
+    },
+    {
+      "epoch": 0.6150556030795552,
+      "grad_norm": 1.0875315186112393,
+      "learning_rate": 3.409304207437749e-06,
+      "loss": 0.0194,
+      "step": 2157
+    },
+    {
+      "epoch": 0.615340747077274,
+      "grad_norm": 1.5282693808083203,
+      "learning_rate": 3.404926221390332e-06,
+      "loss": 0.0218,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6156258910749929,
+      "grad_norm": 0.474632274589884,
+      "learning_rate": 3.400549596371435e-06,
+      "loss": 0.0117,
+      "step": 2159
+    },
+    {
+      "epoch": 0.6159110350727117,
+      "grad_norm": 0.7581340345773641,
+      "learning_rate": 3.3961743361155055e-06,
+      "loss": 0.0197,
+      "step": 2160
+    },
+    {
+      "epoch": 0.6161961790704306,
+      "grad_norm": 1.3727372919022707,
+      "learning_rate": 3.3918004443558163e-06,
+      "loss": 0.0563,
+      "step": 2161
+    },
+    {
+      "epoch": 0.6164813230681494,
+      "grad_norm": 2.2648471754423496,
+      "learning_rate": 3.3874279248244803e-06,
+      "loss": 0.0327,
+      "step": 2162
+    },
+    {
+      "epoch": 0.6167664670658682,
+      "grad_norm": 0.7505272987405108,
+      "learning_rate": 3.383056781252435e-06,
+      "loss": 0.013,
+      "step": 2163
+    },
+    {
+      "epoch": 0.6170516110635871,
+      "grad_norm": 1.1079848809501045,
+      "learning_rate": 3.3786870173694497e-06,
+      "loss": 0.0127,
+      "step": 2164
+    },
+    {
+      "epoch": 0.6173367550613059,
+      "grad_norm": 2.0002129754495295,
+      "learning_rate": 3.37431863690411e-06,
+      "loss": 0.0347,
+      "step": 2165
+    },
+    {
+      "epoch": 0.6176218990590248,
+      "grad_norm": 1.5327965290014722,
+      "learning_rate": 3.369951643583823e-06,
+      "loss": 0.0307,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6179070430567437,
+      "grad_norm": 2.027574254434053,
+      "learning_rate": 3.365586041134815e-06,
+      "loss": 0.0292,
+      "step": 2167
+    },
+    {
+      "epoch": 0.6181921870544625,
+      "grad_norm": 1.1355125852210322,
+      "learning_rate": 3.361221833282122e-06,
+      "loss": 0.0182,
+      "step": 2168
+    },
+    {
+      "epoch": 0.6184773310521814,
+      "grad_norm": 1.0205244715021804,
+      "learning_rate": 3.3568590237495912e-06,
+      "loss": 0.0339,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6187624750499002,
+      "grad_norm": 0.9286220091766371,
+      "learning_rate": 3.3524976162598777e-06,
+      "loss": 0.0223,
+      "step": 2170
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 0.6108615934396132,
+      "learning_rate": 3.34813761453444e-06,
+      "loss": 0.0103,
+      "step": 2171
+    },
+    {
+      "epoch": 0.6193327630453379,
+      "grad_norm": 1.2988343838308507,
+      "learning_rate": 3.343779022293536e-06,
+      "loss": 0.0301,
+      "step": 2172
+    },
+    {
+      "epoch": 0.6196179070430567,
+      "grad_norm": 0.3341719225290523,
+      "learning_rate": 3.3394218432562185e-06,
+      "loss": 0.0082,
+      "step": 2173
+    },
+    {
+      "epoch": 0.6199030510407756,
+      "grad_norm": 1.6943619739632687,
+      "learning_rate": 3.3350660811403425e-06,
+      "loss": 0.0294,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6201881950384944,
+      "grad_norm": 1.1528385894622213,
+      "learning_rate": 3.330711739662545e-06,
+      "loss": 0.0187,
+      "step": 2175
+    },
+    {
+      "epoch": 0.6204733390362133,
+      "grad_norm": 1.4611973629167543,
+      "learning_rate": 3.326358822538258e-06,
+      "loss": 0.0402,
+      "step": 2176
+    },
+    {
+      "epoch": 0.6207584830339321,
+      "grad_norm": 1.7876364426505764,
+      "learning_rate": 3.3220073334816928e-06,
+      "loss": 0.0326,
+      "step": 2177
+    },
+    {
+      "epoch": 0.621043627031651,
+      "grad_norm": 0.7956538149927447,
+      "learning_rate": 3.3176572762058434e-06,
+      "loss": 0.0146,
+      "step": 2178
+    },
+    {
+      "epoch": 0.6213287710293698,
+      "grad_norm": 1.3226414882028097,
+      "learning_rate": 3.313308654422484e-06,
+      "loss": 0.0285,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6216139150270886,
+      "grad_norm": 2.1614175148958616,
+      "learning_rate": 3.3089614718421635e-06,
+      "loss": 0.0377,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6218990590248076,
+      "grad_norm": 0.9471207974194648,
+      "learning_rate": 3.304615732174201e-06,
+      "loss": 0.0145,
+      "step": 2181
+    },
+    {
+      "epoch": 0.6221842030225264,
+      "grad_norm": 1.3292435098511863,
+      "learning_rate": 3.300271439126689e-06,
+      "loss": 0.0234,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6224693470202453,
+      "grad_norm": 0.706611462258107,
+      "learning_rate": 3.2959285964064776e-06,
+      "loss": 0.0147,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6227544910179641,
+      "grad_norm": 3.000133668041504,
+      "learning_rate": 3.291587207719186e-06,
+      "loss": 0.0514,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6230396350156829,
+      "grad_norm": 2.2778332807998405,
+      "learning_rate": 3.2872472767691894e-06,
+      "loss": 0.0488,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6233247790134018,
+      "grad_norm": 1.1901143222547017,
+      "learning_rate": 3.2829088072596226e-06,
+      "loss": 0.0263,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6236099230111206,
+      "grad_norm": 0.8397072051957538,
+      "learning_rate": 3.2785718028923715e-06,
+      "loss": 0.0264,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6238950670088395,
+      "grad_norm": 0.38604583860305863,
+      "learning_rate": 3.2742362673680687e-06,
+      "loss": 0.0074,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6241802110065583,
+      "grad_norm": 1.947982342419954,
+      "learning_rate": 3.2699022043860973e-06,
+      "loss": 0.0267,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6244653550042771,
+      "grad_norm": 1.1626686035488216,
+      "learning_rate": 3.2655696176445852e-06,
+      "loss": 0.0229,
+      "step": 2190
+    },
+    {
+      "epoch": 0.624750499001996,
+      "grad_norm": 1.6043946457619445,
+      "learning_rate": 3.2612385108403955e-06,
+      "loss": 0.0382,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6250356429997148,
+      "grad_norm": 1.8055044074507538,
+      "learning_rate": 3.256908887669134e-06,
+      "loss": 0.0276,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6253207869974337,
+      "grad_norm": 1.6754046006759364,
+      "learning_rate": 3.2525807518251386e-06,
+      "loss": 0.0408,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6256059309951526,
+      "grad_norm": 0.9463034307019634,
+      "learning_rate": 3.248254107001474e-06,
+      "loss": 0.0175,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6258910749928714,
+      "grad_norm": 2.175470555795851,
+      "learning_rate": 3.243928956889938e-06,
+      "loss": 0.0278,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6261762189905903,
+      "grad_norm": 0.6729073385440583,
+      "learning_rate": 3.2396053051810515e-06,
+      "loss": 0.0285,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6264613629883091,
+      "grad_norm": 1.571784137465903,
+      "learning_rate": 3.2352831555640563e-06,
+      "loss": 0.0309,
+      "step": 2197
+    },
+    {
+      "epoch": 0.626746506986028,
+      "grad_norm": 1.1661769802553186,
+      "learning_rate": 3.230962511726915e-06,
+      "loss": 0.025,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6270316509837468,
+      "grad_norm": 0.42123213663257564,
+      "learning_rate": 3.2266433773563e-06,
+      "loss": 0.0115,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6273167949814656,
+      "grad_norm": 1.8338109807568292,
+      "learning_rate": 3.222325756137599e-06,
+      "loss": 0.0367,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6276019389791845,
+      "grad_norm": 1.1277887251821224,
+      "learning_rate": 3.21800965175491e-06,
+      "loss": 0.0311,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6278870829769033,
+      "grad_norm": 1.201306830776855,
+      "learning_rate": 3.213695067891034e-06,
+      "loss": 0.022,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6281722269746222,
+      "grad_norm": 1.2570922828383633,
+      "learning_rate": 3.2093820082274763e-06,
+      "loss": 0.0193,
+      "step": 2203
+    },
+    {
+      "epoch": 0.628457370972341,
+      "grad_norm": 1.5186675511953927,
+      "learning_rate": 3.2050704764444433e-06,
+      "loss": 0.0441,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6287425149700598,
+      "grad_norm": 0.6524174178550722,
+      "learning_rate": 3.200760476220831e-06,
+      "loss": 0.0268,
+      "step": 2205
+    },
+    {
+      "epoch": 0.6290276589677787,
+      "grad_norm": 0.8724476437379903,
+      "learning_rate": 3.1964520112342363e-06,
+      "loss": 0.0223,
+      "step": 2206
+    },
+    {
+      "epoch": 0.6293128029654976,
+      "grad_norm": 1.4186863998600603,
+      "learning_rate": 3.1921450851609404e-06,
+      "loss": 0.0392,
+      "step": 2207
+    },
+    {
+      "epoch": 0.6295979469632165,
+      "grad_norm": 1.336564669853662,
+      "learning_rate": 3.187839701675917e-06,
+      "loss": 0.0222,
+      "step": 2208
+    },
+    {
+      "epoch": 0.6298830909609353,
+      "grad_norm": 1.6468050869178028,
+      "learning_rate": 3.18353586445282e-06,
+      "loss": 0.037,
+      "step": 2209
+    },
+    {
+      "epoch": 0.6301682349586541,
+      "grad_norm": 1.841124429536788,
+      "learning_rate": 3.1792335771639827e-06,
+      "loss": 0.0394,
+      "step": 2210
+    },
+    {
+      "epoch": 0.630453378956373,
+      "grad_norm": 1.5844988548472232,
+      "learning_rate": 3.174932843480416e-06,
+      "loss": 0.0397,
+      "step": 2211
+    },
+    {
+      "epoch": 0.6307385229540918,
+      "grad_norm": 2.0069101566372085,
+      "learning_rate": 3.1706336670718106e-06,
+      "loss": 0.0501,
+      "step": 2212
+    },
+    {
+      "epoch": 0.6310236669518107,
+      "grad_norm": 0.9427410695473213,
+      "learning_rate": 3.166336051606521e-06,
+      "loss": 0.0231,
+      "step": 2213
+    },
+    {
+      "epoch": 0.6313088109495295,
+      "grad_norm": 0.9212663531499328,
+      "learning_rate": 3.1620400007515772e-06,
+      "loss": 0.0114,
+      "step": 2214
+    },
+    {
+      "epoch": 0.6315939549472483,
+      "grad_norm": 0.8808245598066381,
+      "learning_rate": 3.157745518172669e-06,
+      "loss": 0.0078,
+      "step": 2215
+    },
+    {
+      "epoch": 0.6318790989449672,
+      "grad_norm": 1.0702460227921857,
+      "learning_rate": 3.153452607534147e-06,
+      "loss": 0.02,
+      "step": 2216
+    },
+    {
+      "epoch": 0.632164242942686,
+      "grad_norm": 1.2534625169103417,
+      "learning_rate": 3.149161272499024e-06,
+      "loss": 0.0374,
+      "step": 2217
+    },
+    {
+      "epoch": 0.6324493869404049,
+      "grad_norm": 1.5261888629053173,
+      "learning_rate": 3.1448715167289677e-06,
+      "loss": 0.0435,
+      "step": 2218
+    },
+    {
+      "epoch": 0.6327345309381237,
+      "grad_norm": 1.2002510268158786,
+      "learning_rate": 3.140583343884298e-06,
+      "loss": 0.0183,
+      "step": 2219
+    },
+    {
+      "epoch": 0.6330196749358425,
+      "grad_norm": 1.0985692409600272,
+      "learning_rate": 3.1362967576239854e-06,
+      "loss": 0.0262,
+      "step": 2220
+    },
+    {
+      "epoch": 0.6333048189335615,
+      "grad_norm": 1.2072207242963418,
+      "learning_rate": 3.1320117616056413e-06,
+      "loss": 0.0314,
+      "step": 2221
+    },
+    {
+      "epoch": 0.6335899629312803,
+      "grad_norm": 0.467015398806509,
+      "learning_rate": 3.1277283594855267e-06,
+      "loss": 0.0106,
+      "step": 2222
+    },
+    {
+      "epoch": 0.6338751069289992,
+      "grad_norm": 0.8807024802610426,
+      "learning_rate": 3.123446554918538e-06,
+      "loss": 0.011,
+      "step": 2223
+    },
+    {
+      "epoch": 0.634160250926718,
+      "grad_norm": 0.6140299365612738,
+      "learning_rate": 3.1191663515582127e-06,
+      "loss": 0.0248,
+      "step": 2224
+    },
+    {
+      "epoch": 0.6344453949244369,
+      "grad_norm": 1.0559654324827494,
+      "learning_rate": 3.1148877530567177e-06,
+      "loss": 0.0183,
+      "step": 2225
+    },
+    {
+      "epoch": 0.6347305389221557,
+      "grad_norm": 1.0540351749712773,
+      "learning_rate": 3.1106107630648574e-06,
+      "loss": 0.0261,
+      "step": 2226
+    },
+    {
+      "epoch": 0.6350156829198745,
+      "grad_norm": 0.9313773695645087,
+      "learning_rate": 3.106335385232051e-06,
+      "loss": 0.0292,
+      "step": 2227
+    },
+    {
+      "epoch": 0.6353008269175934,
+      "grad_norm": 0.3747808238118342,
+      "learning_rate": 3.102061623206355e-06,
+      "loss": 0.0094,
+      "step": 2228
+    },
+    {
+      "epoch": 0.6355859709153122,
+      "grad_norm": 2.3312916926506504,
+      "learning_rate": 3.0977894806344406e-06,
+      "loss": 0.0527,
+      "step": 2229
+    },
+    {
+      "epoch": 0.635871114913031,
+      "grad_norm": 0.6324565323611943,
+      "learning_rate": 3.093518961161599e-06,
+      "loss": 0.0095,
+      "step": 2230
+    },
+    {
+      "epoch": 0.6361562589107499,
+      "grad_norm": 1.590518034839,
+      "learning_rate": 3.0892500684317386e-06,
+      "loss": 0.0244,
+      "step": 2231
+    },
+    {
+      "epoch": 0.6364414029084687,
+      "grad_norm": 0.7308543148733039,
+      "learning_rate": 3.084982806087372e-06,
+      "loss": 0.024,
+      "step": 2232
+    },
+    {
+      "epoch": 0.6367265469061876,
+      "grad_norm": 1.248369277983139,
+      "learning_rate": 3.080717177769629e-06,
+      "loss": 0.0231,
+      "step": 2233
+    },
+    {
+      "epoch": 0.6370116909039065,
+      "grad_norm": 0.9686441748388779,
+      "learning_rate": 3.0764531871182422e-06,
+      "loss": 0.016,
+      "step": 2234
+    },
+    {
+      "epoch": 0.6372968349016254,
+      "grad_norm": 1.0082773282212865,
+      "learning_rate": 3.072190837771546e-06,
+      "loss": 0.0175,
+      "step": 2235
+    },
+    {
+      "epoch": 0.6375819788993442,
+      "grad_norm": 1.4311529725243861,
+      "learning_rate": 3.067930133366476e-06,
+      "loss": 0.0319,
+      "step": 2236
+    },
+    {
+      "epoch": 0.637867122897063,
+      "grad_norm": 0.8406777760925859,
+      "learning_rate": 3.0636710775385635e-06,
+      "loss": 0.0108,
+      "step": 2237
+    },
+    {
+      "epoch": 0.6381522668947819,
+      "grad_norm": 0.878282487364186,
+      "learning_rate": 3.059413673921931e-06,
+      "loss": 0.0193,
+      "step": 2238
+    },
+    {
+      "epoch": 0.6384374108925007,
+      "grad_norm": 0.9502594240301524,
+      "learning_rate": 3.055157926149293e-06,
+      "loss": 0.0203,
+      "step": 2239
+    },
+    {
+      "epoch": 0.6387225548902196,
+      "grad_norm": 1.308186958668311,
+      "learning_rate": 3.050903837851953e-06,
+      "loss": 0.023,
+      "step": 2240
+    },
+    {
+      "epoch": 0.6390076988879384,
+      "grad_norm": 0.44428606044631475,
+      "learning_rate": 3.0466514126597945e-06,
+      "loss": 0.014,
+      "step": 2241
+    },
+    {
+      "epoch": 0.6392928428856572,
+      "grad_norm": 1.209699851720728,
+      "learning_rate": 3.0424006542012897e-06,
+      "loss": 0.0103,
+      "step": 2242
+    },
+    {
+      "epoch": 0.6395779868833761,
+      "grad_norm": 0.9514393878467423,
+      "learning_rate": 3.038151566103475e-06,
+      "loss": 0.0147,
+      "step": 2243
+    },
+    {
+      "epoch": 0.6398631308810949,
+      "grad_norm": 1.9454407250073225,
+      "learning_rate": 3.0339041519919745e-06,
+      "loss": 0.042,
+      "step": 2244
+    },
+    {
+      "epoch": 0.6401482748788138,
+      "grad_norm": 0.728850597077105,
+      "learning_rate": 3.029658415490977e-06,
+      "loss": 0.0124,
+      "step": 2245
+    },
+    {
+      "epoch": 0.6404334188765326,
+      "grad_norm": 1.1288659293589929,
+      "learning_rate": 3.0254143602232434e-06,
+      "loss": 0.0197,
+      "step": 2246
+    },
+    {
+      "epoch": 0.6407185628742516,
+      "grad_norm": 1.1475249422099956,
+      "learning_rate": 3.021171989810099e-06,
+      "loss": 0.0211,
+      "step": 2247
+    },
+    {
+      "epoch": 0.6410037068719704,
+      "grad_norm": 1.2132830355724133,
+      "learning_rate": 3.0169313078714296e-06,
+      "loss": 0.0293,
+      "step": 2248
+    },
+    {
+      "epoch": 0.6412888508696892,
+      "grad_norm": 1.9363856482086057,
+      "learning_rate": 3.0126923180256806e-06,
+      "loss": 0.0243,
+      "step": 2249
+    },
+    {
+      "epoch": 0.6415739948674081,
+      "grad_norm": 0.7703844388099039,
+      "learning_rate": 3.008455023889857e-06,
+      "loss": 0.0121,
+      "step": 2250
+    },
+    {
+      "epoch": 0.6418591388651269,
+      "grad_norm": 1.3134506657893854,
+      "learning_rate": 3.0042194290795123e-06,
+      "loss": 0.0318,
+      "step": 2251
+    },
+    {
+      "epoch": 0.6421442828628457,
+      "grad_norm": 1.4692545047359329,
+      "learning_rate": 2.999985537208755e-06,
+      "loss": 0.0176,
+      "step": 2252
+    },
+    {
+      "epoch": 0.6424294268605646,
+      "grad_norm": 0.9828572979175608,
+      "learning_rate": 2.9957533518902376e-06,
+      "loss": 0.0196,
+      "step": 2253
+    },
+    {
+      "epoch": 0.6427145708582834,
+      "grad_norm": 0.7049658553143684,
+      "learning_rate": 2.991522876735154e-06,
+      "loss": 0.0229,
+      "step": 2254
+    },
+    {
+      "epoch": 0.6429997148560023,
+      "grad_norm": 0.842198069759342,
+      "learning_rate": 2.987294115353242e-06,
+      "loss": 0.0128,
+      "step": 2255
+    },
+    {
+      "epoch": 0.6432848588537211,
+      "grad_norm": 1.7227804347359337,
+      "learning_rate": 2.9830670713527786e-06,
+      "loss": 0.0335,
+      "step": 2256
+    },
+    {
+      "epoch": 0.64357000285144,
+      "grad_norm": 0.3681229418460075,
+      "learning_rate": 2.9788417483405716e-06,
+      "loss": 0.0083,
+      "step": 2257
+    },
+    {
+      "epoch": 0.6438551468491588,
+      "grad_norm": 1.1818461840946088,
+      "learning_rate": 2.9746181499219627e-06,
+      "loss": 0.0154,
+      "step": 2258
+    },
+    {
+      "epoch": 0.6441402908468776,
+      "grad_norm": 1.411740982709478,
+      "learning_rate": 2.970396279700824e-06,
+      "loss": 0.0184,
+      "step": 2259
+    },
+    {
+      "epoch": 0.6444254348445965,
+      "grad_norm": 2.5844309546840716,
+      "learning_rate": 2.9661761412795465e-06,
+      "loss": 0.0378,
+      "step": 2260
+    },
+    {
+      "epoch": 0.6447105788423154,
+      "grad_norm": 1.5612725743924631,
+      "learning_rate": 2.9619577382590485e-06,
+      "loss": 0.0203,
+      "step": 2261
+    },
+    {
+      "epoch": 0.6449957228400343,
+      "grad_norm": 1.244411728256585,
+      "learning_rate": 2.9577410742387686e-06,
+      "loss": 0.0179,
+      "step": 2262
+    },
+    {
+      "epoch": 0.6452808668377531,
+      "grad_norm": 1.920069290541119,
+      "learning_rate": 2.9535261528166577e-06,
+      "loss": 0.0264,
+      "step": 2263
+    },
+    {
+      "epoch": 0.6455660108354719,
+      "grad_norm": 0.9562289133021995,
+      "learning_rate": 2.949312977589181e-06,
+      "loss": 0.0177,
+      "step": 2264
+    },
+    {
+      "epoch": 0.6458511548331908,
+      "grad_norm": 2.534580020676373,
+      "learning_rate": 2.945101552151317e-06,
+      "loss": 0.0305,
+      "step": 2265
+    },
+    {
+      "epoch": 0.6461362988309096,
+      "grad_norm": 1.3341099226810462,
+      "learning_rate": 2.9408918800965464e-06,
+      "loss": 0.0158,
+      "step": 2266
+    },
+    {
+      "epoch": 0.6464214428286285,
+      "grad_norm": 1.9949800798700283,
+      "learning_rate": 2.936683965016855e-06,
+      "loss": 0.0403,
+      "step": 2267
+    },
+    {
+      "epoch": 0.6467065868263473,
+      "grad_norm": 0.9844526919586044,
+      "learning_rate": 2.9324778105027323e-06,
+      "loss": 0.0137,
+      "step": 2268
+    },
+    {
+      "epoch": 0.6469917308240661,
+      "grad_norm": 1.0381005934303287,
+      "learning_rate": 2.9282734201431627e-06,
+      "loss": 0.0285,
+      "step": 2269
+    },
+    {
+      "epoch": 0.647276874821785,
+      "grad_norm": 1.2860946237813613,
+      "learning_rate": 2.924070797525628e-06,
+      "loss": 0.025,
+      "step": 2270
+    },
+    {
+      "epoch": 0.6475620188195038,
+      "grad_norm": 1.3340813057644438,
+      "learning_rate": 2.919869946236096e-06,
+      "loss": 0.0278,
+      "step": 2271
+    },
+    {
+      "epoch": 0.6478471628172227,
+      "grad_norm": 1.6573747957672138,
+      "learning_rate": 2.9156708698590273e-06,
+      "loss": 0.0301,
+      "step": 2272
+    },
+    {
+      "epoch": 0.6481323068149415,
+      "grad_norm": 0.868842113331832,
+      "learning_rate": 2.9114735719773718e-06,
+      "loss": 0.0206,
+      "step": 2273
+    },
+    {
+      "epoch": 0.6484174508126604,
+      "grad_norm": 0.7243820586100654,
+      "learning_rate": 2.9072780561725543e-06,
+      "loss": 0.0167,
+      "step": 2274
+    },
+    {
+      "epoch": 0.6487025948103793,
+      "grad_norm": 1.185198168923413,
+      "learning_rate": 2.9030843260244834e-06,
+      "loss": 0.023,
+      "step": 2275
+    },
+    {
+      "epoch": 0.6489877388080981,
+      "grad_norm": 0.6293907385655024,
+      "learning_rate": 2.8988923851115425e-06,
+      "loss": 0.0099,
+      "step": 2276
+    },
+    {
+      "epoch": 0.649272882805817,
+      "grad_norm": 2.0445266570057714,
+      "learning_rate": 2.894702237010589e-06,
+      "loss": 0.0186,
+      "step": 2277
+    },
+    {
+      "epoch": 0.6495580268035358,
+      "grad_norm": 2.047384698677883,
+      "learning_rate": 2.8905138852969507e-06,
+      "loss": 0.0268,
+      "step": 2278
+    },
+    {
+      "epoch": 0.6498431708012546,
+      "grad_norm": 1.0653517095398133,
+      "learning_rate": 2.886327333544421e-06,
+      "loss": 0.0161,
+      "step": 2279
+    },
+    {
+      "epoch": 0.6501283147989735,
+      "grad_norm": 1.5006183561890554,
+      "learning_rate": 2.8821425853252603e-06,
+      "loss": 0.035,
+      "step": 2280
+    },
+    {
+      "epoch": 0.6504134587966923,
+      "grad_norm": 0.937416176228414,
+      "learning_rate": 2.8779596442101878e-06,
+      "loss": 0.0229,
+      "step": 2281
+    },
+    {
+      "epoch": 0.6506986027944112,
+      "grad_norm": 1.4517724372844274,
+      "learning_rate": 2.8737785137683815e-06,
+      "loss": 0.027,
+      "step": 2282
+    },
+    {
+      "epoch": 0.65098374679213,
+      "grad_norm": 1.7278224774601127,
+      "learning_rate": 2.8695991975674735e-06,
+      "loss": 0.0237,
+      "step": 2283
+    },
+    {
+      "epoch": 0.6512688907898488,
+      "grad_norm": 1.5785313959183065,
+      "learning_rate": 2.8654216991735504e-06,
+      "loss": 0.0296,
+      "step": 2284
+    },
+    {
+      "epoch": 0.6515540347875677,
+      "grad_norm": 1.0919831115191292,
+      "learning_rate": 2.861246022151143e-06,
+      "loss": 0.02,
+      "step": 2285
+    },
+    {
+      "epoch": 0.6518391787852865,
+      "grad_norm": 0.7275158367500263,
+      "learning_rate": 2.8570721700632354e-06,
+      "loss": 0.0116,
+      "step": 2286
+    },
+    {
+      "epoch": 0.6521243227830055,
+      "grad_norm": 1.3129785911100587,
+      "learning_rate": 2.852900146471249e-06,
+      "loss": 0.0154,
+      "step": 2287
+    },
+    {
+      "epoch": 0.6524094667807243,
+      "grad_norm": 1.6226425506929618,
+      "learning_rate": 2.848729954935042e-06,
+      "loss": 0.019,
+      "step": 2288
+    },
+    {
+      "epoch": 0.6526946107784432,
+      "grad_norm": 1.7305587988312543,
+      "learning_rate": 2.844561599012918e-06,
+      "loss": 0.0267,
+      "step": 2289
+    },
+    {
+      "epoch": 0.652979754776162,
+      "grad_norm": 0.8143957160470554,
+      "learning_rate": 2.8403950822616088e-06,
+      "loss": 0.0085,
+      "step": 2290
+    },
+    {
+      "epoch": 0.6532648987738808,
+      "grad_norm": 1.1122531151983972,
+      "learning_rate": 2.836230408236278e-06,
+      "loss": 0.0215,
+      "step": 2291
+    },
+    {
+      "epoch": 0.6535500427715997,
+      "grad_norm": 0.4083282299767834,
+      "learning_rate": 2.832067580490516e-06,
+      "loss": 0.0063,
+      "step": 2292
+    },
+    {
+      "epoch": 0.6538351867693185,
+      "grad_norm": 0.5640507716536349,
+      "learning_rate": 2.827906602576339e-06,
+      "loss": 0.0087,
+      "step": 2293
+    },
+    {
+      "epoch": 0.6541203307670374,
+      "grad_norm": 1.3067961879293044,
+      "learning_rate": 2.823747478044185e-06,
+      "loss": 0.0138,
+      "step": 2294
+    },
+    {
+      "epoch": 0.6544054747647562,
+      "grad_norm": 1.0110478159071878,
+      "learning_rate": 2.8195902104429084e-06,
+      "loss": 0.0298,
+      "step": 2295
+    },
+    {
+      "epoch": 0.654690618762475,
+      "grad_norm": 0.5904570142107487,
+      "learning_rate": 2.815434803319783e-06,
+      "loss": 0.0111,
+      "step": 2296
+    },
+    {
+      "epoch": 0.6549757627601939,
+      "grad_norm": 0.245921457034098,
+      "learning_rate": 2.8112812602204885e-06,
+      "loss": 0.0026,
+      "step": 2297
+    },
+    {
+      "epoch": 0.6552609067579127,
+      "grad_norm": 0.7085877553818561,
+      "learning_rate": 2.8071295846891256e-06,
+      "loss": 0.0093,
+      "step": 2298
+    },
+    {
+      "epoch": 0.6555460507556315,
+      "grad_norm": 1.6452998075190501,
+      "learning_rate": 2.802979780268188e-06,
+      "loss": 0.0244,
+      "step": 2299
+    },
+    {
+      "epoch": 0.6558311947533504,
+      "grad_norm": 0.6904886542892218,
+      "learning_rate": 2.7988318504985817e-06,
+      "loss": 0.0174,
+      "step": 2300
+    },
+    {
+      "epoch": 0.6561163387510693,
+      "grad_norm": 0.5850968737884185,
+      "learning_rate": 2.7946857989196076e-06,
+      "loss": 0.0069,
+      "step": 2301
+    },
+    {
+      "epoch": 0.6564014827487882,
+      "grad_norm": 1.8185293469917627,
+      "learning_rate": 2.7905416290689717e-06,
+      "loss": 0.0412,
+      "step": 2302
+    },
+    {
+      "epoch": 0.656686626746507,
+      "grad_norm": 0.8965190611007764,
+      "learning_rate": 2.7863993444827697e-06,
+      "loss": 0.0226,
+      "step": 2303
+    },
+    {
+      "epoch": 0.6569717707442259,
+      "grad_norm": 1.3182053225055796,
+      "learning_rate": 2.782258948695481e-06,
+      "loss": 0.0175,
+      "step": 2304
+    },
+    {
+      "epoch": 0.6572569147419447,
+      "grad_norm": 0.9339402915101415,
+      "learning_rate": 2.778120445239989e-06,
+      "loss": 0.0122,
+      "step": 2305
+    },
+    {
+      "epoch": 0.6575420587396635,
+      "grad_norm": 1.6156977930324048,
+      "learning_rate": 2.773983837647551e-06,
+      "loss": 0.0389,
+      "step": 2306
+    },
+    {
+      "epoch": 0.6578272027373824,
+      "grad_norm": 0.6651074350777881,
+      "learning_rate": 2.76984912944781e-06,
+      "loss": 0.0146,
+      "step": 2307
+    },
+    {
+      "epoch": 0.6581123467351012,
+      "grad_norm": 1.8452127520754087,
+      "learning_rate": 2.765716324168789e-06,
+      "loss": 0.0227,
+      "step": 2308
+    },
+    {
+      "epoch": 0.6583974907328201,
+      "grad_norm": 0.9652739173619617,
+      "learning_rate": 2.761585425336886e-06,
+      "loss": 0.0267,
+      "step": 2309
+    },
+    {
+      "epoch": 0.6586826347305389,
+      "grad_norm": 1.2330026270001035,
+      "learning_rate": 2.757456436476873e-06,
+      "loss": 0.023,
+      "step": 2310
+    },
+    {
+      "epoch": 0.6589677787282577,
+      "grad_norm": 0.9565278016472557,
+      "learning_rate": 2.7533293611118923e-06,
+      "loss": 0.0131,
+      "step": 2311
+    },
+    {
+      "epoch": 0.6592529227259766,
+      "grad_norm": 0.5029893671434901,
+      "learning_rate": 2.7492042027634525e-06,
+      "loss": 0.0074,
+      "step": 2312
+    },
+    {
+      "epoch": 0.6595380667236954,
+      "grad_norm": 1.6048936560011449,
+      "learning_rate": 2.7450809649514265e-06,
+      "loss": 0.048,
+      "step": 2313
+    },
+    {
+      "epoch": 0.6598232107214144,
+      "grad_norm": 2.4275543233286503,
+      "learning_rate": 2.740959651194054e-06,
+      "loss": 0.0222,
+      "step": 2314
+    },
+    {
+      "epoch": 0.6601083547191332,
+      "grad_norm": 2.8531321009807247,
+      "learning_rate": 2.7368402650079228e-06,
+      "loss": 0.0597,
+      "step": 2315
+    },
+    {
+      "epoch": 0.660393498716852,
+      "grad_norm": 1.4630738141417063,
+      "learning_rate": 2.7327228099079826e-06,
+      "loss": 0.0229,
+      "step": 2316
+    },
+    {
+      "epoch": 0.6606786427145709,
+      "grad_norm": 0.6779278399548759,
+      "learning_rate": 2.728607289407534e-06,
+      "loss": 0.0069,
+      "step": 2317
+    },
+    {
+      "epoch": 0.6609637867122897,
+      "grad_norm": 2.398616876487329,
+      "learning_rate": 2.7244937070182286e-06,
+      "loss": 0.0297,
+      "step": 2318
+    },
+    {
+      "epoch": 0.6612489307100086,
+      "grad_norm": 0.7855959677784145,
+      "learning_rate": 2.7203820662500625e-06,
+      "loss": 0.0356,
+      "step": 2319
+    },
+    {
+      "epoch": 0.6615340747077274,
+      "grad_norm": 0.24261892957888342,
+      "learning_rate": 2.716272370611375e-06,
+      "loss": 0.004,
+      "step": 2320
+    },
+    {
+      "epoch": 0.6618192187054462,
+      "grad_norm": 1.5723455114774483,
+      "learning_rate": 2.712164623608844e-06,
+      "loss": 0.0285,
+      "step": 2321
+    },
+    {
+      "epoch": 0.6621043627031651,
+      "grad_norm": 0.24075125669855854,
+      "learning_rate": 2.7080588287474885e-06,
+      "loss": 0.0038,
+      "step": 2322
+    },
+    {
+      "epoch": 0.6623895067008839,
+      "grad_norm": 0.6828692169124314,
+      "learning_rate": 2.7039549895306593e-06,
+      "loss": 0.0154,
+      "step": 2323
+    },
+    {
+      "epoch": 0.6626746506986028,
+      "grad_norm": 0.6957630191781368,
+      "learning_rate": 2.699853109460039e-06,
+      "loss": 0.0097,
+      "step": 2324
+    },
+    {
+      "epoch": 0.6629597946963216,
+      "grad_norm": 2.9059174955559723,
+      "learning_rate": 2.695753192035639e-06,
+      "loss": 0.0604,
+      "step": 2325
+    },
+    {
+      "epoch": 0.6632449386940404,
+      "grad_norm": 1.1734601436082144,
+      "learning_rate": 2.691655240755795e-06,
+      "loss": 0.0181,
+      "step": 2326
+    },
+    {
+      "epoch": 0.6635300826917594,
+      "grad_norm": 1.3146469546781792,
+      "learning_rate": 2.6875592591171663e-06,
+      "loss": 0.0249,
+      "step": 2327
+    },
+    {
+      "epoch": 0.6638152266894782,
+      "grad_norm": 0.6224899882604887,
+      "learning_rate": 2.6834652506147297e-06,
+      "loss": 0.0094,
+      "step": 2328
+    },
+    {
+      "epoch": 0.6641003706871971,
+      "grad_norm": 0.8606657609714811,
+      "learning_rate": 2.67937321874178e-06,
+      "loss": 0.0117,
+      "step": 2329
+    },
+    {
+      "epoch": 0.6643855146849159,
+      "grad_norm": 1.237561338520917,
+      "learning_rate": 2.675283166989926e-06,
+      "loss": 0.0251,
+      "step": 2330
+    },
+    {
+      "epoch": 0.6646706586826348,
+      "grad_norm": 1.3971850707966957,
+      "learning_rate": 2.671195098849089e-06,
+      "loss": 0.0557,
+      "step": 2331
+    },
+    {
+      "epoch": 0.6649558026803536,
+      "grad_norm": 0.8052143234506958,
+      "learning_rate": 2.6671090178074878e-06,
+      "loss": 0.0147,
+      "step": 2332
+    },
+    {
+      "epoch": 0.6652409466780724,
+      "grad_norm": 1.6731504755642395,
+      "learning_rate": 2.663024927351655e-06,
+      "loss": 0.0187,
+      "step": 2333
+    },
+    {
+      "epoch": 0.6655260906757913,
+      "grad_norm": 3.2684569195727837,
+      "learning_rate": 2.658942830966425e-06,
+      "loss": 0.0474,
+      "step": 2334
+    },
+    {
+      "epoch": 0.6658112346735101,
+      "grad_norm": 1.6686161366860728,
+      "learning_rate": 2.654862732134926e-06,
+      "loss": 0.0957,
+      "step": 2335
+    },
+    {
+      "epoch": 0.666096378671229,
+      "grad_norm": 1.8905217359113058,
+      "learning_rate": 2.6507846343385862e-06,
+      "loss": 0.0332,
+      "step": 2336
+    },
+    {
+      "epoch": 0.6663815226689478,
+      "grad_norm": 0.8396385938218617,
+      "learning_rate": 2.6467085410571175e-06,
+      "loss": 0.0125,
+      "step": 2337
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.6943892112720279,
+      "learning_rate": 2.6426344557685342e-06,
+      "loss": 0.0151,
+      "step": 2338
+    },
+    {
+      "epoch": 0.6669518106643855,
+      "grad_norm": 0.8584969710658708,
+      "learning_rate": 2.6385623819491278e-06,
+      "loss": 0.0174,
+      "step": 2339
+    },
+    {
+      "epoch": 0.6672369546621043,
+      "grad_norm": 1.1323015265299097,
+      "learning_rate": 2.6344923230734786e-06,
+      "loss": 0.0344,
+      "step": 2340
+    },
+    {
+      "epoch": 0.6675220986598233,
+      "grad_norm": 1.2498573670424775,
+      "learning_rate": 2.630424282614441e-06,
+      "loss": 0.0258,
+      "step": 2341
+    },
+    {
+      "epoch": 0.6678072426575421,
+      "grad_norm": 1.6270286307278543,
+      "learning_rate": 2.6263582640431595e-06,
+      "loss": 0.0707,
+      "step": 2342
+    },
+    {
+      "epoch": 0.6680923866552609,
+      "grad_norm": 0.9402570992440751,
+      "learning_rate": 2.622294270829039e-06,
+      "loss": 0.0289,
+      "step": 2343
+    },
+    {
+      "epoch": 0.6683775306529798,
+      "grad_norm": 1.1740285977390892,
+      "learning_rate": 2.6182323064397645e-06,
+      "loss": 0.0236,
+      "step": 2344
+    },
+    {
+      "epoch": 0.6686626746506986,
+      "grad_norm": 2.7212202569279995,
+      "learning_rate": 2.614172374341287e-06,
+      "loss": 0.0568,
+      "step": 2345
+    },
+    {
+      "epoch": 0.6689478186484175,
+      "grad_norm": 0.6205323905446688,
+      "learning_rate": 2.610114477997827e-06,
+      "loss": 0.0109,
+      "step": 2346
+    },
+    {
+      "epoch": 0.6692329626461363,
+      "grad_norm": 2.0357402784016387,
+      "learning_rate": 2.6060586208718673e-06,
+      "loss": 0.0345,
+      "step": 2347
+    },
+    {
+      "epoch": 0.6695181066438551,
+      "grad_norm": 1.3978184198875205,
+      "learning_rate": 2.602004806424144e-06,
+      "loss": 0.0201,
+      "step": 2348
+    },
+    {
+      "epoch": 0.669803250641574,
+      "grad_norm": 0.8378057178472931,
+      "learning_rate": 2.597953038113655e-06,
+      "loss": 0.013,
+      "step": 2349
+    },
+    {
+      "epoch": 0.6700883946392928,
+      "grad_norm": 1.8429505668991712,
+      "learning_rate": 2.5939033193976567e-06,
+      "loss": 0.034,
+      "step": 2350
+    },
+    {
+      "epoch": 0.6703735386370117,
+      "grad_norm": 0.6951733957823362,
+      "learning_rate": 2.58985565373165e-06,
+      "loss": 0.0149,
+      "step": 2351
+    },
+    {
+      "epoch": 0.6706586826347305,
+      "grad_norm": 1.254638572408408,
+      "learning_rate": 2.585810044569387e-06,
+      "loss": 0.0198,
+      "step": 2352
+    },
+    {
+      "epoch": 0.6709438266324493,
+      "grad_norm": 1.2282829118276561,
+      "learning_rate": 2.581766495362864e-06,
+      "loss": 0.0153,
+      "step": 2353
+    },
+    {
+      "epoch": 0.6712289706301683,
+      "grad_norm": 1.8682120897475893,
+      "learning_rate": 2.57772500956232e-06,
+      "loss": 0.0286,
+      "step": 2354
+    },
+    {
+      "epoch": 0.6715141146278871,
+      "grad_norm": 0.9128992862378721,
+      "learning_rate": 2.573685590616234e-06,
+      "loss": 0.0178,
+      "step": 2355
+    },
+    {
+      "epoch": 0.671799258625606,
+      "grad_norm": 1.556527433042271,
+      "learning_rate": 2.5696482419713197e-06,
+      "loss": 0.0441,
+      "step": 2356
+    },
+    {
+      "epoch": 0.6720844026233248,
+      "grad_norm": 0.2516116743394582,
+      "learning_rate": 2.565612967072525e-06,
+      "loss": 0.0058,
+      "step": 2357
+    },
+    {
+      "epoch": 0.6723695466210436,
+      "grad_norm": 1.191372773291596,
+      "learning_rate": 2.5615797693630328e-06,
+      "loss": 0.0368,
+      "step": 2358
+    },
+    {
+      "epoch": 0.6726546906187625,
+      "grad_norm": 1.091184520957333,
+      "learning_rate": 2.557548652284245e-06,
+      "loss": 0.0259,
+      "step": 2359
+    },
+    {
+      "epoch": 0.6729398346164813,
+      "grad_norm": 1.0730756075730936,
+      "learning_rate": 2.553519619275794e-06,
+      "loss": 0.0185,
+      "step": 2360
+    },
+    {
+      "epoch": 0.6732249786142002,
+      "grad_norm": 1.8119984458090785,
+      "learning_rate": 2.5494926737755306e-06,
+      "loss": 0.0379,
+      "step": 2361
+    },
+    {
+      "epoch": 0.673510122611919,
+      "grad_norm": 0.5991375226390366,
+      "learning_rate": 2.5454678192195305e-06,
+      "loss": 0.0154,
+      "step": 2362
+    },
+    {
+      "epoch": 0.6737952666096378,
+      "grad_norm": 0.8446181744229467,
+      "learning_rate": 2.5414450590420793e-06,
+      "loss": 0.016,
+      "step": 2363
+    },
+    {
+      "epoch": 0.6740804106073567,
+      "grad_norm": 1.2506916116716331,
+      "learning_rate": 2.5374243966756782e-06,
+      "loss": 0.0179,
+      "step": 2364
+    },
+    {
+      "epoch": 0.6743655546050755,
+      "grad_norm": 0.7844351247089717,
+      "learning_rate": 2.5334058355510337e-06,
+      "loss": 0.0142,
+      "step": 2365
+    },
+    {
+      "epoch": 0.6746506986027944,
+      "grad_norm": 1.2753595775969064,
+      "learning_rate": 2.529389379097067e-06,
+      "loss": 0.0584,
+      "step": 2366
+    },
+    {
+      "epoch": 0.6749358426005133,
+      "grad_norm": 2.130096709691072,
+      "learning_rate": 2.5253750307408996e-06,
+      "loss": 0.0433,
+      "step": 2367
+    },
+    {
+      "epoch": 0.6752209865982322,
+      "grad_norm": 2.4889057844910223,
+      "learning_rate": 2.5213627939078534e-06,
+      "loss": 0.0529,
+      "step": 2368
+    },
+    {
+      "epoch": 0.675506130595951,
+      "grad_norm": 2.0733191308785397,
+      "learning_rate": 2.517352672021449e-06,
+      "loss": 0.0735,
+      "step": 2369
+    },
+    {
+      "epoch": 0.6757912745936698,
+      "grad_norm": 0.8721844668658577,
+      "learning_rate": 2.5133446685034048e-06,
+      "loss": 0.02,
+      "step": 2370
+    },
+    {
+      "epoch": 0.6760764185913887,
+      "grad_norm": 0.5662408732865061,
+      "learning_rate": 2.5093387867736275e-06,
+      "loss": 0.0185,
+      "step": 2371
+    },
+    {
+      "epoch": 0.6763615625891075,
+      "grad_norm": 0.7771568621772995,
+      "learning_rate": 2.5053350302502167e-06,
+      "loss": 0.016,
+      "step": 2372
+    },
+    {
+      "epoch": 0.6766467065868264,
+      "grad_norm": 1.699224004876367,
+      "learning_rate": 2.5013334023494556e-06,
+      "loss": 0.0308,
+      "step": 2373
+    },
+    {
+      "epoch": 0.6769318505845452,
+      "grad_norm": 0.6987391064530609,
+      "learning_rate": 2.4973339064858163e-06,
+      "loss": 0.0159,
+      "step": 2374
+    },
+    {
+      "epoch": 0.677216994582264,
+      "grad_norm": 1.6646679512778157,
+      "learning_rate": 2.49333654607195e-06,
+      "loss": 0.0236,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6775021385799829,
+      "grad_norm": 1.6455502915346387,
+      "learning_rate": 2.489341324518678e-06,
+      "loss": 0.0621,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6777872825777017,
+      "grad_norm": 0.7215796317053166,
+      "learning_rate": 2.4853482452350048e-06,
+      "loss": 0.0194,
+      "step": 2377
+    },
+    {
+      "epoch": 0.6780724265754206,
+      "grad_norm": 0.9265039963157378,
+      "learning_rate": 2.4813573116281083e-06,
+      "loss": 0.0287,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6783575705731394,
+      "grad_norm": 1.386647458592848,
+      "learning_rate": 2.477368527103329e-06,
+      "loss": 0.0276,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6786427145708582,
+      "grad_norm": 1.0194721860507407,
+      "learning_rate": 2.473381895064181e-06,
+      "loss": 0.0569,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6789278585685772,
+      "grad_norm": 1.5781901137615435,
+      "learning_rate": 2.46939741891233e-06,
+      "loss": 0.0219,
+      "step": 2381
+    },
+    {
+      "epoch": 0.679213002566296,
+      "grad_norm": 0.9271661128511948,
+      "learning_rate": 2.4654151020476165e-06,
+      "loss": 0.0154,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6794981465640149,
+      "grad_norm": 1.2929840875659686,
+      "learning_rate": 2.461434947868029e-06,
+      "loss": 0.0336,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6797832905617337,
+      "grad_norm": 0.8356221756560887,
+      "learning_rate": 2.4574569597697145e-06,
+      "loss": 0.0202,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6800684345594525,
+      "grad_norm": 0.664795804036026,
+      "learning_rate": 2.4534811411469704e-06,
+      "loss": 0.0104,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6803535785571714,
+      "grad_norm": 1.9929916630870672,
+      "learning_rate": 2.4495074953922425e-06,
+      "loss": 0.0277,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6806387225548902,
+      "grad_norm": 1.5262298229248654,
+      "learning_rate": 2.4455360258961247e-06,
+      "loss": 0.0325,
+      "step": 2387
+    },
+    {
+      "epoch": 0.6809238665526091,
+      "grad_norm": 0.7087902063036329,
+      "learning_rate": 2.4415667360473518e-06,
+      "loss": 0.0165,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6812090105503279,
+      "grad_norm": 0.7561419087572873,
+      "learning_rate": 2.437599629232797e-06,
+      "loss": 0.011,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6814941545480467,
+      "grad_norm": 1.045806215953938,
+      "learning_rate": 2.433634708837478e-06,
+      "loss": 0.0236,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6817792985457656,
+      "grad_norm": 1.3375850382967915,
+      "learning_rate": 2.4296719782445422e-06,
+      "loss": 0.0322,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6820644425434844,
+      "grad_norm": 1.442188021651849,
+      "learning_rate": 2.4257114408352646e-06,
+      "loss": 0.0382,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6823495865412033,
+      "grad_norm": 1.4811177425124444,
+      "learning_rate": 2.421753099989052e-06,
+      "loss": 0.0399,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6826347305389222,
+      "grad_norm": 1.2151409515623008,
+      "learning_rate": 2.4177969590834425e-06,
+      "loss": 0.0344,
+      "step": 2394
+    },
+    {
+      "epoch": 0.682919874536641,
+      "grad_norm": 2.7056727418479958,
+      "learning_rate": 2.4138430214940906e-06,
+      "loss": 0.0551,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6832050185343599,
+      "grad_norm": 0.5969770040702388,
+      "learning_rate": 2.4098912905947712e-06,
+      "loss": 0.015,
+      "step": 2396
+    },
+    {
+      "epoch": 0.6834901625320787,
+      "grad_norm": 1.4726176220224632,
+      "learning_rate": 2.4059417697573782e-06,
+      "loss": 0.0191,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6837753065297976,
+      "grad_norm": 0.6985703906149163,
+      "learning_rate": 2.4019944623519194e-06,
+      "loss": 0.0217,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6840604505275164,
+      "grad_norm": 1.3556092377104045,
+      "learning_rate": 2.3980493717465124e-06,
+      "loss": 0.0314,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6843455945252352,
+      "grad_norm": 1.6320428555492388,
+      "learning_rate": 2.394106501307386e-06,
+      "loss": 0.0406,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6846307385229541,
+      "grad_norm": 0.3152755510356312,
+      "learning_rate": 2.390165854398872e-06,
+      "loss": 0.0076,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6849158825206729,
+      "grad_norm": 0.5496493802305953,
+      "learning_rate": 2.386227434383407e-06,
+      "loss": 0.0095,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6852010265183918,
+      "grad_norm": 0.6520718654651182,
+      "learning_rate": 2.3822912446215244e-06,
+      "loss": 0.0179,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6854861705161106,
+      "grad_norm": 1.2473935565493588,
+      "learning_rate": 2.3783572884718592e-06,
+      "loss": 0.0319,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6857713145138294,
+      "grad_norm": 1.0535903429708244,
+      "learning_rate": 2.3744255692911345e-06,
+      "loss": 0.015,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6860564585115483,
+      "grad_norm": 1.153321160939757,
+      "learning_rate": 2.370496090434172e-06,
+      "loss": 0.0243,
+      "step": 2406
+    },
+    {
+      "epoch": 0.6863416025092672,
+      "grad_norm": 0.8326855476595857,
+      "learning_rate": 2.3665688552538767e-06,
+      "loss": 0.016,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6866267465069861,
+      "grad_norm": 1.4628607016553867,
+      "learning_rate": 2.3626438671012412e-06,
+      "loss": 0.0411,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6869118905047049,
+      "grad_norm": 0.5985342385584901,
+      "learning_rate": 2.358721129325336e-06,
+      "loss": 0.0153,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6871970345024238,
+      "grad_norm": 0.8315001817874339,
+      "learning_rate": 2.354800645273319e-06,
+      "loss": 0.0314,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6874821785001426,
+      "grad_norm": 0.6885528098745334,
+      "learning_rate": 2.3508824182904207e-06,
+      "loss": 0.0124,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6877673224978614,
+      "grad_norm": 1.9141392122493273,
+      "learning_rate": 2.3469664517199463e-06,
+      "loss": 0.0354,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6880524664955803,
+      "grad_norm": 1.1665319631123647,
+      "learning_rate": 2.3430527489032723e-06,
+      "loss": 0.0215,
+      "step": 2413
+    },
+    {
+      "epoch": 0.6883376104932991,
+      "grad_norm": 1.0810004667670476,
+      "learning_rate": 2.339141313179844e-06,
+      "loss": 0.0321,
+      "step": 2414
+    },
+    {
+      "epoch": 0.688622754491018,
+      "grad_norm": 0.8222383468643456,
+      "learning_rate": 2.3352321478871717e-06,
+      "loss": 0.023,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6889078984887368,
+      "grad_norm": 1.1477329288258749,
+      "learning_rate": 2.331325256360828e-06,
+      "loss": 0.0147,
+      "step": 2416
+    },
+    {
+      "epoch": 0.6891930424864556,
+      "grad_norm": 0.4781898959973014,
+      "learning_rate": 2.327420641934447e-06,
+      "loss": 0.0059,
+      "step": 2417
+    },
+    {
+      "epoch": 0.6894781864841745,
+      "grad_norm": 1.1660744412330555,
+      "learning_rate": 2.323518307939717e-06,
+      "loss": 0.0294,
+      "step": 2418
+    },
+    {
+      "epoch": 0.6897633304818933,
+      "grad_norm": 0.5872242883588114,
+      "learning_rate": 2.3196182577063868e-06,
+      "loss": 0.0121,
+      "step": 2419
+    },
+    {
+      "epoch": 0.6900484744796122,
+      "grad_norm": 2.212762462397327,
+      "learning_rate": 2.315720494562248e-06,
+      "loss": 0.0324,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6903336184773311,
+      "grad_norm": 1.6373711381327187,
+      "learning_rate": 2.3118250218331463e-06,
+      "loss": 0.0185,
+      "step": 2421
+    },
+    {
+      "epoch": 0.6906187624750499,
+      "grad_norm": 0.7770141753292534,
+      "learning_rate": 2.307931842842968e-06,
+      "loss": 0.0299,
+      "step": 2422
+    },
+    {
+      "epoch": 0.6909039064727688,
+      "grad_norm": 1.518201141107165,
+      "learning_rate": 2.3040409609136515e-06,
+      "loss": 0.0463,
+      "step": 2423
+    },
+    {
+      "epoch": 0.6911890504704876,
+      "grad_norm": 0.43241598666358966,
+      "learning_rate": 2.3001523793651688e-06,
+      "loss": 0.0124,
+      "step": 2424
+    },
+    {
+      "epoch": 0.6914741944682065,
+      "grad_norm": 1.0771130938316928,
+      "learning_rate": 2.2962661015155234e-06,
+      "loss": 0.0561,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6917593384659253,
+      "grad_norm": 0.9154394384341286,
+      "learning_rate": 2.292382130680766e-06,
+      "loss": 0.0182,
+      "step": 2426
+    },
+    {
+      "epoch": 0.6920444824636441,
+      "grad_norm": 0.9492377155776165,
+      "learning_rate": 2.2885004701749695e-06,
+      "loss": 0.0181,
+      "step": 2427
+    },
+    {
+      "epoch": 0.692329626461363,
+      "grad_norm": 1.1013460236131432,
+      "learning_rate": 2.2846211233102387e-06,
+      "loss": 0.03,
+      "step": 2428
+    },
+    {
+      "epoch": 0.6926147704590818,
+      "grad_norm": 1.1127112319248678,
+      "learning_rate": 2.2807440933967034e-06,
+      "loss": 0.021,
+      "step": 2429
+    },
+    {
+      "epoch": 0.6928999144568007,
+      "grad_norm": 0.9531578556692747,
+      "learning_rate": 2.276869383742517e-06,
+      "loss": 0.0149,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6931850584545195,
+      "grad_norm": 1.0679032804811803,
+      "learning_rate": 2.2729969976538524e-06,
+      "loss": 0.0187,
+      "step": 2431
+    },
+    {
+      "epoch": 0.6934702024522383,
+      "grad_norm": 2.528890187837627,
+      "learning_rate": 2.2691269384349007e-06,
+      "loss": 0.0397,
+      "step": 2432
+    },
+    {
+      "epoch": 0.6937553464499572,
+      "grad_norm": 1.366923157548722,
+      "learning_rate": 2.265259209387867e-06,
+      "loss": 0.0273,
+      "step": 2433
+    },
+    {
+      "epoch": 0.6940404904476761,
+      "grad_norm": 0.6850783877005101,
+      "learning_rate": 2.261393813812966e-06,
+      "loss": 0.015,
+      "step": 2434
+    },
+    {
+      "epoch": 0.694325634445395,
+      "grad_norm": 1.0072621570466294,
+      "learning_rate": 2.2575307550084295e-06,
+      "loss": 0.0234,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6946107784431138,
+      "grad_norm": 1.3465168819525803,
+      "learning_rate": 2.2536700362704846e-06,
+      "loss": 0.0213,
+      "step": 2436
+    },
+    {
+      "epoch": 0.6948959224408326,
+      "grad_norm": 1.0391955609835548,
+      "learning_rate": 2.2498116608933673e-06,
+      "loss": 0.0157,
+      "step": 2437
+    },
+    {
+      "epoch": 0.6951810664385515,
+      "grad_norm": 2.814771203840875,
+      "learning_rate": 2.2459556321693123e-06,
+      "loss": 0.0552,
+      "step": 2438
+    },
+    {
+      "epoch": 0.6954662104362703,
+      "grad_norm": 0.6121746444624311,
+      "learning_rate": 2.242101953388556e-06,
+      "loss": 0.0112,
+      "step": 2439
+    },
+    {
+      "epoch": 0.6957513544339892,
+      "grad_norm": 1.3291384876570984,
+      "learning_rate": 2.238250627839325e-06,
+      "loss": 0.0131,
+      "step": 2440
+    },
+    {
+      "epoch": 0.696036498431708,
+      "grad_norm": 1.1148763168142317,
+      "learning_rate": 2.2344016588078403e-06,
+      "loss": 0.0354,
+      "step": 2441
+    },
+    {
+      "epoch": 0.6963216424294268,
+      "grad_norm": 1.36958013088009,
+      "learning_rate": 2.230555049578312e-06,
+      "loss": 0.029,
+      "step": 2442
+    },
+    {
+      "epoch": 0.6966067864271457,
+      "grad_norm": 0.7790771298504943,
+      "learning_rate": 2.2267108034329343e-06,
+      "loss": 0.0127,
+      "step": 2443
+    },
+    {
+      "epoch": 0.6968919304248645,
+      "grad_norm": 1.1333344045819018,
+      "learning_rate": 2.222868923651888e-06,
+      "loss": 0.0225,
+      "step": 2444
+    },
+    {
+      "epoch": 0.6971770744225834,
+      "grad_norm": 1.2892580769517954,
+      "learning_rate": 2.2190294135133334e-06,
+      "loss": 0.016,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6974622184203022,
+      "grad_norm": 2.2596154885188304,
+      "learning_rate": 2.2151922762934096e-06,
+      "loss": 0.0359,
+      "step": 2446
+    },
+    {
+      "epoch": 0.6977473624180212,
+      "grad_norm": 1.5204543539753177,
+      "learning_rate": 2.2113575152662304e-06,
+      "loss": 0.0225,
+      "step": 2447
+    },
+    {
+      "epoch": 0.69803250641574,
+      "grad_norm": 2.129816314096946,
+      "learning_rate": 2.207525133703881e-06,
+      "loss": 0.0312,
+      "step": 2448
+    },
+    {
+      "epoch": 0.6983176504134588,
+      "grad_norm": 0.8726329016607753,
+      "learning_rate": 2.203695134876419e-06,
+      "loss": 0.0129,
+      "step": 2449
+    },
+    {
+      "epoch": 0.6986027944111777,
+      "grad_norm": 0.981509542326578,
+      "learning_rate": 2.199867522051865e-06,
+      "loss": 0.0306,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6988879384088965,
+      "grad_norm": 1.2233864788734838,
+      "learning_rate": 2.1960422984962094e-06,
+      "loss": 0.0457,
+      "step": 2451
+    },
+    {
+      "epoch": 0.6991730824066154,
+      "grad_norm": 1.121663408796296,
+      "learning_rate": 2.1922194674734003e-06,
+      "loss": 0.0278,
+      "step": 2452
+    },
+    {
+      "epoch": 0.6994582264043342,
+      "grad_norm": 1.612742447635186,
+      "learning_rate": 2.1883990322453414e-06,
+      "loss": 0.0537,
+      "step": 2453
+    },
+    {
+      "epoch": 0.699743370402053,
+      "grad_norm": 0.9032206069159022,
+      "learning_rate": 2.184580996071895e-06,
+      "loss": 0.0382,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7000285143997719,
+      "grad_norm": 1.3020251741485003,
+      "learning_rate": 2.1807653622108797e-06,
+      "loss": 0.0511,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7003136583974907,
+      "grad_norm": 0.927814184904209,
+      "learning_rate": 2.1769521339180604e-06,
+      "loss": 0.0291,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7005988023952096,
+      "grad_norm": 1.4080121788852615,
+      "learning_rate": 2.17314131444715e-06,
+      "loss": 0.0168,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7008839463929284,
+      "grad_norm": 1.9115355677255668,
+      "learning_rate": 2.1693329070498057e-06,
+      "loss": 0.0243,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7011690903906472,
+      "grad_norm": 0.6119517264756272,
+      "learning_rate": 2.165526914975628e-06,
+      "loss": 0.0128,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7014542343883661,
+      "grad_norm": 1.2212817821331179,
+      "learning_rate": 2.1617233414721546e-06,
+      "loss": 0.0124,
+      "step": 2460
+    },
+    {
+      "epoch": 0.701739378386085,
+      "grad_norm": 1.0456467000073637,
+      "learning_rate": 2.1579221897848608e-06,
+      "loss": 0.0225,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7020245223838039,
+      "grad_norm": 1.9001029978437518,
+      "learning_rate": 2.1541234631571533e-06,
+      "loss": 0.0347,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7023096663815227,
+      "grad_norm": 1.1170803439317123,
+      "learning_rate": 2.1503271648303776e-06,
+      "loss": 0.0158,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7025948103792415,
+      "grad_norm": 1.2513944489653663,
+      "learning_rate": 2.1465332980437937e-06,
+      "loss": 0.0185,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7028799543769604,
+      "grad_norm": 0.9513327072225457,
+      "learning_rate": 2.1427418660345978e-06,
+      "loss": 0.0162,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7031650983746792,
+      "grad_norm": 1.0188208991697878,
+      "learning_rate": 2.138952872037902e-06,
+      "loss": 0.044,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7034502423723981,
+      "grad_norm": 1.5772768139954325,
+      "learning_rate": 2.135166319286745e-06,
+      "loss": 0.05,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7037353863701169,
+      "grad_norm": 1.8088612980621237,
+      "learning_rate": 2.1313822110120787e-06,
+      "loss": 0.0434,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7040205303678357,
+      "grad_norm": 0.589663362629804,
+      "learning_rate": 2.1276005504427643e-06,
+      "loss": 0.0089,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7043056743655546,
+      "grad_norm": 1.0455626616060543,
+      "learning_rate": 2.1238213408055806e-06,
+      "loss": 0.0252,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7045908183632734,
+      "grad_norm": 1.2419091857004885,
+      "learning_rate": 2.1200445853252165e-06,
+      "loss": 0.0313,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7048759623609923,
+      "grad_norm": 1.309158885132024,
+      "learning_rate": 2.116270287224262e-06,
+      "loss": 0.016,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7051611063587111,
+      "grad_norm": 2.46200008191099,
+      "learning_rate": 2.1124984497232127e-06,
+      "loss": 0.0427,
+      "step": 2473
+    },
+    {
+      "epoch": 0.70544625035643,
+      "grad_norm": 1.1686060450283622,
+      "learning_rate": 2.1087290760404634e-06,
+      "loss": 0.0226,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7057313943541489,
+      "grad_norm": 1.4653606459109394,
+      "learning_rate": 2.1049621693923084e-06,
+      "loss": 0.0413,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7060165383518677,
+      "grad_norm": 0.7719168267016131,
+      "learning_rate": 2.101197732992935e-06,
+      "loss": 0.0205,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7063016823495866,
+      "grad_norm": 0.4502961937417831,
+      "learning_rate": 2.0974357700544244e-06,
+      "loss": 0.0117,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7065868263473054,
+      "grad_norm": 1.385954489989293,
+      "learning_rate": 2.0936762837867445e-06,
+      "loss": 0.0242,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7068719703450242,
+      "grad_norm": 2.1799652784336216,
+      "learning_rate": 2.0899192773977574e-06,
+      "loss": 0.0569,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7071571143427431,
+      "grad_norm": 1.1110636882709009,
+      "learning_rate": 2.086164754093198e-06,
+      "loss": 0.0316,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7074422583404619,
+      "grad_norm": 1.4941307442971459,
+      "learning_rate": 2.0824127170766904e-06,
+      "loss": 0.0239,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7077274023381808,
+      "grad_norm": 0.6801030858403582,
+      "learning_rate": 2.0786631695497335e-06,
+      "loss": 0.0135,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7080125463358996,
+      "grad_norm": 1.1364604361529025,
+      "learning_rate": 2.074916114711706e-06,
+      "loss": 0.0166,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7082976903336184,
+      "grad_norm": 1.2388624592631503,
+      "learning_rate": 2.071171555759856e-06,
+      "loss": 0.02,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7085828343313373,
+      "grad_norm": 0.9231133048338788,
+      "learning_rate": 2.0674294958893052e-06,
+      "loss": 0.0132,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7088679783290561,
+      "grad_norm": 1.0122357410820453,
+      "learning_rate": 2.0636899382930357e-06,
+      "loss": 0.0195,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7091531223267751,
+      "grad_norm": 1.0915144391953777,
+      "learning_rate": 2.0599528861619046e-06,
+      "loss": 0.0218,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7094382663244939,
+      "grad_norm": 0.5781606335754661,
+      "learning_rate": 2.056218342684624e-06,
+      "loss": 0.0138,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7097234103222128,
+      "grad_norm": 1.2034483165455416,
+      "learning_rate": 2.0524863110477683e-06,
+      "loss": 0.0208,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7100085543199316,
+      "grad_norm": 1.0223652929411267,
+      "learning_rate": 2.0487567944357658e-06,
+      "loss": 0.0177,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7102936983176504,
+      "grad_norm": 0.7589363858158285,
+      "learning_rate": 2.0450297960309057e-06,
+      "loss": 0.0191,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7105788423153693,
+      "grad_norm": 1.7624288774750982,
+      "learning_rate": 2.0413053190133198e-06,
+      "loss": 0.0462,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7108639863130881,
+      "grad_norm": 0.9123674547573875,
+      "learning_rate": 2.0375833665609927e-06,
+      "loss": 0.0349,
+      "step": 2493
+    },
+    {
+      "epoch": 0.711149130310807,
+      "grad_norm": 1.1539174639109313,
+      "learning_rate": 2.033863941849754e-06,
+      "loss": 0.0383,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7114342743085258,
+      "grad_norm": 0.9661273666986842,
+      "learning_rate": 2.0301470480532803e-06,
+      "loss": 0.0296,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7117194183062446,
+      "grad_norm": 1.101711554212656,
+      "learning_rate": 2.026432688343085e-06,
+      "loss": 0.0322,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7120045623039635,
+      "grad_norm": 1.700057021557032,
+      "learning_rate": 2.0227208658885167e-06,
+      "loss": 0.0282,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7122897063016823,
+      "grad_norm": 0.6125316137983147,
+      "learning_rate": 2.019011583856761e-06,
+      "loss": 0.009,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7125748502994012,
+      "grad_norm": 2.1882827226129273,
+      "learning_rate": 2.015304845412841e-06,
+      "loss": 0.0422,
+      "step": 2499
+    },
+    {
+      "epoch": 0.71285999429712,
+      "grad_norm": 0.9211104223469543,
+      "learning_rate": 2.0116006537196033e-06,
+      "loss": 0.0327,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7131451382948389,
+      "grad_norm": 1.0694540209104813,
+      "learning_rate": 2.0078990119377233e-06,
+      "loss": 0.0279,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7134302822925578,
+      "grad_norm": 1.388175054161624,
+      "learning_rate": 2.004199923225701e-06,
+      "loss": 0.0259,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7137154262902766,
+      "grad_norm": 1.0988981162449094,
+      "learning_rate": 2.0005033907398574e-06,
+      "loss": 0.0212,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7140005702879955,
+      "grad_norm": 1.1482399987583405,
+      "learning_rate": 1.9968094176343322e-06,
+      "loss": 0.0281,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.5501580823692963,
+      "learning_rate": 1.9931180070610823e-06,
+      "loss": 0.012,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7145708582834331,
+      "grad_norm": 1.2633694375174702,
+      "learning_rate": 1.989429162169876e-06,
+      "loss": 0.0304,
+      "step": 2506
+    },
+    {
+      "epoch": 0.714856002281152,
+      "grad_norm": 0.6924873154457065,
+      "learning_rate": 1.9857428861082976e-06,
+      "loss": 0.0141,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7151411462788708,
+      "grad_norm": 0.6065820485902311,
+      "learning_rate": 1.9820591820217315e-06,
+      "loss": 0.0207,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7154262902765897,
+      "grad_norm": 0.7765390907982586,
+      "learning_rate": 1.978378053053373e-06,
+      "loss": 0.0139,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7157114342743085,
+      "grad_norm": 2.0988006081981316,
+      "learning_rate": 1.9746995023442177e-06,
+      "loss": 0.0231,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7159965782720273,
+      "grad_norm": 1.0519010093126555,
+      "learning_rate": 1.9710235330330656e-06,
+      "loss": 0.0181,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7162817222697462,
+      "grad_norm": 1.1614853715616795,
+      "learning_rate": 1.9673501482565083e-06,
+      "loss": 0.0341,
+      "step": 2512
+    },
+    {
+      "epoch": 0.716566866267465,
+      "grad_norm": 0.8821473934969634,
+      "learning_rate": 1.9636793511489377e-06,
+      "loss": 0.0229,
+      "step": 2513
+    },
+    {
+      "epoch": 0.716852010265184,
+      "grad_norm": 0.5250158672929854,
+      "learning_rate": 1.9600111448425285e-06,
+      "loss": 0.0159,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7171371542629028,
+      "grad_norm": 1.9676827735955886,
+      "learning_rate": 1.9563455324672566e-06,
+      "loss": 0.033,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7174222982606216,
+      "grad_norm": 0.648474303996466,
+      "learning_rate": 1.952682517150877e-06,
+      "loss": 0.0115,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7177074422583405,
+      "grad_norm": 1.3772601549572416,
+      "learning_rate": 1.9490221020189306e-06,
+      "loss": 0.0301,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7179925862560593,
+      "grad_norm": 1.526408416109426,
+      "learning_rate": 1.945364290194739e-06,
+      "loss": 0.0301,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7182777302537782,
+      "grad_norm": 1.00941806936837,
+      "learning_rate": 1.941709084799404e-06,
+      "loss": 0.0403,
+      "step": 2519
+    },
+    {
+      "epoch": 0.718562874251497,
+      "grad_norm": 0.7841790764743731,
+      "learning_rate": 1.9380564889518027e-06,
+      "loss": 0.0206,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7188480182492158,
+      "grad_norm": 0.3196980040678001,
+      "learning_rate": 1.9344065057685844e-06,
+      "loss": 0.0117,
+      "step": 2521
+    },
+    {
+      "epoch": 0.7191331622469347,
+      "grad_norm": 1.2494691029010219,
+      "learning_rate": 1.9307591383641704e-06,
+      "loss": 0.0271,
+      "step": 2522
+    },
+    {
+      "epoch": 0.7194183062446535,
+      "grad_norm": 3.35964679264394,
+      "learning_rate": 1.927114389850749e-06,
+      "loss": 0.0591,
+      "step": 2523
+    },
+    {
+      "epoch": 0.7197034502423724,
+      "grad_norm": 1.6156104346637272,
+      "learning_rate": 1.923472263338278e-06,
+      "loss": 0.0249,
+      "step": 2524
+    },
+    {
+      "epoch": 0.7199885942400912,
+      "grad_norm": 1.1690876897119913,
+      "learning_rate": 1.919832761934471e-06,
+      "loss": 0.0336,
+      "step": 2525
+    },
+    {
+      "epoch": 0.72027373823781,
+      "grad_norm": 0.4901043573896946,
+      "learning_rate": 1.9161958887448036e-06,
+      "loss": 0.0087,
+      "step": 2526
+    },
+    {
+      "epoch": 0.720558882235529,
+      "grad_norm": 0.5766787110046747,
+      "learning_rate": 1.912561646872515e-06,
+      "loss": 0.0194,
+      "step": 2527
+    },
+    {
+      "epoch": 0.7208440262332478,
+      "grad_norm": 0.4067719479340745,
+      "learning_rate": 1.908930039418593e-06,
+      "loss": 0.0074,
+      "step": 2528
+    },
+    {
+      "epoch": 0.7211291702309667,
+      "grad_norm": 0.8190616283681241,
+      "learning_rate": 1.9053010694817792e-06,
+      "loss": 0.0151,
+      "step": 2529
+    },
+    {
+      "epoch": 0.7214143142286855,
+      "grad_norm": 1.4745099441171658,
+      "learning_rate": 1.9016747401585612e-06,
+      "loss": 0.0192,
+      "step": 2530
+    },
+    {
+      "epoch": 0.7216994582264044,
+      "grad_norm": 1.1626077125676921,
+      "learning_rate": 1.8980510545431813e-06,
+      "loss": 0.0336,
+      "step": 2531
+    },
+    {
+      "epoch": 0.7219846022241232,
+      "grad_norm": 2.3232659525233785,
+      "learning_rate": 1.89443001572762e-06,
+      "loss": 0.0488,
+      "step": 2532
+    },
+    {
+      "epoch": 0.722269746221842,
+      "grad_norm": 0.4293645535795284,
+      "learning_rate": 1.8908116268016009e-06,
+      "loss": 0.0186,
+      "step": 2533
+    },
+    {
+      "epoch": 0.7225548902195609,
+      "grad_norm": 0.7057630192922806,
+      "learning_rate": 1.8871958908525861e-06,
+      "loss": 0.0165,
+      "step": 2534
+    },
+    {
+      "epoch": 0.7228400342172797,
+      "grad_norm": 0.7801979051938523,
+      "learning_rate": 1.883582810965775e-06,
+      "loss": 0.0173,
+      "step": 2535
+    },
+    {
+      "epoch": 0.7231251782149986,
+      "grad_norm": 0.6054884481515708,
+      "learning_rate": 1.8799723902240995e-06,
+      "loss": 0.012,
+      "step": 2536
+    },
+    {
+      "epoch": 0.7234103222127174,
+      "grad_norm": 0.7295030558864235,
+      "learning_rate": 1.8763646317082234e-06,
+      "loss": 0.0143,
+      "step": 2537
+    },
+    {
+      "epoch": 0.7236954662104362,
+      "grad_norm": 0.2348493963859001,
+      "learning_rate": 1.872759538496539e-06,
+      "loss": 0.0063,
+      "step": 2538
+    },
+    {
+      "epoch": 0.7239806102081551,
+      "grad_norm": 1.5317613085812878,
+      "learning_rate": 1.869157113665162e-06,
+      "loss": 0.0177,
+      "step": 2539
+    },
+    {
+      "epoch": 0.7242657542058739,
+      "grad_norm": 1.1185611844047032,
+      "learning_rate": 1.8655573602879384e-06,
+      "loss": 0.0406,
+      "step": 2540
+    },
+    {
+      "epoch": 0.7245508982035929,
+      "grad_norm": 0.7553842956065492,
+      "learning_rate": 1.8619602814364241e-06,
+      "loss": 0.0142,
+      "step": 2541
+    },
+    {
+      "epoch": 0.7248360422013117,
+      "grad_norm": 1.613711705936186,
+      "learning_rate": 1.8583658801798988e-06,
+      "loss": 0.021,
+      "step": 2542
+    },
+    {
+      "epoch": 0.7251211861990305,
+      "grad_norm": 0.6639471861383226,
+      "learning_rate": 1.8547741595853603e-06,
+      "loss": 0.0152,
+      "step": 2543
+    },
+    {
+      "epoch": 0.7254063301967494,
+      "grad_norm": 1.043444007239291,
+      "learning_rate": 1.8511851227175142e-06,
+      "loss": 0.0163,
+      "step": 2544
+    },
+    {
+      "epoch": 0.7256914741944682,
+      "grad_norm": 1.3859476624660125,
+      "learning_rate": 1.8475987726387783e-06,
+      "loss": 0.028,
+      "step": 2545
+    },
+    {
+      "epoch": 0.7259766181921871,
+      "grad_norm": 1.2707053562031854,
+      "learning_rate": 1.8440151124092764e-06,
+      "loss": 0.0203,
+      "step": 2546
+    },
+    {
+      "epoch": 0.7262617621899059,
+      "grad_norm": 0.7743398165083601,
+      "learning_rate": 1.8404341450868385e-06,
+      "loss": 0.0225,
+      "step": 2547
+    },
+    {
+      "epoch": 0.7265469061876247,
+      "grad_norm": 1.4948566545401907,
+      "learning_rate": 1.836855873726997e-06,
+      "loss": 0.0289,
+      "step": 2548
+    },
+    {
+      "epoch": 0.7268320501853436,
+      "grad_norm": 1.4926542884907947,
+      "learning_rate": 1.8332803013829824e-06,
+      "loss": 0.0314,
+      "step": 2549
+    },
+    {
+      "epoch": 0.7271171941830624,
+      "grad_norm": 1.7158591542352042,
+      "learning_rate": 1.8297074311057233e-06,
+      "loss": 0.0264,
+      "step": 2550
+    },
+    {
+      "epoch": 0.7274023381807813,
+      "grad_norm": 0.7051197790810769,
+      "learning_rate": 1.826137265943843e-06,
+      "loss": 0.0123,
+      "step": 2551
+    },
+    {
+      "epoch": 0.7276874821785001,
+      "grad_norm": 1.2332019677453634,
+      "learning_rate": 1.822569808943656e-06,
+      "loss": 0.0255,
+      "step": 2552
+    },
+    {
+      "epoch": 0.7279726261762189,
+      "grad_norm": 1.4227488885844022,
+      "learning_rate": 1.8190050631491662e-06,
+      "loss": 0.0186,
+      "step": 2553
+    },
+    {
+      "epoch": 0.7282577701739379,
+      "grad_norm": 0.888470253708542,
+      "learning_rate": 1.8154430316020638e-06,
+      "loss": 0.0091,
+      "step": 2554
+    },
+    {
+      "epoch": 0.7285429141716567,
+      "grad_norm": 0.6362280906290803,
+      "learning_rate": 1.811883717341722e-06,
+      "loss": 0.0227,
+      "step": 2555
+    },
+    {
+      "epoch": 0.7288280581693756,
+      "grad_norm": 2.0627256070084683,
+      "learning_rate": 1.8083271234051991e-06,
+      "loss": 0.0383,
+      "step": 2556
+    },
+    {
+      "epoch": 0.7291132021670944,
+      "grad_norm": 1.1067413056479523,
+      "learning_rate": 1.804773252827231e-06,
+      "loss": 0.0333,
+      "step": 2557
+    },
+    {
+      "epoch": 0.7293983461648132,
+      "grad_norm": 2.0036690116211333,
+      "learning_rate": 1.8012221086402226e-06,
+      "loss": 0.0302,
+      "step": 2558
+    },
+    {
+      "epoch": 0.7296834901625321,
+      "grad_norm": 1.7330904351751883,
+      "learning_rate": 1.7976736938742646e-06,
+      "loss": 0.0202,
+      "step": 2559
+    },
+    {
+      "epoch": 0.7299686341602509,
+      "grad_norm": 1.9066727681418387,
+      "learning_rate": 1.7941280115571103e-06,
+      "loss": 0.0521,
+      "step": 2560
+    },
+    {
+      "epoch": 0.7302537781579698,
+      "grad_norm": 0.8350093333223596,
+      "learning_rate": 1.7905850647141842e-06,
+      "loss": 0.0174,
+      "step": 2561
+    },
+    {
+      "epoch": 0.7305389221556886,
+      "grad_norm": 0.8935901878058463,
+      "learning_rate": 1.787044856368576e-06,
+      "loss": 0.0124,
+      "step": 2562
+    },
+    {
+      "epoch": 0.7308240661534074,
+      "grad_norm": 0.2609415835600623,
+      "learning_rate": 1.7835073895410393e-06,
+      "loss": 0.0034,
+      "step": 2563
+    },
+    {
+      "epoch": 0.7311092101511263,
+      "grad_norm": 0.7383110088545776,
+      "learning_rate": 1.779972667249989e-06,
+      "loss": 0.0132,
+      "step": 2564
+    },
+    {
+      "epoch": 0.7313943541488451,
+      "grad_norm": 0.82098631234755,
+      "learning_rate": 1.7764406925114957e-06,
+      "loss": 0.0181,
+      "step": 2565
+    },
+    {
+      "epoch": 0.731679498146564,
+      "grad_norm": 1.2122779163003397,
+      "learning_rate": 1.7729114683392889e-06,
+      "loss": 0.0186,
+      "step": 2566
+    },
+    {
+      "epoch": 0.7319646421442829,
+      "grad_norm": 1.0409332876987545,
+      "learning_rate": 1.769384997744747e-06,
+      "loss": 0.0218,
+      "step": 2567
+    },
+    {
+      "epoch": 0.7322497861420018,
+      "grad_norm": 1.0657193928152229,
+      "learning_rate": 1.7658612837369065e-06,
+      "loss": 0.0255,
+      "step": 2568
+    },
+    {
+      "epoch": 0.7325349301397206,
+      "grad_norm": 2.009539824029055,
+      "learning_rate": 1.7623403293224423e-06,
+      "loss": 0.0335,
+      "step": 2569
+    },
+    {
+      "epoch": 0.7328200741374394,
+      "grad_norm": 0.47135702796339735,
+      "learning_rate": 1.7588221375056797e-06,
+      "loss": 0.0124,
+      "step": 2570
+    },
+    {
+      "epoch": 0.7331052181351583,
+      "grad_norm": 0.5183616314191448,
+      "learning_rate": 1.7553067112885846e-06,
+      "loss": 0.0047,
+      "step": 2571
+    },
+    {
+      "epoch": 0.7333903621328771,
+      "grad_norm": 0.9381090221129461,
+      "learning_rate": 1.751794053670769e-06,
+      "loss": 0.0172,
+      "step": 2572
+    },
+    {
+      "epoch": 0.733675506130596,
+      "grad_norm": 1.5789937408733317,
+      "learning_rate": 1.7482841676494766e-06,
+      "loss": 0.0205,
+      "step": 2573
+    },
+    {
+      "epoch": 0.7339606501283148,
+      "grad_norm": 0.5928552002577685,
+      "learning_rate": 1.7447770562195831e-06,
+      "loss": 0.0137,
+      "step": 2574
+    },
+    {
+      "epoch": 0.7342457941260336,
+      "grad_norm": 1.1734406658202,
+      "learning_rate": 1.741272722373607e-06,
+      "loss": 0.0276,
+      "step": 2575
+    },
+    {
+      "epoch": 0.7345309381237525,
+      "grad_norm": 1.3469595275728594,
+      "learning_rate": 1.7377711691016885e-06,
+      "loss": 0.0291,
+      "step": 2576
+    },
+    {
+      "epoch": 0.7348160821214713,
+      "grad_norm": 1.0553587040388435,
+      "learning_rate": 1.7342723993915984e-06,
+      "loss": 0.0337,
+      "step": 2577
+    },
+    {
+      "epoch": 0.7351012261191902,
+      "grad_norm": 0.6104053843204919,
+      "learning_rate": 1.730776416228731e-06,
+      "loss": 0.0106,
+      "step": 2578
+    },
+    {
+      "epoch": 0.735386370116909,
+      "grad_norm": 1.180139523029748,
+      "learning_rate": 1.727283222596105e-06,
+      "loss": 0.0286,
+      "step": 2579
+    },
+    {
+      "epoch": 0.7356715141146278,
+      "grad_norm": 0.8943007056784072,
+      "learning_rate": 1.723792821474356e-06,
+      "loss": 0.0327,
+      "step": 2580
+    },
+    {
+      "epoch": 0.7359566581123468,
+      "grad_norm": 1.407472268943547,
+      "learning_rate": 1.7203052158417395e-06,
+      "loss": 0.039,
+      "step": 2581
+    },
+    {
+      "epoch": 0.7362418021100656,
+      "grad_norm": 1.2414400485072907,
+      "learning_rate": 1.7168204086741242e-06,
+      "loss": 0.0365,
+      "step": 2582
+    },
+    {
+      "epoch": 0.7365269461077845,
+      "grad_norm": 1.045965446497316,
+      "learning_rate": 1.7133384029449895e-06,
+      "loss": 0.0134,
+      "step": 2583
+    },
+    {
+      "epoch": 0.7368120901055033,
+      "grad_norm": 1.3401194939322665,
+      "learning_rate": 1.7098592016254318e-06,
+      "loss": 0.0192,
+      "step": 2584
+    },
+    {
+      "epoch": 0.7370972341032221,
+      "grad_norm": 0.4330957022031905,
+      "learning_rate": 1.7063828076841433e-06,
+      "loss": 0.0042,
+      "step": 2585
+    },
+    {
+      "epoch": 0.737382378100941,
+      "grad_norm": 0.6130445246080449,
+      "learning_rate": 1.7029092240874284e-06,
+      "loss": 0.0089,
+      "step": 2586
+    },
+    {
+      "epoch": 0.7376675220986598,
+      "grad_norm": 0.8820092232571117,
+      "learning_rate": 1.6994384537991898e-06,
+      "loss": 0.011,
+      "step": 2587
+    },
+    {
+      "epoch": 0.7379526660963787,
+      "grad_norm": 1.0509254864322741,
+      "learning_rate": 1.6959704997809355e-06,
+      "loss": 0.0291,
+      "step": 2588
+    },
+    {
+      "epoch": 0.7382378100940975,
+      "grad_norm": 1.3113452555468992,
+      "learning_rate": 1.6925053649917645e-06,
+      "loss": 0.0255,
+      "step": 2589
+    },
+    {
+      "epoch": 0.7385229540918163,
+      "grad_norm": 1.1114160968697633,
+      "learning_rate": 1.6890430523883715e-06,
+      "loss": 0.0176,
+      "step": 2590
+    },
+    {
+      "epoch": 0.7388080980895352,
+      "grad_norm": 1.4047944529889076,
+      "learning_rate": 1.6855835649250446e-06,
+      "loss": 0.0161,
+      "step": 2591
+    },
+    {
+      "epoch": 0.739093242087254,
+      "grad_norm": 1.268771098085473,
+      "learning_rate": 1.6821269055536604e-06,
+      "loss": 0.0359,
+      "step": 2592
+    },
+    {
+      "epoch": 0.7393783860849729,
+      "grad_norm": 1.0412207129886348,
+      "learning_rate": 1.678673077223682e-06,
+      "loss": 0.0111,
+      "step": 2593
+    },
+    {
+      "epoch": 0.7396635300826918,
+      "grad_norm": 1.9488662026753016,
+      "learning_rate": 1.6752220828821574e-06,
+      "loss": 0.0355,
+      "step": 2594
+    },
+    {
+      "epoch": 0.7399486740804107,
+      "grad_norm": 0.6388958326044936,
+      "learning_rate": 1.671773925473717e-06,
+      "loss": 0.0161,
+      "step": 2595
+    },
+    {
+      "epoch": 0.7402338180781295,
+      "grad_norm": 1.3676324169071015,
+      "learning_rate": 1.6683286079405692e-06,
+      "loss": 0.0446,
+      "step": 2596
+    },
+    {
+      "epoch": 0.7405189620758483,
+      "grad_norm": 0.7729955227718246,
+      "learning_rate": 1.6648861332225002e-06,
+      "loss": 0.0168,
+      "step": 2597
+    },
+    {
+      "epoch": 0.7408041060735672,
+      "grad_norm": 0.6524911607576815,
+      "learning_rate": 1.66144650425687e-06,
+      "loss": 0.0067,
+      "step": 2598
+    },
+    {
+      "epoch": 0.741089250071286,
+      "grad_norm": 0.9737064422050019,
+      "learning_rate": 1.6580097239786096e-06,
+      "loss": 0.0215,
+      "step": 2599
+    },
+    {
+      "epoch": 0.7413743940690048,
+      "grad_norm": 2.1748723955431237,
+      "learning_rate": 1.654575795320223e-06,
+      "loss": 0.0546,
+      "step": 2600
+    },
+    {
+      "epoch": 0.7416595380667237,
+      "grad_norm": 1.4285912646142453,
+      "learning_rate": 1.6511447212117786e-06,
+      "loss": 0.0346,
+      "step": 2601
+    },
+    {
+      "epoch": 0.7419446820644425,
+      "grad_norm": 2.7288639089504745,
+      "learning_rate": 1.6477165045809052e-06,
+      "loss": 0.0358,
+      "step": 2602
+    },
+    {
+      "epoch": 0.7422298260621614,
+      "grad_norm": 2.3650702688759018,
+      "learning_rate": 1.6442911483527978e-06,
+      "loss": 0.0669,
+      "step": 2603
+    },
+    {
+      "epoch": 0.7425149700598802,
+      "grad_norm": 1.2690767225620623,
+      "learning_rate": 1.6408686554502124e-06,
+      "loss": 0.0169,
+      "step": 2604
+    },
+    {
+      "epoch": 0.742800114057599,
+      "grad_norm": 1.4869660462533452,
+      "learning_rate": 1.637449028793458e-06,
+      "loss": 0.0294,
+      "step": 2605
+    },
+    {
+      "epoch": 0.7430852580553179,
+      "grad_norm": 1.3737598171297696,
+      "learning_rate": 1.6340322713003992e-06,
+      "loss": 0.026,
+      "step": 2606
+    },
+    {
+      "epoch": 0.7433704020530368,
+      "grad_norm": 0.5276542234582096,
+      "learning_rate": 1.6306183858864528e-06,
+      "loss": 0.0108,
+      "step": 2607
+    },
+    {
+      "epoch": 0.7436555460507557,
+      "grad_norm": 1.208035745863832,
+      "learning_rate": 1.6272073754645845e-06,
+      "loss": 0.0434,
+      "step": 2608
+    },
+    {
+      "epoch": 0.7439406900484745,
+      "grad_norm": 1.2060462983723554,
+      "learning_rate": 1.623799242945307e-06,
+      "loss": 0.0599,
+      "step": 2609
+    },
+    {
+      "epoch": 0.7442258340461934,
+      "grad_norm": 1.613815953946869,
+      "learning_rate": 1.6203939912366768e-06,
+      "loss": 0.0241,
+      "step": 2610
+    },
+    {
+      "epoch": 0.7445109780439122,
+      "grad_norm": 1.460570451296447,
+      "learning_rate": 1.6169916232442923e-06,
+      "loss": 0.0306,
+      "step": 2611
+    },
+    {
+      "epoch": 0.744796122041631,
+      "grad_norm": 1.0008169723964753,
+      "learning_rate": 1.6135921418712959e-06,
+      "loss": 0.0215,
+      "step": 2612
+    },
+    {
+      "epoch": 0.7450812660393499,
+      "grad_norm": 2.086617218943197,
+      "learning_rate": 1.610195550018358e-06,
+      "loss": 0.0274,
+      "step": 2613
+    },
+    {
+      "epoch": 0.7453664100370687,
+      "grad_norm": 1.6041573788510077,
+      "learning_rate": 1.6068018505836901e-06,
+      "loss": 0.019,
+      "step": 2614
+    },
+    {
+      "epoch": 0.7456515540347876,
+      "grad_norm": 0.7713053016927894,
+      "learning_rate": 1.6034110464630325e-06,
+      "loss": 0.0232,
+      "step": 2615
+    },
+    {
+      "epoch": 0.7459366980325064,
+      "grad_norm": 0.5511320080746058,
+      "learning_rate": 1.6000231405496602e-06,
+      "loss": 0.0191,
+      "step": 2616
+    },
+    {
+      "epoch": 0.7462218420302252,
+      "grad_norm": 1.0634637982976405,
+      "learning_rate": 1.5966381357343708e-06,
+      "loss": 0.0175,
+      "step": 2617
+    },
+    {
+      "epoch": 0.7465069860279441,
+      "grad_norm": 1.0374330207474594,
+      "learning_rate": 1.5932560349054838e-06,
+      "loss": 0.0245,
+      "step": 2618
+    },
+    {
+      "epoch": 0.7467921300256629,
+      "grad_norm": 1.0108319276571467,
+      "learning_rate": 1.5898768409488447e-06,
+      "loss": 0.0136,
+      "step": 2619
+    },
+    {
+      "epoch": 0.7470772740233818,
+      "grad_norm": 1.6396915594949968,
+      "learning_rate": 1.5865005567478215e-06,
+      "loss": 0.0251,
+      "step": 2620
+    },
+    {
+      "epoch": 0.7473624180211007,
+      "grad_norm": 0.9627999719461374,
+      "learning_rate": 1.5831271851832937e-06,
+      "loss": 0.0391,
+      "step": 2621
+    },
+    {
+      "epoch": 0.7476475620188195,
+      "grad_norm": 0.8253787374724513,
+      "learning_rate": 1.5797567291336586e-06,
+      "loss": 0.0122,
+      "step": 2622
+    },
+    {
+      "epoch": 0.7479327060165384,
+      "grad_norm": 0.6661195439268087,
+      "learning_rate": 1.5763891914748241e-06,
+      "loss": 0.0221,
+      "step": 2623
+    },
+    {
+      "epoch": 0.7482178500142572,
+      "grad_norm": 0.8779901773734643,
+      "learning_rate": 1.5730245750802103e-06,
+      "loss": 0.0195,
+      "step": 2624
+    },
+    {
+      "epoch": 0.7485029940119761,
+      "grad_norm": 1.6634272557568168,
+      "learning_rate": 1.5696628828207421e-06,
+      "loss": 0.037,
+      "step": 2625
+    },
+    {
+      "epoch": 0.7487881380096949,
+      "grad_norm": 1.4516278438233137,
+      "learning_rate": 1.5663041175648513e-06,
+      "loss": 0.0219,
+      "step": 2626
+    },
+    {
+      "epoch": 0.7490732820074137,
+      "grad_norm": 1.907475947775475,
+      "learning_rate": 1.5629482821784691e-06,
+      "loss": 0.0314,
+      "step": 2627
+    },
+    {
+      "epoch": 0.7493584260051326,
+      "grad_norm": 1.2003200972703238,
+      "learning_rate": 1.5595953795250346e-06,
+      "loss": 0.0242,
+      "step": 2628
+    },
+    {
+      "epoch": 0.7496435700028514,
+      "grad_norm": 1.1021426342768983,
+      "learning_rate": 1.556245412465473e-06,
+      "loss": 0.034,
+      "step": 2629
+    },
+    {
+      "epoch": 0.7499287140005703,
+      "grad_norm": 1.0302550777919475,
+      "learning_rate": 1.5528983838582129e-06,
+      "loss": 0.0285,
+      "step": 2630
+    },
+    {
+      "epoch": 0.7502138579982891,
+      "grad_norm": 0.8308789044386139,
+      "learning_rate": 1.5495542965591709e-06,
+      "loss": 0.0114,
+      "step": 2631
+    },
+    {
+      "epoch": 0.7504990019960079,
+      "grad_norm": 0.8819505544759334,
+      "learning_rate": 1.5462131534217607e-06,
+      "loss": 0.0159,
+      "step": 2632
+    },
+    {
+      "epoch": 0.7507841459937268,
+      "grad_norm": 0.9835443580092531,
+      "learning_rate": 1.542874957296876e-06,
+      "loss": 0.0154,
+      "step": 2633
+    },
+    {
+      "epoch": 0.7510692899914457,
+      "grad_norm": 1.0665665377487932,
+      "learning_rate": 1.5395397110329024e-06,
+      "loss": 0.0353,
+      "step": 2634
+    },
+    {
+      "epoch": 0.7513544339891646,
+      "grad_norm": 0.8164632784798294,
+      "learning_rate": 1.5362074174756998e-06,
+      "loss": 0.0134,
+      "step": 2635
+    },
+    {
+      "epoch": 0.7516395779868834,
+      "grad_norm": 1.646300986945349,
+      "learning_rate": 1.5328780794686188e-06,
+      "loss": 0.0394,
+      "step": 2636
+    },
+    {
+      "epoch": 0.7519247219846023,
+      "grad_norm": 1.1301313594828286,
+      "learning_rate": 1.5295516998524823e-06,
+      "loss": 0.0142,
+      "step": 2637
+    },
+    {
+      "epoch": 0.7522098659823211,
+      "grad_norm": 0.5764109215471761,
+      "learning_rate": 1.5262282814655893e-06,
+      "loss": 0.0142,
+      "step": 2638
+    },
+    {
+      "epoch": 0.7524950099800399,
+      "grad_norm": 1.3610524683318708,
+      "learning_rate": 1.5229078271437141e-06,
+      "loss": 0.034,
+      "step": 2639
+    },
+    {
+      "epoch": 0.7527801539777588,
+      "grad_norm": 0.477649205446183,
+      "learning_rate": 1.5195903397200996e-06,
+      "loss": 0.0109,
+      "step": 2640
+    },
+    {
+      "epoch": 0.7530652979754776,
+      "grad_norm": 1.515573238930559,
+      "learning_rate": 1.5162758220254586e-06,
+      "loss": 0.0166,
+      "step": 2641
+    },
+    {
+      "epoch": 0.7533504419731964,
+      "grad_norm": 0.7091310562536813,
+      "learning_rate": 1.5129642768879687e-06,
+      "loss": 0.0135,
+      "step": 2642
+    },
+    {
+      "epoch": 0.7536355859709153,
+      "grad_norm": 1.5074902375844876,
+      "learning_rate": 1.5096557071332712e-06,
+      "loss": 0.0253,
+      "step": 2643
+    },
+    {
+      "epoch": 0.7539207299686341,
+      "grad_norm": 1.0767879630125734,
+      "learning_rate": 1.5063501155844723e-06,
+      "loss": 0.0167,
+      "step": 2644
+    },
+    {
+      "epoch": 0.754205873966353,
+      "grad_norm": 3.1348957847232226,
+      "learning_rate": 1.5030475050621336e-06,
+      "loss": 0.0288,
+      "step": 2645
+    },
+    {
+      "epoch": 0.7544910179640718,
+      "grad_norm": 0.6684133082529641,
+      "learning_rate": 1.49974787838427e-06,
+      "loss": 0.0084,
+      "step": 2646
+    },
+    {
+      "epoch": 0.7547761619617908,
+      "grad_norm": 0.7727753084015943,
+      "learning_rate": 1.4964512383663544e-06,
+      "loss": 0.0219,
+      "step": 2647
+    },
+    {
+      "epoch": 0.7550613059595096,
+      "grad_norm": 2.0188024106365887,
+      "learning_rate": 1.4931575878213127e-06,
+      "loss": 0.0395,
+      "step": 2648
+    },
+    {
+      "epoch": 0.7553464499572284,
+      "grad_norm": 0.4906606761025358,
+      "learning_rate": 1.4898669295595181e-06,
+      "loss": 0.0069,
+      "step": 2649
+    },
+    {
+      "epoch": 0.7556315939549473,
+      "grad_norm": 1.260687367143981,
+      "learning_rate": 1.4865792663887907e-06,
+      "loss": 0.0319,
+      "step": 2650
+    },
+    {
+      "epoch": 0.7559167379526661,
+      "grad_norm": 1.0883273196471233,
+      "learning_rate": 1.4832946011143906e-06,
+      "loss": 0.0264,
+      "step": 2651
+    },
+    {
+      "epoch": 0.756201881950385,
+      "grad_norm": 0.35999385019644614,
+      "learning_rate": 1.4800129365390282e-06,
+      "loss": 0.0055,
+      "step": 2652
+    },
+    {
+      "epoch": 0.7564870259481038,
+      "grad_norm": 0.4561544926370098,
+      "learning_rate": 1.4767342754628477e-06,
+      "loss": 0.0095,
+      "step": 2653
+    },
+    {
+      "epoch": 0.7567721699458226,
+      "grad_norm": 0.8901887574735574,
+      "learning_rate": 1.4734586206834323e-06,
+      "loss": 0.0198,
+      "step": 2654
+    },
+    {
+      "epoch": 0.7570573139435415,
+      "grad_norm": 0.7083720617572299,
+      "learning_rate": 1.4701859749958004e-06,
+      "loss": 0.0152,
+      "step": 2655
+    },
+    {
+      "epoch": 0.7573424579412603,
+      "grad_norm": 1.1369840531128947,
+      "learning_rate": 1.466916341192401e-06,
+      "loss": 0.0141,
+      "step": 2656
+    },
+    {
+      "epoch": 0.7576276019389792,
+      "grad_norm": 1.2957588766875028,
+      "learning_rate": 1.4636497220631145e-06,
+      "loss": 0.0336,
+      "step": 2657
+    },
+    {
+      "epoch": 0.757912745936698,
+      "grad_norm": 2.0713263046299173,
+      "learning_rate": 1.4603861203952502e-06,
+      "loss": 0.0302,
+      "step": 2658
+    },
+    {
+      "epoch": 0.7581978899344168,
+      "grad_norm": 1.1733273590818454,
+      "learning_rate": 1.4571255389735385e-06,
+      "loss": 0.0234,
+      "step": 2659
+    },
+    {
+      "epoch": 0.7584830339321357,
+      "grad_norm": 2.1819052963632717,
+      "learning_rate": 1.4538679805801386e-06,
+      "loss": 0.0329,
+      "step": 2660
+    },
+    {
+      "epoch": 0.7587681779298546,
+      "grad_norm": 0.3495163345720255,
+      "learning_rate": 1.4506134479946281e-06,
+      "loss": 0.0098,
+      "step": 2661
+    },
+    {
+      "epoch": 0.7590533219275735,
+      "grad_norm": 0.5490129444641568,
+      "learning_rate": 1.4473619439939985e-06,
+      "loss": 0.0065,
+      "step": 2662
+    },
+    {
+      "epoch": 0.7593384659252923,
+      "grad_norm": 1.581308447857722,
+      "learning_rate": 1.4441134713526595e-06,
+      "loss": 0.057,
+      "step": 2663
+    },
+    {
+      "epoch": 0.7596236099230111,
+      "grad_norm": 1.0260023381149048,
+      "learning_rate": 1.440868032842439e-06,
+      "loss": 0.0185,
+      "step": 2664
+    },
+    {
+      "epoch": 0.75990875392073,
+      "grad_norm": 0.5787379526406348,
+      "learning_rate": 1.437625631232571e-06,
+      "loss": 0.0151,
+      "step": 2665
+    },
+    {
+      "epoch": 0.7601938979184488,
+      "grad_norm": 1.5032898745586536,
+      "learning_rate": 1.4343862692896986e-06,
+      "loss": 0.0191,
+      "step": 2666
+    },
+    {
+      "epoch": 0.7604790419161677,
+      "grad_norm": 0.9364587983512026,
+      "learning_rate": 1.431149949777873e-06,
+      "loss": 0.0138,
+      "step": 2667
+    },
+    {
+      "epoch": 0.7607641859138865,
+      "grad_norm": 0.7579871237730098,
+      "learning_rate": 1.4279166754585472e-06,
+      "loss": 0.0236,
+      "step": 2668
+    },
+    {
+      "epoch": 0.7610493299116053,
+      "grad_norm": 1.2909526522887504,
+      "learning_rate": 1.4246864490905776e-06,
+      "loss": 0.0228,
+      "step": 2669
+    },
+    {
+      "epoch": 0.7613344739093242,
+      "grad_norm": 1.313113386385982,
+      "learning_rate": 1.421459273430219e-06,
+      "loss": 0.0234,
+      "step": 2670
+    },
+    {
+      "epoch": 0.761619617907043,
+      "grad_norm": 0.8158285471748641,
+      "learning_rate": 1.4182351512311237e-06,
+      "loss": 0.027,
+      "step": 2671
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 1.1047514162563927,
+      "learning_rate": 1.4150140852443389e-06,
+      "loss": 0.016,
+      "step": 2672
+    },
+    {
+      "epoch": 0.7621899059024807,
+      "grad_norm": 0.7487315401677382,
+      "learning_rate": 1.4117960782183021e-06,
+      "loss": 0.0075,
+      "step": 2673
+    },
+    {
+      "epoch": 0.7624750499001997,
+      "grad_norm": 1.5926773015253632,
+      "learning_rate": 1.4085811328988425e-06,
+      "loss": 0.0245,
+      "step": 2674
+    },
+    {
+      "epoch": 0.7627601938979185,
+      "grad_norm": 1.0230513489809057,
+      "learning_rate": 1.405369252029175e-06,
+      "loss": 0.0108,
+      "step": 2675
+    },
+    {
+      "epoch": 0.7630453378956373,
+      "grad_norm": 0.6835988253710413,
+      "learning_rate": 1.4021604383499044e-06,
+      "loss": 0.0141,
+      "step": 2676
+    },
+    {
+      "epoch": 0.7633304818933562,
+      "grad_norm": 1.2256591335523719,
+      "learning_rate": 1.3989546945990129e-06,
+      "loss": 0.0158,
+      "step": 2677
+    },
+    {
+      "epoch": 0.763615625891075,
+      "grad_norm": 1.2930161491847667,
+      "learning_rate": 1.395752023511867e-06,
+      "loss": 0.0195,
+      "step": 2678
+    },
+    {
+      "epoch": 0.7639007698887939,
+      "grad_norm": 0.8994233014350279,
+      "learning_rate": 1.392552427821205e-06,
+      "loss": 0.0087,
+      "step": 2679
+    },
+    {
+      "epoch": 0.7641859138865127,
+      "grad_norm": 0.6770559844138675,
+      "learning_rate": 1.3893559102571502e-06,
+      "loss": 0.0141,
+      "step": 2680
+    },
+    {
+      "epoch": 0.7644710578842315,
+      "grad_norm": 1.4930856086234503,
+      "learning_rate": 1.386162473547194e-06,
+      "loss": 0.0343,
+      "step": 2681
+    },
+    {
+      "epoch": 0.7647562018819504,
+      "grad_norm": 1.682448635821019,
+      "learning_rate": 1.3829721204162e-06,
+      "loss": 0.0297,
+      "step": 2682
+    },
+    {
+      "epoch": 0.7650413458796692,
+      "grad_norm": 1.5491061182481676,
+      "learning_rate": 1.3797848535863995e-06,
+      "loss": 0.0326,
+      "step": 2683
+    },
+    {
+      "epoch": 0.765326489877388,
+      "grad_norm": 1.175627444484784,
+      "learning_rate": 1.3766006757773932e-06,
+      "loss": 0.0218,
+      "step": 2684
+    },
+    {
+      "epoch": 0.7656116338751069,
+      "grad_norm": 0.9840348835612843,
+      "learning_rate": 1.373419589706143e-06,
+      "loss": 0.0211,
+      "step": 2685
+    },
+    {
+      "epoch": 0.7658967778728257,
+      "grad_norm": 1.0057768113881969,
+      "learning_rate": 1.3702415980869743e-06,
+      "loss": 0.0464,
+      "step": 2686
+    },
+    {
+      "epoch": 0.7661819218705447,
+      "grad_norm": 1.7830889645336818,
+      "learning_rate": 1.3670667036315728e-06,
+      "loss": 0.0185,
+      "step": 2687
+    },
+    {
+      "epoch": 0.7664670658682635,
+      "grad_norm": 0.4224535629956841,
+      "learning_rate": 1.3638949090489772e-06,
+      "loss": 0.0081,
+      "step": 2688
+    },
+    {
+      "epoch": 0.7667522098659824,
+      "grad_norm": 0.6377033854455558,
+      "learning_rate": 1.360726217045591e-06,
+      "loss": 0.01,
+      "step": 2689
+    },
+    {
+      "epoch": 0.7670373538637012,
+      "grad_norm": 1.6543881718541624,
+      "learning_rate": 1.3575606303251582e-06,
+      "loss": 0.0274,
+      "step": 2690
+    },
+    {
+      "epoch": 0.76732249786142,
+      "grad_norm": 0.705004441035942,
+      "learning_rate": 1.3543981515887788e-06,
+      "loss": 0.0168,
+      "step": 2691
+    },
+    {
+      "epoch": 0.7676076418591389,
+      "grad_norm": 1.5935922533656512,
+      "learning_rate": 1.3512387835349045e-06,
+      "loss": 0.0248,
+      "step": 2692
+    },
+    {
+      "epoch": 0.7678927858568577,
+      "grad_norm": 0.7488281251727223,
+      "learning_rate": 1.3480825288593274e-06,
+      "loss": 0.0139,
+      "step": 2693
+    },
+    {
+      "epoch": 0.7681779298545766,
+      "grad_norm": 0.8165161964838176,
+      "learning_rate": 1.3449293902551857e-06,
+      "loss": 0.0199,
+      "step": 2694
+    },
+    {
+      "epoch": 0.7684630738522954,
+      "grad_norm": 1.5726186357766658,
+      "learning_rate": 1.341779370412954e-06,
+      "loss": 0.0364,
+      "step": 2695
+    },
+    {
+      "epoch": 0.7687482178500142,
+      "grad_norm": 0.8244759122464347,
+      "learning_rate": 1.3386324720204542e-06,
+      "loss": 0.0119,
+      "step": 2696
+    },
+    {
+      "epoch": 0.7690333618477331,
+      "grad_norm": 1.5946273315050323,
+      "learning_rate": 1.3354886977628374e-06,
+      "loss": 0.0267,
+      "step": 2697
+    },
+    {
+      "epoch": 0.7693185058454519,
+      "grad_norm": 0.8033194869459227,
+      "learning_rate": 1.3323480503225933e-06,
+      "loss": 0.0086,
+      "step": 2698
+    },
+    {
+      "epoch": 0.7696036498431708,
+      "grad_norm": 1.1393108615672012,
+      "learning_rate": 1.3292105323795406e-06,
+      "loss": 0.0277,
+      "step": 2699
+    },
+    {
+      "epoch": 0.7698887938408896,
+      "grad_norm": 1.9860907344454382,
+      "learning_rate": 1.3260761466108302e-06,
+      "loss": 0.0211,
+      "step": 2700
+    },
+    {
+      "epoch": 0.7701739378386085,
+      "grad_norm": 1.5082456222402305,
+      "learning_rate": 1.3229448956909385e-06,
+      "loss": 0.022,
+      "step": 2701
+    },
+    {
+      "epoch": 0.7704590818363274,
+      "grad_norm": 1.5344921371462583,
+      "learning_rate": 1.3198167822916685e-06,
+      "loss": 0.0245,
+      "step": 2702
+    },
+    {
+      "epoch": 0.7707442258340462,
+      "grad_norm": 0.7564484297441108,
+      "learning_rate": 1.3166918090821445e-06,
+      "loss": 0.0141,
+      "step": 2703
+    },
+    {
+      "epoch": 0.7710293698317651,
+      "grad_norm": 0.43057532343881677,
+      "learning_rate": 1.3135699787288114e-06,
+      "loss": 0.0059,
+      "step": 2704
+    },
+    {
+      "epoch": 0.7713145138294839,
+      "grad_norm": 1.0858462972323473,
+      "learning_rate": 1.3104512938954373e-06,
+      "loss": 0.0327,
+      "step": 2705
+    },
+    {
+      "epoch": 0.7715996578272027,
+      "grad_norm": 1.1709018893965801,
+      "learning_rate": 1.3073357572430984e-06,
+      "loss": 0.032,
+      "step": 2706
+    },
+    {
+      "epoch": 0.7718848018249216,
+      "grad_norm": 0.28314832345320107,
+      "learning_rate": 1.3042233714301893e-06,
+      "loss": 0.0063,
+      "step": 2707
+    },
+    {
+      "epoch": 0.7721699458226404,
+      "grad_norm": 1.0670337951747642,
+      "learning_rate": 1.3011141391124138e-06,
+      "loss": 0.017,
+      "step": 2708
+    },
+    {
+      "epoch": 0.7724550898203593,
+      "grad_norm": 1.5924118733503132,
+      "learning_rate": 1.2980080629427904e-06,
+      "loss": 0.0261,
+      "step": 2709
+    },
+    {
+      "epoch": 0.7727402338180781,
+      "grad_norm": 0.8829070105724145,
+      "learning_rate": 1.2949051455716378e-06,
+      "loss": 0.0107,
+      "step": 2710
+    },
+    {
+      "epoch": 0.7730253778157969,
+      "grad_norm": 0.4216528848394497,
+      "learning_rate": 1.2918053896465826e-06,
+      "loss": 0.012,
+      "step": 2711
+    },
+    {
+      "epoch": 0.7733105218135158,
+      "grad_norm": 0.7050612966316768,
+      "learning_rate": 1.2887087978125535e-06,
+      "loss": 0.0093,
+      "step": 2712
+    },
+    {
+      "epoch": 0.7735956658112346,
+      "grad_norm": 1.1605581061863133,
+      "learning_rate": 1.2856153727117792e-06,
+      "loss": 0.0212,
+      "step": 2713
+    },
+    {
+      "epoch": 0.7738808098089536,
+      "grad_norm": 1.3212374854178839,
+      "learning_rate": 1.2825251169837865e-06,
+      "loss": 0.0224,
+      "step": 2714
+    },
+    {
+      "epoch": 0.7741659538066724,
+      "grad_norm": 1.1885252640350221,
+      "learning_rate": 1.2794380332653976e-06,
+      "loss": 0.0208,
+      "step": 2715
+    },
+    {
+      "epoch": 0.7744510978043913,
+      "grad_norm": 1.7760519011776683,
+      "learning_rate": 1.2763541241907268e-06,
+      "loss": 0.0569,
+      "step": 2716
+    },
+    {
+      "epoch": 0.7747362418021101,
+      "grad_norm": 0.9620142840295433,
+      "learning_rate": 1.2732733923911854e-06,
+      "loss": 0.0161,
+      "step": 2717
+    },
+    {
+      "epoch": 0.7750213857998289,
+      "grad_norm": 1.1933906064161974,
+      "learning_rate": 1.270195840495465e-06,
+      "loss": 0.0181,
+      "step": 2718
+    },
+    {
+      "epoch": 0.7753065297975478,
+      "grad_norm": 1.0984270689396793,
+      "learning_rate": 1.2671214711295492e-06,
+      "loss": 0.0144,
+      "step": 2719
+    },
+    {
+      "epoch": 0.7755916737952666,
+      "grad_norm": 0.7603521204159235,
+      "learning_rate": 1.2640502869167043e-06,
+      "loss": 0.014,
+      "step": 2720
+    },
+    {
+      "epoch": 0.7758768177929855,
+      "grad_norm": 0.42403568104580863,
+      "learning_rate": 1.260982290477482e-06,
+      "loss": 0.0083,
+      "step": 2721
+    },
+    {
+      "epoch": 0.7761619617907043,
+      "grad_norm": 0.865345796891974,
+      "learning_rate": 1.2579174844297114e-06,
+      "loss": 0.0112,
+      "step": 2722
+    },
+    {
+      "epoch": 0.7764471057884231,
+      "grad_norm": 1.3649714875018568,
+      "learning_rate": 1.2548558713884963e-06,
+      "loss": 0.0109,
+      "step": 2723
+    },
+    {
+      "epoch": 0.776732249786142,
+      "grad_norm": 0.8805998540651545,
+      "learning_rate": 1.2517974539662198e-06,
+      "loss": 0.0181,
+      "step": 2724
+    },
+    {
+      "epoch": 0.7770173937838608,
+      "grad_norm": 1.133487480597252,
+      "learning_rate": 1.2487422347725397e-06,
+      "loss": 0.0439,
+      "step": 2725
+    },
+    {
+      "epoch": 0.7773025377815796,
+      "grad_norm": 1.077854176177537,
+      "learning_rate": 1.2456902164143813e-06,
+      "loss": 0.0107,
+      "step": 2726
+    },
+    {
+      "epoch": 0.7775876817792986,
+      "grad_norm": 1.4419228139613494,
+      "learning_rate": 1.2426414014959409e-06,
+      "loss": 0.0252,
+      "step": 2727
+    },
+    {
+      "epoch": 0.7778728257770174,
+      "grad_norm": 0.40646062715513975,
+      "learning_rate": 1.2395957926186802e-06,
+      "loss": 0.0048,
+      "step": 2728
+    },
+    {
+      "epoch": 0.7781579697747363,
+      "grad_norm": 0.8225502003586711,
+      "learning_rate": 1.2365533923813256e-06,
+      "loss": 0.0199,
+      "step": 2729
+    },
+    {
+      "epoch": 0.7784431137724551,
+      "grad_norm": 1.59562236103222,
+      "learning_rate": 1.2335142033798658e-06,
+      "loss": 0.0508,
+      "step": 2730
+    },
+    {
+      "epoch": 0.778728257770174,
+      "grad_norm": 1.233387689768871,
+      "learning_rate": 1.23047822820755e-06,
+      "loss": 0.0287,
+      "step": 2731
+    },
+    {
+      "epoch": 0.7790134017678928,
+      "grad_norm": 0.5500387337491803,
+      "learning_rate": 1.227445469454882e-06,
+      "loss": 0.0065,
+      "step": 2732
+    },
+    {
+      "epoch": 0.7792985457656116,
+      "grad_norm": 2.0299603071873533,
+      "learning_rate": 1.2244159297096291e-06,
+      "loss": 0.0415,
+      "step": 2733
+    },
+    {
+      "epoch": 0.7795836897633305,
+      "grad_norm": 1.353228704932132,
+      "learning_rate": 1.2213896115568024e-06,
+      "loss": 0.0387,
+      "step": 2734
+    },
+    {
+      "epoch": 0.7798688337610493,
+      "grad_norm": 1.6415096320435094,
+      "learning_rate": 1.2183665175786686e-06,
+      "loss": 0.0424,
+      "step": 2735
+    },
+    {
+      "epoch": 0.7801539777587682,
+      "grad_norm": 2.374020647253202,
+      "learning_rate": 1.2153466503547417e-06,
+      "loss": 0.0519,
+      "step": 2736
+    },
+    {
+      "epoch": 0.780439121756487,
+      "grad_norm": 1.8977748402119863,
+      "learning_rate": 1.2123300124617876e-06,
+      "loss": 0.0722,
+      "step": 2737
+    },
+    {
+      "epoch": 0.7807242657542058,
+      "grad_norm": 0.7295480552079158,
+      "learning_rate": 1.2093166064738098e-06,
+      "loss": 0.0135,
+      "step": 2738
+    },
+    {
+      "epoch": 0.7810094097519247,
+      "grad_norm": 0.8229549593104721,
+      "learning_rate": 1.2063064349620596e-06,
+      "loss": 0.0259,
+      "step": 2739
+    },
+    {
+      "epoch": 0.7812945537496435,
+      "grad_norm": 0.6925997286050343,
+      "learning_rate": 1.2032995004950216e-06,
+      "loss": 0.0146,
+      "step": 2740
+    },
+    {
+      "epoch": 0.7815796977473625,
+      "grad_norm": 0.9014027542875058,
+      "learning_rate": 1.2002958056384262e-06,
+      "loss": 0.0234,
+      "step": 2741
+    },
+    {
+      "epoch": 0.7818648417450813,
+      "grad_norm": 0.706354278112295,
+      "learning_rate": 1.197295352955235e-06,
+      "loss": 0.0086,
+      "step": 2742
+    },
+    {
+      "epoch": 0.7821499857428001,
+      "grad_norm": 0.8645190729336512,
+      "learning_rate": 1.194298145005644e-06,
+      "loss": 0.017,
+      "step": 2743
+    },
+    {
+      "epoch": 0.782435129740519,
+      "grad_norm": 0.37228866967958824,
+      "learning_rate": 1.1913041843470807e-06,
+      "loss": 0.0096,
+      "step": 2744
+    },
+    {
+      "epoch": 0.7827202737382378,
+      "grad_norm": 0.6034253285490447,
+      "learning_rate": 1.1883134735342023e-06,
+      "loss": 0.0153,
+      "step": 2745
+    },
+    {
+      "epoch": 0.7830054177359567,
+      "grad_norm": 1.1035155409154056,
+      "learning_rate": 1.1853260151188912e-06,
+      "loss": 0.0318,
+      "step": 2746
+    },
+    {
+      "epoch": 0.7832905617336755,
+      "grad_norm": 0.7012984171334737,
+      "learning_rate": 1.1823418116502566e-06,
+      "loss": 0.0133,
+      "step": 2747
+    },
+    {
+      "epoch": 0.7835757057313943,
+      "grad_norm": 2.2145145784676674,
+      "learning_rate": 1.1793608656746286e-06,
+      "loss": 0.0276,
+      "step": 2748
+    },
+    {
+      "epoch": 0.7838608497291132,
+      "grad_norm": 1.4050671677912367,
+      "learning_rate": 1.1763831797355612e-06,
+      "loss": 0.0201,
+      "step": 2749
+    },
+    {
+      "epoch": 0.784145993726832,
+      "grad_norm": 0.8314849154599873,
+      "learning_rate": 1.1734087563738245e-06,
+      "loss": 0.0146,
+      "step": 2750
+    },
+    {
+      "epoch": 0.7844311377245509,
+      "grad_norm": 1.6532863006133396,
+      "learning_rate": 1.1704375981274019e-06,
+      "loss": 0.0234,
+      "step": 2751
+    },
+    {
+      "epoch": 0.7847162817222697,
+      "grad_norm": 0.6670573448259365,
+      "learning_rate": 1.1674697075314928e-06,
+      "loss": 0.0135,
+      "step": 2752
+    },
+    {
+      "epoch": 0.7850014257199885,
+      "grad_norm": 0.6240914012825134,
+      "learning_rate": 1.1645050871185121e-06,
+      "loss": 0.0116,
+      "step": 2753
+    },
+    {
+      "epoch": 0.7852865697177075,
+      "grad_norm": 0.9870483425644136,
+      "learning_rate": 1.1615437394180806e-06,
+      "loss": 0.0107,
+      "step": 2754
+    },
+    {
+      "epoch": 0.7855717137154263,
+      "grad_norm": 1.9210648543710227,
+      "learning_rate": 1.1585856669570277e-06,
+      "loss": 0.0399,
+      "step": 2755
+    },
+    {
+      "epoch": 0.7858568577131452,
+      "grad_norm": 1.7634186455826524,
+      "learning_rate": 1.1556308722593845e-06,
+      "loss": 0.0589,
+      "step": 2756
+    },
+    {
+      "epoch": 0.786142001710864,
+      "grad_norm": 1.2268952626774527,
+      "learning_rate": 1.152679357846392e-06,
+      "loss": 0.0232,
+      "step": 2757
+    },
+    {
+      "epoch": 0.7864271457085829,
+      "grad_norm": 1.456806678520593,
+      "learning_rate": 1.1497311262364874e-06,
+      "loss": 0.0243,
+      "step": 2758
+    },
+    {
+      "epoch": 0.7867122897063017,
+      "grad_norm": 0.4950354070457815,
+      "learning_rate": 1.1467861799453084e-06,
+      "loss": 0.0072,
+      "step": 2759
+    },
+    {
+      "epoch": 0.7869974337040205,
+      "grad_norm": 0.8080613169821007,
+      "learning_rate": 1.143844521485688e-06,
+      "loss": 0.016,
+      "step": 2760
+    },
+    {
+      "epoch": 0.7872825777017394,
+      "grad_norm": 1.600494453613366,
+      "learning_rate": 1.1409061533676586e-06,
+      "loss": 0.0267,
+      "step": 2761
+    },
+    {
+      "epoch": 0.7875677216994582,
+      "grad_norm": 1.9230161162665296,
+      "learning_rate": 1.1379710780984376e-06,
+      "loss": 0.0278,
+      "step": 2762
+    },
+    {
+      "epoch": 0.787852865697177,
+      "grad_norm": 1.04041200038653,
+      "learning_rate": 1.1350392981824375e-06,
+      "loss": 0.0226,
+      "step": 2763
+    },
+    {
+      "epoch": 0.7881380096948959,
+      "grad_norm": 0.20832185978045045,
+      "learning_rate": 1.1321108161212574e-06,
+      "loss": 0.0049,
+      "step": 2764
+    },
+    {
+      "epoch": 0.7884231536926147,
+      "grad_norm": 0.6598830632065913,
+      "learning_rate": 1.1291856344136853e-06,
+      "loss": 0.0072,
+      "step": 2765
+    },
+    {
+      "epoch": 0.7887082976903336,
+      "grad_norm": 1.6941306324384335,
+      "learning_rate": 1.1262637555556905e-06,
+      "loss": 0.0444,
+      "step": 2766
+    },
+    {
+      "epoch": 0.7889934416880525,
+      "grad_norm": 2.3563085514868853,
+      "learning_rate": 1.1233451820404222e-06,
+      "loss": 0.0641,
+      "step": 2767
+    },
+    {
+      "epoch": 0.7892785856857714,
+      "grad_norm": 0.5794506056340978,
+      "learning_rate": 1.1204299163582117e-06,
+      "loss": 0.0099,
+      "step": 2768
+    },
+    {
+      "epoch": 0.7895637296834902,
+      "grad_norm": 0.5242308434486851,
+      "learning_rate": 1.1175179609965697e-06,
+      "loss": 0.0082,
+      "step": 2769
+    },
+    {
+      "epoch": 0.789848873681209,
+      "grad_norm": 1.1466102878255973,
+      "learning_rate": 1.1146093184401791e-06,
+      "loss": 0.0196,
+      "step": 2770
+    },
+    {
+      "epoch": 0.7901340176789279,
+      "grad_norm": 1.8361430845752722,
+      "learning_rate": 1.1117039911708966e-06,
+      "loss": 0.039,
+      "step": 2771
+    },
+    {
+      "epoch": 0.7904191616766467,
+      "grad_norm": 2.1710130177501252,
+      "learning_rate": 1.108801981667752e-06,
+      "loss": 0.0378,
+      "step": 2772
+    },
+    {
+      "epoch": 0.7907043056743656,
+      "grad_norm": 1.0607481844235185,
+      "learning_rate": 1.1059032924069419e-06,
+      "loss": 0.0198,
+      "step": 2773
+    },
+    {
+      "epoch": 0.7909894496720844,
+      "grad_norm": 1.1534272328002837,
+      "learning_rate": 1.1030079258618303e-06,
+      "loss": 0.0181,
+      "step": 2774
+    },
+    {
+      "epoch": 0.7912745936698032,
+      "grad_norm": 1.4451323963964195,
+      "learning_rate": 1.1001158845029475e-06,
+      "loss": 0.028,
+      "step": 2775
+    },
+    {
+      "epoch": 0.7915597376675221,
+      "grad_norm": 1.0170979316582653,
+      "learning_rate": 1.0972271707979837e-06,
+      "loss": 0.0151,
+      "step": 2776
+    },
+    {
+      "epoch": 0.7918448816652409,
+      "grad_norm": 0.4474136512465243,
+      "learning_rate": 1.0943417872117956e-06,
+      "loss": 0.0113,
+      "step": 2777
+    },
+    {
+      "epoch": 0.7921300256629598,
+      "grad_norm": 1.0050859141496182,
+      "learning_rate": 1.09145973620639e-06,
+      "loss": 0.0114,
+      "step": 2778
+    },
+    {
+      "epoch": 0.7924151696606786,
+      "grad_norm": 1.7556272164199616,
+      "learning_rate": 1.0885810202409358e-06,
+      "loss": 0.0284,
+      "step": 2779
+    },
+    {
+      "epoch": 0.7927003136583974,
+      "grad_norm": 0.8890931457178195,
+      "learning_rate": 1.0857056417717538e-06,
+      "loss": 0.0171,
+      "step": 2780
+    },
+    {
+      "epoch": 0.7929854576561164,
+      "grad_norm": 0.9710326204173798,
+      "learning_rate": 1.0828336032523206e-06,
+      "loss": 0.0303,
+      "step": 2781
+    },
+    {
+      "epoch": 0.7932706016538352,
+      "grad_norm": 1.23379631524315,
+      "learning_rate": 1.0799649071332585e-06,
+      "loss": 0.0203,
+      "step": 2782
+    },
+    {
+      "epoch": 0.7935557456515541,
+      "grad_norm": 1.0048302038142245,
+      "learning_rate": 1.077099555862342e-06,
+      "loss": 0.0184,
+      "step": 2783
+    },
+    {
+      "epoch": 0.7938408896492729,
+      "grad_norm": 0.5781905231147105,
+      "learning_rate": 1.0742375518844845e-06,
+      "loss": 0.0065,
+      "step": 2784
+    },
+    {
+      "epoch": 0.7941260336469917,
+      "grad_norm": 0.6193101566716163,
+      "learning_rate": 1.0713788976417522e-06,
+      "loss": 0.0142,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7944111776447106,
+      "grad_norm": 0.4952614177724453,
+      "learning_rate": 1.068523595573348e-06,
+      "loss": 0.0058,
+      "step": 2786
+    },
+    {
+      "epoch": 0.7946963216424294,
+      "grad_norm": 1.2992089082042755,
+      "learning_rate": 1.0656716481156144e-06,
+      "loss": 0.0278,
+      "step": 2787
+    },
+    {
+      "epoch": 0.7949814656401483,
+      "grad_norm": 1.0260825841582102,
+      "learning_rate": 1.0628230577020327e-06,
+      "loss": 0.0125,
+      "step": 2788
+    },
+    {
+      "epoch": 0.7952666096378671,
+      "grad_norm": 1.650422933470823,
+      "learning_rate": 1.05997782676322e-06,
+      "loss": 0.0284,
+      "step": 2789
+    },
+    {
+      "epoch": 0.795551753635586,
+      "grad_norm": 1.6922243882463914,
+      "learning_rate": 1.0571359577269263e-06,
+      "loss": 0.034,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7958368976333048,
+      "grad_norm": 2.064635080746134,
+      "learning_rate": 1.0542974530180327e-06,
+      "loss": 0.0331,
+      "step": 2791
+    },
+    {
+      "epoch": 0.7961220416310236,
+      "grad_norm": 0.6979973531032883,
+      "learning_rate": 1.0514623150585484e-06,
+      "loss": 0.015,
+      "step": 2792
+    },
+    {
+      "epoch": 0.7964071856287425,
+      "grad_norm": 1.9320077547472618,
+      "learning_rate": 1.048630546267615e-06,
+      "loss": 0.048,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7966923296264614,
+      "grad_norm": 0.9851340654285308,
+      "learning_rate": 1.045802149061495e-06,
+      "loss": 0.0095,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7969774736241803,
+      "grad_norm": 0.33088819901168887,
+      "learning_rate": 1.0429771258535726e-06,
+      "loss": 0.007,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7972626176218991,
+      "grad_norm": 1.4719706011002405,
+      "learning_rate": 1.0401554790543545e-06,
+      "loss": 0.0243,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7975477616196179,
+      "grad_norm": 2.20284051649092,
+      "learning_rate": 1.0373372110714697e-06,
+      "loss": 0.0381,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7978329056173368,
+      "grad_norm": 0.9042831305808354,
+      "learning_rate": 1.0345223243096614e-06,
+      "loss": 0.0133,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7981180496150556,
+      "grad_norm": 0.4357457490944433,
+      "learning_rate": 1.0317108211707883e-06,
+      "loss": 0.0059,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7984031936127745,
+      "grad_norm": 1.4454010907254553,
+      "learning_rate": 1.0289027040538174e-06,
+      "loss": 0.0271,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7986883376104933,
+      "grad_norm": 0.8325294383532024,
+      "learning_rate": 1.0260979753548356e-06,
+      "loss": 0.0242,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7989734816082121,
+      "grad_norm": 0.6799031003016142,
+      "learning_rate": 1.023296637467031e-06,
+      "loss": 0.0115,
+      "step": 2802
+    },
+    {
+      "epoch": 0.799258625605931,
+      "grad_norm": 0.8000424760149701,
+      "learning_rate": 1.0204986927807026e-06,
+      "loss": 0.0139,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7995437696036498,
+      "grad_norm": 1.0437386516730733,
+      "learning_rate": 1.0177041436832508e-06,
+      "loss": 0.0234,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7998289136013687,
+      "grad_norm": 0.9328110888666781,
+      "learning_rate": 1.0149129925591816e-06,
+      "loss": 0.0411,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8001140575990875,
+      "grad_norm": 0.8919536940706596,
+      "learning_rate": 1.0121252417901e-06,
+      "loss": 0.0243,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8003992015968064,
+      "grad_norm": 1.7879385125082785,
+      "learning_rate": 1.00934089375471e-06,
+      "loss": 0.0299,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8006843455945253,
+      "grad_norm": 1.0080187097759719,
+      "learning_rate": 1.006559950828812e-06,
+      "loss": 0.0275,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8009694895922441,
+      "grad_norm": 2.297194259534309,
+      "learning_rate": 1.0037824153852993e-06,
+      "loss": 0.0589,
+      "step": 2809
+    },
+    {
+      "epoch": 0.801254633589963,
+      "grad_norm": 1.3722790217531777,
+      "learning_rate": 1.0010082897941642e-06,
+      "loss": 0.0468,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8015397775876818,
+      "grad_norm": 2.880885415554408,
+      "learning_rate": 9.982375764224805e-07,
+      "loss": 0.0544,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8018249215854006,
+      "grad_norm": 1.3779641066305213,
+      "learning_rate": 9.954702776344144e-07,
+      "loss": 0.0498,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8021100655831195,
+      "grad_norm": 1.578974235716462,
+      "learning_rate": 9.927063957912214e-07,
+      "loss": 0.0309,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8023952095808383,
+      "grad_norm": 0.5250722169025638,
+      "learning_rate": 9.89945933251238e-07,
+      "loss": 0.0073,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8026803535785572,
+      "grad_norm": 1.214875683960437,
+      "learning_rate": 9.871888923698836e-07,
+      "loss": 0.0313,
+      "step": 2815
+    },
+    {
+      "epoch": 0.802965497576276,
+      "grad_norm": 1.899479207342997,
+      "learning_rate": 9.844352754996578e-07,
+      "loss": 0.0394,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8032506415739948,
+      "grad_norm": 1.7956615761315469,
+      "learning_rate": 9.816850849901404e-07,
+      "loss": 0.037,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8035357855717137,
+      "grad_norm": 0.987617818563438,
+      "learning_rate": 9.78938323187985e-07,
+      "loss": 0.0249,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8038209295694325,
+      "grad_norm": 2.2873969718098177,
+      "learning_rate": 9.761949924369217e-07,
+      "loss": 0.0508,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8041060735671514,
+      "grad_norm": 0.6727560710264749,
+      "learning_rate": 9.734550950777523e-07,
+      "loss": 0.0135,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8043912175648703,
+      "grad_norm": 1.1083108655277412,
+      "learning_rate": 9.707186334483477e-07,
+      "loss": 0.0301,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8046763615625891,
+      "grad_norm": 1.4347478292304672,
+      "learning_rate": 9.679856098836498e-07,
+      "loss": 0.018,
+      "step": 2822
+    },
+    {
+      "epoch": 0.804961505560308,
+      "grad_norm": 0.7012755730521804,
+      "learning_rate": 9.652560267156647e-07,
+      "loss": 0.0096,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8052466495580268,
+      "grad_norm": 0.76616377946827,
+      "learning_rate": 9.625298862734645e-07,
+      "loss": 0.0136,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8055317935557457,
+      "grad_norm": 1.510172264857123,
+      "learning_rate": 9.598071908831802e-07,
+      "loss": 0.0371,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8058169375534645,
+      "grad_norm": 0.9247247630615653,
+      "learning_rate": 9.570879428680097e-07,
+      "loss": 0.0162,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8061020815511833,
+      "grad_norm": 1.4471204790194672,
+      "learning_rate": 9.543721445482051e-07,
+      "loss": 0.0414,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8063872255489022,
+      "grad_norm": 1.1261570935477987,
+      "learning_rate": 9.516597982410719e-07,
+      "loss": 0.0222,
+      "step": 2828
+    },
+    {
+      "epoch": 0.806672369546621,
+      "grad_norm": 1.213228339092222,
+      "learning_rate": 9.489509062609776e-07,
+      "loss": 0.019,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8069575135443399,
+      "grad_norm": 1.0568593386045195,
+      "learning_rate": 9.46245470919337e-07,
+      "loss": 0.0162,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8072426575420587,
+      "grad_norm": 1.0598129850467366,
+      "learning_rate": 9.435434945246164e-07,
+      "loss": 0.0115,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8075278015397775,
+      "grad_norm": 1.148087127959207,
+      "learning_rate": 9.408449793823316e-07,
+      "loss": 0.0186,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8078129455374964,
+      "grad_norm": 0.7838943287296171,
+      "learning_rate": 9.38149927795044e-07,
+      "loss": 0.0145,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8080980895352153,
+      "grad_norm": 0.6450557189362891,
+      "learning_rate": 9.354583420623603e-07,
+      "loss": 0.0178,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8083832335329342,
+      "grad_norm": 1.2580295549839888,
+      "learning_rate": 9.327702244809295e-07,
+      "loss": 0.0334,
+      "step": 2835
+    },
+    {
+      "epoch": 0.808668377530653,
+      "grad_norm": 0.5841231963670347,
+      "learning_rate": 9.300855773444422e-07,
+      "loss": 0.0171,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8089535215283719,
+      "grad_norm": 1.0744692555164113,
+      "learning_rate": 9.274044029436252e-07,
+      "loss": 0.0231,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8092386655260907,
+      "grad_norm": 0.5347627958333485,
+      "learning_rate": 9.247267035662483e-07,
+      "loss": 0.0106,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.8639389678174983,
+      "learning_rate": 9.220524814971082e-07,
+      "loss": 0.0098,
+      "step": 2839
+    },
+    {
+      "epoch": 0.8098089535215284,
+      "grad_norm": 1.4394556398547473,
+      "learning_rate": 9.193817390180387e-07,
+      "loss": 0.0158,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8100940975192472,
+      "grad_norm": 1.7003245750972533,
+      "learning_rate": 9.167144784079035e-07,
+      "loss": 0.0287,
+      "step": 2841
+    },
+    {
+      "epoch": 0.810379241516966,
+      "grad_norm": 1.3577296521728226,
+      "learning_rate": 9.140507019425981e-07,
+      "loss": 0.031,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8106643855146849,
+      "grad_norm": 1.3152852591455102,
+      "learning_rate": 9.11390411895044e-07,
+      "loss": 0.0271,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8109495295124037,
+      "grad_norm": 0.8196188314690228,
+      "learning_rate": 9.087336105351813e-07,
+      "loss": 0.0207,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8112346735101226,
+      "grad_norm": 0.7446020625868897,
+      "learning_rate": 9.060803001299833e-07,
+      "loss": 0.023,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8115198175078414,
+      "grad_norm": 1.6959863019895527,
+      "learning_rate": 9.034304829434387e-07,
+      "loss": 0.0425,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8118049615055604,
+      "grad_norm": 1.5565435765306315,
+      "learning_rate": 9.007841612365575e-07,
+      "loss": 0.0324,
+      "step": 2847
+    },
+    {
+      "epoch": 0.8120901055032792,
+      "grad_norm": 1.4501860535298694,
+      "learning_rate": 8.981413372673659e-07,
+      "loss": 0.0276,
+      "step": 2848
+    },
+    {
+      "epoch": 0.812375249500998,
+      "grad_norm": 0.8662824114962682,
+      "learning_rate": 8.955020132909059e-07,
+      "loss": 0.0222,
+      "step": 2849
+    },
+    {
+      "epoch": 0.8126603934987169,
+      "grad_norm": 0.6872747440077209,
+      "learning_rate": 8.928661915592335e-07,
+      "loss": 0.0064,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8129455374964357,
+      "grad_norm": 0.8899013455152293,
+      "learning_rate": 8.902338743214167e-07,
+      "loss": 0.0133,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8132306814941546,
+      "grad_norm": 0.75181554380744,
+      "learning_rate": 8.876050638235323e-07,
+      "loss": 0.0204,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8135158254918734,
+      "grad_norm": 0.830293688967669,
+      "learning_rate": 8.849797623086643e-07,
+      "loss": 0.011,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8138009694895922,
+      "grad_norm": 1.0573364540507733,
+      "learning_rate": 8.823579720169068e-07,
+      "loss": 0.038,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8140861134873111,
+      "grad_norm": 1.2831202813735518,
+      "learning_rate": 8.797396951853515e-07,
+      "loss": 0.0322,
+      "step": 2855
+    },
+    {
+      "epoch": 0.8143712574850299,
+      "grad_norm": 0.5543201708828293,
+      "learning_rate": 8.771249340480959e-07,
+      "loss": 0.0094,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8146564014827488,
+      "grad_norm": 0.9366333141939963,
+      "learning_rate": 8.745136908362367e-07,
+      "loss": 0.0124,
+      "step": 2857
+    },
+    {
+      "epoch": 0.8149415454804676,
+      "grad_norm": 1.1448966352215733,
+      "learning_rate": 8.719059677778712e-07,
+      "loss": 0.0348,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8152266894781864,
+      "grad_norm": 0.6801824833631003,
+      "learning_rate": 8.693017670980903e-07,
+      "loss": 0.0153,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8155118334759053,
+      "grad_norm": 1.490041583362656,
+      "learning_rate": 8.667010910189794e-07,
+      "loss": 0.039,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8157969774736242,
+      "grad_norm": 1.3198201360592343,
+      "learning_rate": 8.641039417596181e-07,
+      "loss": 0.0233,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8160821214713431,
+      "grad_norm": 0.8793603147206949,
+      "learning_rate": 8.615103215360754e-07,
+      "loss": 0.0206,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8163672654690619,
+      "grad_norm": 0.9461948476174028,
+      "learning_rate": 8.589202325614094e-07,
+      "loss": 0.0248,
+      "step": 2863
+    },
+    {
+      "epoch": 0.8166524094667807,
+      "grad_norm": 1.016498277100784,
+      "learning_rate": 8.563336770456654e-07,
+      "loss": 0.012,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8169375534644996,
+      "grad_norm": 2.102929530837856,
+      "learning_rate": 8.537506571958736e-07,
+      "loss": 0.0296,
+      "step": 2865
+    },
+    {
+      "epoch": 0.8172226974622184,
+      "grad_norm": 1.5361923525365948,
+      "learning_rate": 8.511711752160467e-07,
+      "loss": 0.0197,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8175078414599373,
+      "grad_norm": 1.2400610374990475,
+      "learning_rate": 8.485952333071801e-07,
+      "loss": 0.0212,
+      "step": 2867
+    },
+    {
+      "epoch": 0.8177929854576561,
+      "grad_norm": 0.9973748714157373,
+      "learning_rate": 8.460228336672466e-07,
+      "loss": 0.0389,
+      "step": 2868
+    },
+    {
+      "epoch": 0.818078129455375,
+      "grad_norm": 2.49728872513892,
+      "learning_rate": 8.434539784911966e-07,
+      "loss": 0.0497,
+      "step": 2869
+    },
+    {
+      "epoch": 0.8183632734530938,
+      "grad_norm": 1.3237326648218986,
+      "learning_rate": 8.408886699709601e-07,
+      "loss": 0.0351,
+      "step": 2870
+    },
+    {
+      "epoch": 0.8186484174508126,
+      "grad_norm": 0.8497672015331845,
+      "learning_rate": 8.383269102954367e-07,
+      "loss": 0.0264,
+      "step": 2871
+    },
+    {
+      "epoch": 0.8189335614485315,
+      "grad_norm": 1.065455590155432,
+      "learning_rate": 8.357687016504972e-07,
+      "loss": 0.0256,
+      "step": 2872
+    },
+    {
+      "epoch": 0.8192187054462503,
+      "grad_norm": 1.5358434738505333,
+      "learning_rate": 8.332140462189841e-07,
+      "loss": 0.0439,
+      "step": 2873
+    },
+    {
+      "epoch": 0.8195038494439693,
+      "grad_norm": 1.4294989218591867,
+      "learning_rate": 8.306629461807109e-07,
+      "loss": 0.0286,
+      "step": 2874
+    },
+    {
+      "epoch": 0.8197889934416881,
+      "grad_norm": 1.9130676592818923,
+      "learning_rate": 8.281154037124523e-07,
+      "loss": 0.0196,
+      "step": 2875
+    },
+    {
+      "epoch": 0.8200741374394069,
+      "grad_norm": 1.406613009250302,
+      "learning_rate": 8.255714209879506e-07,
+      "loss": 0.0262,
+      "step": 2876
+    },
+    {
+      "epoch": 0.8203592814371258,
+      "grad_norm": 0.6268363665660219,
+      "learning_rate": 8.230310001779096e-07,
+      "loss": 0.0121,
+      "step": 2877
+    },
+    {
+      "epoch": 0.8206444254348446,
+      "grad_norm": 0.7357882911831951,
+      "learning_rate": 8.204941434499941e-07,
+      "loss": 0.0099,
+      "step": 2878
+    },
+    {
+      "epoch": 0.8209295694325635,
+      "grad_norm": 1.8175431785227907,
+      "learning_rate": 8.179608529688276e-07,
+      "loss": 0.0249,
+      "step": 2879
+    },
+    {
+      "epoch": 0.8212147134302823,
+      "grad_norm": 0.2394768774671226,
+      "learning_rate": 8.154311308959911e-07,
+      "loss": 0.0064,
+      "step": 2880
+    },
+    {
+      "epoch": 0.8214998574280011,
+      "grad_norm": 1.0082037597372302,
+      "learning_rate": 8.129049793900185e-07,
+      "loss": 0.0152,
+      "step": 2881
+    },
+    {
+      "epoch": 0.82178500142572,
+      "grad_norm": 0.9206006167023392,
+      "learning_rate": 8.103824006064032e-07,
+      "loss": 0.0267,
+      "step": 2882
+    },
+    {
+      "epoch": 0.8220701454234388,
+      "grad_norm": 1.184257876582493,
+      "learning_rate": 8.078633966975818e-07,
+      "loss": 0.0344,
+      "step": 2883
+    },
+    {
+      "epoch": 0.8223552894211577,
+      "grad_norm": 1.119928137630302,
+      "learning_rate": 8.053479698129463e-07,
+      "loss": 0.021,
+      "step": 2884
+    },
+    {
+      "epoch": 0.8226404334188765,
+      "grad_norm": 0.7752796412552893,
+      "learning_rate": 8.028361220988334e-07,
+      "loss": 0.0217,
+      "step": 2885
+    },
+    {
+      "epoch": 0.8229255774165953,
+      "grad_norm": 1.0723505832855476,
+      "learning_rate": 8.003278556985295e-07,
+      "loss": 0.0404,
+      "step": 2886
+    },
+    {
+      "epoch": 0.8232107214143143,
+      "grad_norm": 0.370869296241954,
+      "learning_rate": 7.978231727522634e-07,
+      "loss": 0.0091,
+      "step": 2887
+    },
+    {
+      "epoch": 0.8234958654120331,
+      "grad_norm": 1.6501205644493508,
+      "learning_rate": 7.953220753972029e-07,
+      "loss": 0.0302,
+      "step": 2888
+    },
+    {
+      "epoch": 0.823781009409752,
+      "grad_norm": 3.387858235076237,
+      "learning_rate": 7.928245657674599e-07,
+      "loss": 0.0449,
+      "step": 2889
+    },
+    {
+      "epoch": 0.8240661534074708,
+      "grad_norm": 0.7939624399299788,
+      "learning_rate": 7.903306459940863e-07,
+      "loss": 0.0149,
+      "step": 2890
+    },
+    {
+      "epoch": 0.8243512974051896,
+      "grad_norm": 0.8919761379481936,
+      "learning_rate": 7.87840318205067e-07,
+      "loss": 0.0121,
+      "step": 2891
+    },
+    {
+      "epoch": 0.8246364414029085,
+      "grad_norm": 0.8945707327162529,
+      "learning_rate": 7.853535845253252e-07,
+      "loss": 0.0167,
+      "step": 2892
+    },
+    {
+      "epoch": 0.8249215854006273,
+      "grad_norm": 1.4960085975377995,
+      "learning_rate": 7.828704470767151e-07,
+      "loss": 0.0239,
+      "step": 2893
+    },
+    {
+      "epoch": 0.8252067293983462,
+      "grad_norm": 1.7361926971910235,
+      "learning_rate": 7.803909079780237e-07,
+      "loss": 0.0196,
+      "step": 2894
+    },
+    {
+      "epoch": 0.825491873396065,
+      "grad_norm": 2.1069882551766312,
+      "learning_rate": 7.779149693449666e-07,
+      "loss": 0.0331,
+      "step": 2895
+    },
+    {
+      "epoch": 0.8257770173937838,
+      "grad_norm": 1.640025015859863,
+      "learning_rate": 7.754426332901888e-07,
+      "loss": 0.0393,
+      "step": 2896
+    },
+    {
+      "epoch": 0.8260621613915027,
+      "grad_norm": 0.6030323002263968,
+      "learning_rate": 7.729739019232579e-07,
+      "loss": 0.0108,
+      "step": 2897
+    },
+    {
+      "epoch": 0.8263473053892215,
+      "grad_norm": 0.526741493347755,
+      "learning_rate": 7.705087773506731e-07,
+      "loss": 0.016,
+      "step": 2898
+    },
+    {
+      "epoch": 0.8266324493869404,
+      "grad_norm": 1.5742017652828122,
+      "learning_rate": 7.680472616758467e-07,
+      "loss": 0.0298,
+      "step": 2899
+    },
+    {
+      "epoch": 0.8269175933846592,
+      "grad_norm": 0.934331197910886,
+      "learning_rate": 7.655893569991175e-07,
+      "loss": 0.0105,
+      "step": 2900
+    },
+    {
+      "epoch": 0.8272027373823782,
+      "grad_norm": 0.7000170882506863,
+      "learning_rate": 7.631350654177405e-07,
+      "loss": 0.0158,
+      "step": 2901
+    },
+    {
+      "epoch": 0.827487881380097,
+      "grad_norm": 0.9680259614112673,
+      "learning_rate": 7.606843890258914e-07,
+      "loss": 0.0239,
+      "step": 2902
+    },
+    {
+      "epoch": 0.8277730253778158,
+      "grad_norm": 1.494423198330321,
+      "learning_rate": 7.582373299146578e-07,
+      "loss": 0.0392,
+      "step": 2903
+    },
+    {
+      "epoch": 0.8280581693755347,
+      "grad_norm": 1.207768362788276,
+      "learning_rate": 7.557938901720418e-07,
+      "loss": 0.0151,
+      "step": 2904
+    },
+    {
+      "epoch": 0.8283433133732535,
+      "grad_norm": 1.1032689271047176,
+      "learning_rate": 7.533540718829547e-07,
+      "loss": 0.0482,
+      "step": 2905
+    },
+    {
+      "epoch": 0.8286284573709723,
+      "grad_norm": 1.376001495335339,
+      "learning_rate": 7.509178771292231e-07,
+      "loss": 0.0188,
+      "step": 2906
+    },
+    {
+      "epoch": 0.8289136013686912,
+      "grad_norm": 1.0040653742355359,
+      "learning_rate": 7.484853079895782e-07,
+      "loss": 0.014,
+      "step": 2907
+    },
+    {
+      "epoch": 0.82919874536641,
+      "grad_norm": 0.853257302586088,
+      "learning_rate": 7.460563665396569e-07,
+      "loss": 0.019,
+      "step": 2908
+    },
+    {
+      "epoch": 0.8294838893641289,
+      "grad_norm": 0.41199696502392325,
+      "learning_rate": 7.436310548520037e-07,
+      "loss": 0.0078,
+      "step": 2909
+    },
+    {
+      "epoch": 0.8297690333618477,
+      "grad_norm": 1.344739178123292,
+      "learning_rate": 7.412093749960625e-07,
+      "loss": 0.0381,
+      "step": 2910
+    },
+    {
+      "epoch": 0.8300541773595665,
+      "grad_norm": 1.4175396586193671,
+      "learning_rate": 7.38791329038181e-07,
+      "loss": 0.0152,
+      "step": 2911
+    },
+    {
+      "epoch": 0.8303393213572854,
+      "grad_norm": 1.8019194230806457,
+      "learning_rate": 7.363769190416048e-07,
+      "loss": 0.0206,
+      "step": 2912
+    },
+    {
+      "epoch": 0.8306244653550042,
+      "grad_norm": 0.9576926416107516,
+      "learning_rate": 7.339661470664761e-07,
+      "loss": 0.0179,
+      "step": 2913
+    },
+    {
+      "epoch": 0.8309096093527232,
+      "grad_norm": 1.1324523780149793,
+      "learning_rate": 7.315590151698371e-07,
+      "loss": 0.0238,
+      "step": 2914
+    },
+    {
+      "epoch": 0.831194753350442,
+      "grad_norm": 1.4722177708236213,
+      "learning_rate": 7.291555254056198e-07,
+      "loss": 0.0186,
+      "step": 2915
+    },
+    {
+      "epoch": 0.8314798973481609,
+      "grad_norm": 0.9504375580728752,
+      "learning_rate": 7.267556798246489e-07,
+      "loss": 0.0433,
+      "step": 2916
+    },
+    {
+      "epoch": 0.8317650413458797,
+      "grad_norm": 1.3361157091698659,
+      "learning_rate": 7.243594804746401e-07,
+      "loss": 0.0271,
+      "step": 2917
+    },
+    {
+      "epoch": 0.8320501853435985,
+      "grad_norm": 1.5034432617659543,
+      "learning_rate": 7.219669294002002e-07,
+      "loss": 0.0161,
+      "step": 2918
+    },
+    {
+      "epoch": 0.8323353293413174,
+      "grad_norm": 0.8175025379771647,
+      "learning_rate": 7.195780286428206e-07,
+      "loss": 0.0126,
+      "step": 2919
+    },
+    {
+      "epoch": 0.8326204733390362,
+      "grad_norm": 0.8373585191851456,
+      "learning_rate": 7.17192780240879e-07,
+      "loss": 0.0164,
+      "step": 2920
+    },
+    {
+      "epoch": 0.832905617336755,
+      "grad_norm": 0.4003185300249988,
+      "learning_rate": 7.148111862296331e-07,
+      "loss": 0.0117,
+      "step": 2921
+    },
+    {
+      "epoch": 0.8331907613344739,
+      "grad_norm": 1.5025437383078868,
+      "learning_rate": 7.124332486412289e-07,
+      "loss": 0.0204,
+      "step": 2922
+    },
+    {
+      "epoch": 0.8334759053321927,
+      "grad_norm": 1.3047499112241643,
+      "learning_rate": 7.100589695046883e-07,
+      "loss": 0.022,
+      "step": 2923
+    },
+    {
+      "epoch": 0.8337610493299116,
+      "grad_norm": 0.568459371870428,
+      "learning_rate": 7.076883508459115e-07,
+      "loss": 0.0078,
+      "step": 2924
+    },
+    {
+      "epoch": 0.8340461933276304,
+      "grad_norm": 2.1528288915735145,
+      "learning_rate": 7.053213946876769e-07,
+      "loss": 0.0592,
+      "step": 2925
+    },
+    {
+      "epoch": 0.8343313373253493,
+      "grad_norm": 0.7861846035271016,
+      "learning_rate": 7.029581030496368e-07,
+      "loss": 0.025,
+      "step": 2926
+    },
+    {
+      "epoch": 0.8346164813230682,
+      "grad_norm": 2.3660338087684987,
+      "learning_rate": 7.005984779483166e-07,
+      "loss": 0.0348,
+      "step": 2927
+    },
+    {
+      "epoch": 0.834901625320787,
+      "grad_norm": 1.2003912939811823,
+      "learning_rate": 6.982425213971145e-07,
+      "loss": 0.0263,
+      "step": 2928
+    },
+    {
+      "epoch": 0.8351867693185059,
+      "grad_norm": 0.80495647903881,
+      "learning_rate": 6.958902354062952e-07,
+      "loss": 0.0153,
+      "step": 2929
+    },
+    {
+      "epoch": 0.8354719133162247,
+      "grad_norm": 0.9565190013742579,
+      "learning_rate": 6.93541621982996e-07,
+      "loss": 0.0163,
+      "step": 2930
+    },
+    {
+      "epoch": 0.8357570573139436,
+      "grad_norm": 1.5292130072943486,
+      "learning_rate": 6.911966831312189e-07,
+      "loss": 0.0256,
+      "step": 2931
+    },
+    {
+      "epoch": 0.8360422013116624,
+      "grad_norm": 1.3532017392811009,
+      "learning_rate": 6.888554208518272e-07,
+      "loss": 0.019,
+      "step": 2932
+    },
+    {
+      "epoch": 0.8363273453093812,
+      "grad_norm": 0.9094015042109889,
+      "learning_rate": 6.86517837142549e-07,
+      "loss": 0.015,
+      "step": 2933
+    },
+    {
+      "epoch": 0.8366124893071001,
+      "grad_norm": 0.7082302676502451,
+      "learning_rate": 6.841839339979778e-07,
+      "loss": 0.0187,
+      "step": 2934
+    },
+    {
+      "epoch": 0.8368976333048189,
+      "grad_norm": 0.9879617552325342,
+      "learning_rate": 6.818537134095604e-07,
+      "loss": 0.0201,
+      "step": 2935
+    },
+    {
+      "epoch": 0.8371827773025378,
+      "grad_norm": 0.564486638835549,
+      "learning_rate": 6.795271773656054e-07,
+      "loss": 0.0208,
+      "step": 2936
+    },
+    {
+      "epoch": 0.8374679213002566,
+      "grad_norm": 0.2814308779842693,
+      "learning_rate": 6.772043278512747e-07,
+      "loss": 0.0068,
+      "step": 2937
+    },
+    {
+      "epoch": 0.8377530652979754,
+      "grad_norm": 0.7178043054217914,
+      "learning_rate": 6.748851668485873e-07,
+      "loss": 0.0098,
+      "step": 2938
+    },
+    {
+      "epoch": 0.8380382092956943,
+      "grad_norm": 0.6805116484205672,
+      "learning_rate": 6.725696963364126e-07,
+      "loss": 0.0081,
+      "step": 2939
+    },
+    {
+      "epoch": 0.8383233532934131,
+      "grad_norm": 1.0022569621899031,
+      "learning_rate": 6.702579182904723e-07,
+      "loss": 0.0264,
+      "step": 2940
+    },
+    {
+      "epoch": 0.8386084972911321,
+      "grad_norm": 1.904193877543303,
+      "learning_rate": 6.679498346833374e-07,
+      "loss": 0.042,
+      "step": 2941
+    },
+    {
+      "epoch": 0.8388936412888509,
+      "grad_norm": 0.29653976115050246,
+      "learning_rate": 6.656454474844248e-07,
+      "loss": 0.0059,
+      "step": 2942
+    },
+    {
+      "epoch": 0.8391787852865698,
+      "grad_norm": 0.6705392516340656,
+      "learning_rate": 6.633447586600028e-07,
+      "loss": 0.0118,
+      "step": 2943
+    },
+    {
+      "epoch": 0.8394639292842886,
+      "grad_norm": 0.4684828232836064,
+      "learning_rate": 6.61047770173176e-07,
+      "loss": 0.0083,
+      "step": 2944
+    },
+    {
+      "epoch": 0.8397490732820074,
+      "grad_norm": 0.6894065679425363,
+      "learning_rate": 6.587544839838961e-07,
+      "loss": 0.0139,
+      "step": 2945
+    },
+    {
+      "epoch": 0.8400342172797263,
+      "grad_norm": 1.1770716615384926,
+      "learning_rate": 6.564649020489566e-07,
+      "loss": 0.0188,
+      "step": 2946
+    },
+    {
+      "epoch": 0.8403193612774451,
+      "grad_norm": 0.8594451557210869,
+      "learning_rate": 6.541790263219894e-07,
+      "loss": 0.0171,
+      "step": 2947
+    },
+    {
+      "epoch": 0.840604505275164,
+      "grad_norm": 0.9031875486057582,
+      "learning_rate": 6.518968587534635e-07,
+      "loss": 0.025,
+      "step": 2948
+    },
+    {
+      "epoch": 0.8408896492728828,
+      "grad_norm": 1.9432753386854618,
+      "learning_rate": 6.496184012906814e-07,
+      "loss": 0.0232,
+      "step": 2949
+    },
+    {
+      "epoch": 0.8411747932706016,
+      "grad_norm": 0.7020884836813761,
+      "learning_rate": 6.473436558777846e-07,
+      "loss": 0.0086,
+      "step": 2950
+    },
+    {
+      "epoch": 0.8414599372683205,
+      "grad_norm": 0.5466929368926949,
+      "learning_rate": 6.450726244557448e-07,
+      "loss": 0.012,
+      "step": 2951
+    },
+    {
+      "epoch": 0.8417450812660393,
+      "grad_norm": 1.7507269669821028,
+      "learning_rate": 6.428053089623648e-07,
+      "loss": 0.0178,
+      "step": 2952
+    },
+    {
+      "epoch": 0.8420302252637581,
+      "grad_norm": 1.4983504947040904,
+      "learning_rate": 6.405417113322765e-07,
+      "loss": 0.0259,
+      "step": 2953
+    },
+    {
+      "epoch": 0.8423153692614771,
+      "grad_norm": 0.6649231360890827,
+      "learning_rate": 6.382818334969399e-07,
+      "loss": 0.0152,
+      "step": 2954
+    },
+    {
+      "epoch": 0.8426005132591959,
+      "grad_norm": 0.5291504974490259,
+      "learning_rate": 6.360256773846402e-07,
+      "loss": 0.0093,
+      "step": 2955
+    },
+    {
+      "epoch": 0.8428856572569148,
+      "grad_norm": 1.0922105163879545,
+      "learning_rate": 6.337732449204886e-07,
+      "loss": 0.0149,
+      "step": 2956
+    },
+    {
+      "epoch": 0.8431708012546336,
+      "grad_norm": 1.0162775087631601,
+      "learning_rate": 6.315245380264179e-07,
+      "loss": 0.0238,
+      "step": 2957
+    },
+    {
+      "epoch": 0.8434559452523525,
+      "grad_norm": 1.035067751129879,
+      "learning_rate": 6.292795586211803e-07,
+      "loss": 0.0184,
+      "step": 2958
+    },
+    {
+      "epoch": 0.8437410892500713,
+      "grad_norm": 0.4314581076678392,
+      "learning_rate": 6.27038308620353e-07,
+      "loss": 0.0069,
+      "step": 2959
+    },
+    {
+      "epoch": 0.8440262332477901,
+      "grad_norm": 0.9172564461045714,
+      "learning_rate": 6.24800789936324e-07,
+      "loss": 0.013,
+      "step": 2960
+    },
+    {
+      "epoch": 0.844311377245509,
+      "grad_norm": 1.8064356661478254,
+      "learning_rate": 6.225670044783011e-07,
+      "loss": 0.0285,
+      "step": 2961
+    },
+    {
+      "epoch": 0.8445965212432278,
+      "grad_norm": 0.9846608581771469,
+      "learning_rate": 6.203369541523075e-07,
+      "loss": 0.0181,
+      "step": 2962
+    },
+    {
+      "epoch": 0.8448816652409467,
+      "grad_norm": 0.5341057881948537,
+      "learning_rate": 6.181106408611781e-07,
+      "loss": 0.0092,
+      "step": 2963
+    },
+    {
+      "epoch": 0.8451668092386655,
+      "grad_norm": 0.5404376217996557,
+      "learning_rate": 6.158880665045586e-07,
+      "loss": 0.0104,
+      "step": 2964
+    },
+    {
+      "epoch": 0.8454519532363843,
+      "grad_norm": 0.5547122276300936,
+      "learning_rate": 6.136692329789046e-07,
+      "loss": 0.0138,
+      "step": 2965
+    },
+    {
+      "epoch": 0.8457370972341032,
+      "grad_norm": 0.870237694120721,
+      "learning_rate": 6.114541421774811e-07,
+      "loss": 0.0194,
+      "step": 2966
+    },
+    {
+      "epoch": 0.8460222412318221,
+      "grad_norm": 1.0347448612702808,
+      "learning_rate": 6.092427959903574e-07,
+      "loss": 0.0283,
+      "step": 2967
+    },
+    {
+      "epoch": 0.846307385229541,
+      "grad_norm": 1.4250265517501355,
+      "learning_rate": 6.070351963044091e-07,
+      "loss": 0.0508,
+      "step": 2968
+    },
+    {
+      "epoch": 0.8465925292272598,
+      "grad_norm": 0.34595863279716216,
+      "learning_rate": 6.04831345003315e-07,
+      "loss": 0.0062,
+      "step": 2969
+    },
+    {
+      "epoch": 0.8468776732249786,
+      "grad_norm": 1.2934086285223172,
+      "learning_rate": 6.026312439675553e-07,
+      "loss": 0.0239,
+      "step": 2970
+    },
+    {
+      "epoch": 0.8471628172226975,
+      "grad_norm": 1.1738506805733755,
+      "learning_rate": 6.004348950744094e-07,
+      "loss": 0.0134,
+      "step": 2971
+    },
+    {
+      "epoch": 0.8474479612204163,
+      "grad_norm": 0.9407356078115766,
+      "learning_rate": 5.982423001979559e-07,
+      "loss": 0.0395,
+      "step": 2972
+    },
+    {
+      "epoch": 0.8477331052181352,
+      "grad_norm": 1.899663507801717,
+      "learning_rate": 5.960534612090707e-07,
+      "loss": 0.029,
+      "step": 2973
+    },
+    {
+      "epoch": 0.848018249215854,
+      "grad_norm": 1.635086087665949,
+      "learning_rate": 5.93868379975423e-07,
+      "loss": 0.0204,
+      "step": 2974
+    },
+    {
+      "epoch": 0.8483033932135728,
+      "grad_norm": 2.7131462813485427,
+      "learning_rate": 5.916870583614792e-07,
+      "loss": 0.0511,
+      "step": 2975
+    },
+    {
+      "epoch": 0.8485885372112917,
+      "grad_norm": 0.6000044831511491,
+      "learning_rate": 5.895094982284949e-07,
+      "loss": 0.0151,
+      "step": 2976
+    },
+    {
+      "epoch": 0.8488736812090105,
+      "grad_norm": 0.8428158736884612,
+      "learning_rate": 5.873357014345143e-07,
+      "loss": 0.0275,
+      "step": 2977
+    },
+    {
+      "epoch": 0.8491588252067294,
+      "grad_norm": 1.1200799726027786,
+      "learning_rate": 5.851656698343761e-07,
+      "loss": 0.0311,
+      "step": 2978
+    },
+    {
+      "epoch": 0.8494439692044482,
+      "grad_norm": 0.7513879350557346,
+      "learning_rate": 5.829994052797011e-07,
+      "loss": 0.0189,
+      "step": 2979
+    },
+    {
+      "epoch": 0.849729113202167,
+      "grad_norm": 0.5097265708096947,
+      "learning_rate": 5.808369096188981e-07,
+      "loss": 0.0101,
+      "step": 2980
+    },
+    {
+      "epoch": 0.850014257199886,
+      "grad_norm": 1.5460511983229916,
+      "learning_rate": 5.786781846971601e-07,
+      "loss": 0.0304,
+      "step": 2981
+    },
+    {
+      "epoch": 0.8502994011976048,
+      "grad_norm": 1.677115356095477,
+      "learning_rate": 5.765232323564617e-07,
+      "loss": 0.0635,
+      "step": 2982
+    },
+    {
+      "epoch": 0.8505845451953237,
+      "grad_norm": 0.880229931432272,
+      "learning_rate": 5.743720544355597e-07,
+      "loss": 0.0363,
+      "step": 2983
+    },
+    {
+      "epoch": 0.8508696891930425,
+      "grad_norm": 0.5325984514944866,
+      "learning_rate": 5.722246527699887e-07,
+      "loss": 0.0049,
+      "step": 2984
+    },
+    {
+      "epoch": 0.8511548331907614,
+      "grad_norm": 1.031023684922083,
+      "learning_rate": 5.700810291920628e-07,
+      "loss": 0.0192,
+      "step": 2985
+    },
+    {
+      "epoch": 0.8514399771884802,
+      "grad_norm": 0.8501593340127255,
+      "learning_rate": 5.679411855308697e-07,
+      "loss": 0.0173,
+      "step": 2986
+    },
+    {
+      "epoch": 0.851725121186199,
+      "grad_norm": 1.2342740400840233,
+      "learning_rate": 5.658051236122774e-07,
+      "loss": 0.023,
+      "step": 2987
+    },
+    {
+      "epoch": 0.8520102651839179,
+      "grad_norm": 1.060959958106904,
+      "learning_rate": 5.636728452589196e-07,
+      "loss": 0.0154,
+      "step": 2988
+    },
+    {
+      "epoch": 0.8522954091816367,
+      "grad_norm": 1.2602365805850364,
+      "learning_rate": 5.615443522902076e-07,
+      "loss": 0.0231,
+      "step": 2989
+    },
+    {
+      "epoch": 0.8525805531793555,
+      "grad_norm": 0.764129932845986,
+      "learning_rate": 5.594196465223184e-07,
+      "loss": 0.0094,
+      "step": 2990
+    },
+    {
+      "epoch": 0.8528656971770744,
+      "grad_norm": 1.5099273606359904,
+      "learning_rate": 5.57298729768202e-07,
+      "loss": 0.0173,
+      "step": 2991
+    },
+    {
+      "epoch": 0.8531508411747932,
+      "grad_norm": 0.9820827550126608,
+      "learning_rate": 5.551816038375729e-07,
+      "loss": 0.0134,
+      "step": 2992
+    },
+    {
+      "epoch": 0.8534359851725121,
+      "grad_norm": 1.1727880595649633,
+      "learning_rate": 5.530682705369084e-07,
+      "loss": 0.0258,
+      "step": 2993
+    },
+    {
+      "epoch": 0.853721129170231,
+      "grad_norm": 1.0065318180810632,
+      "learning_rate": 5.509587316694537e-07,
+      "loss": 0.0207,
+      "step": 2994
+    },
+    {
+      "epoch": 0.8540062731679499,
+      "grad_norm": 0.7848127018539507,
+      "learning_rate": 5.488529890352157e-07,
+      "loss": 0.0083,
+      "step": 2995
+    },
+    {
+      "epoch": 0.8542914171656687,
+      "grad_norm": 1.0156083213820275,
+      "learning_rate": 5.467510444309609e-07,
+      "loss": 0.0119,
+      "step": 2996
+    },
+    {
+      "epoch": 0.8545765611633875,
+      "grad_norm": 1.0197812302503872,
+      "learning_rate": 5.446528996502149e-07,
+      "loss": 0.0128,
+      "step": 2997
+    },
+    {
+      "epoch": 0.8548617051611064,
+      "grad_norm": 0.8285221051674888,
+      "learning_rate": 5.425585564832625e-07,
+      "loss": 0.0093,
+      "step": 2998
+    },
+    {
+      "epoch": 0.8551468491588252,
+      "grad_norm": 0.8350102395268476,
+      "learning_rate": 5.404680167171427e-07,
+      "loss": 0.0209,
+      "step": 2999
+    },
+    {
+      "epoch": 0.8554319931565441,
+      "grad_norm": 0.8296267711722471,
+      "learning_rate": 5.38381282135651e-07,
+      "loss": 0.0103,
+      "step": 3000
+    },
+    {
+      "epoch": 0.8557171371542629,
+      "grad_norm": 1.1354258345442707,
+      "learning_rate": 5.362983545193351e-07,
+      "loss": 0.0132,
+      "step": 3001
+    },
+    {
+      "epoch": 0.8560022811519817,
+      "grad_norm": 1.3807885834714135,
+      "learning_rate": 5.34219235645494e-07,
+      "loss": 0.0184,
+      "step": 3002
+    },
+    {
+      "epoch": 0.8562874251497006,
+      "grad_norm": 0.4618296825629643,
+      "learning_rate": 5.321439272881795e-07,
+      "loss": 0.0101,
+      "step": 3003
+    },
+    {
+      "epoch": 0.8565725691474194,
+      "grad_norm": 0.7977559295975786,
+      "learning_rate": 5.300724312181876e-07,
+      "loss": 0.0149,
+      "step": 3004
+    },
+    {
+      "epoch": 0.8568577131451383,
+      "grad_norm": 0.5816427329433224,
+      "learning_rate": 5.280047492030638e-07,
+      "loss": 0.0177,
+      "step": 3005
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.47258730273410254,
+      "learning_rate": 5.259408830070989e-07,
+      "loss": 0.0071,
+      "step": 3006
+    },
+    {
+      "epoch": 0.857428001140576,
+      "grad_norm": 0.7070970759261074,
+      "learning_rate": 5.238808343913299e-07,
+      "loss": 0.0247,
+      "step": 3007
+    },
+    {
+      "epoch": 0.8577131451382949,
+      "grad_norm": 0.6869352056980774,
+      "learning_rate": 5.218246051135323e-07,
+      "loss": 0.0137,
+      "step": 3008
+    },
+    {
+      "epoch": 0.8579982891360137,
+      "grad_norm": 1.0038883944832988,
+      "learning_rate": 5.197721969282271e-07,
+      "loss": 0.0257,
+      "step": 3009
+    },
+    {
+      "epoch": 0.8582834331337326,
+      "grad_norm": 1.2068882204622162,
+      "learning_rate": 5.177236115866685e-07,
+      "loss": 0.0209,
+      "step": 3010
+    },
+    {
+      "epoch": 0.8585685771314514,
+      "grad_norm": 0.9311944031546174,
+      "learning_rate": 5.156788508368565e-07,
+      "loss": 0.0231,
+      "step": 3011
+    },
+    {
+      "epoch": 0.8588537211291702,
+      "grad_norm": 1.5397233224377365,
+      "learning_rate": 5.136379164235222e-07,
+      "loss": 0.0241,
+      "step": 3012
+    },
+    {
+      "epoch": 0.8591388651268891,
+      "grad_norm": 1.6349055154960528,
+      "learning_rate": 5.116008100881348e-07,
+      "loss": 0.0272,
+      "step": 3013
+    },
+    {
+      "epoch": 0.8594240091246079,
+      "grad_norm": 1.4003109679024432,
+      "learning_rate": 5.095675335688949e-07,
+      "loss": 0.0391,
+      "step": 3014
+    },
+    {
+      "epoch": 0.8597091531223268,
+      "grad_norm": 0.6121106764703516,
+      "learning_rate": 5.075380886007369e-07,
+      "loss": 0.0158,
+      "step": 3015
+    },
+    {
+      "epoch": 0.8599942971200456,
+      "grad_norm": 1.343535205043131,
+      "learning_rate": 5.055124769153247e-07,
+      "loss": 0.0401,
+      "step": 3016
+    },
+    {
+      "epoch": 0.8602794411177644,
+      "grad_norm": 0.3419700286934388,
+      "learning_rate": 5.034907002410517e-07,
+      "loss": 0.0048,
+      "step": 3017
+    },
+    {
+      "epoch": 0.8605645851154833,
+      "grad_norm": 0.9895620726753768,
+      "learning_rate": 5.014727603030389e-07,
+      "loss": 0.0221,
+      "step": 3018
+    },
+    {
+      "epoch": 0.8608497291132021,
+      "grad_norm": 1.3614488033863745,
+      "learning_rate": 4.994586588231354e-07,
+      "loss": 0.0183,
+      "step": 3019
+    },
+    {
+      "epoch": 0.861134873110921,
+      "grad_norm": 1.7974814733359572,
+      "learning_rate": 4.974483975199135e-07,
+      "loss": 0.053,
+      "step": 3020
+    },
+    {
+      "epoch": 0.8614200171086399,
+      "grad_norm": 0.6222206952249488,
+      "learning_rate": 4.954419781086672e-07,
+      "loss": 0.0126,
+      "step": 3021
+    },
+    {
+      "epoch": 0.8617051611063588,
+      "grad_norm": 0.5890050784706143,
+      "learning_rate": 4.934394023014133e-07,
+      "loss": 0.0097,
+      "step": 3022
+    },
+    {
+      "epoch": 0.8619903051040776,
+      "grad_norm": 0.2498934201504258,
+      "learning_rate": 4.914406718068914e-07,
+      "loss": 0.0037,
+      "step": 3023
+    },
+    {
+      "epoch": 0.8622754491017964,
+      "grad_norm": 1.370346135075879,
+      "learning_rate": 4.894457883305576e-07,
+      "loss": 0.0189,
+      "step": 3024
+    },
+    {
+      "epoch": 0.8625605930995153,
+      "grad_norm": 0.35639843396867693,
+      "learning_rate": 4.874547535745872e-07,
+      "loss": 0.0055,
+      "step": 3025
+    },
+    {
+      "epoch": 0.8628457370972341,
+      "grad_norm": 1.6416261884833374,
+      "learning_rate": 4.854675692378669e-07,
+      "loss": 0.0296,
+      "step": 3026
+    },
+    {
+      "epoch": 0.863130881094953,
+      "grad_norm": 1.1136701347202882,
+      "learning_rate": 4.834842370160048e-07,
+      "loss": 0.0164,
+      "step": 3027
+    },
+    {
+      "epoch": 0.8634160250926718,
+      "grad_norm": 1.368275106996081,
+      "learning_rate": 4.815047586013172e-07,
+      "loss": 0.0336,
+      "step": 3028
+    },
+    {
+      "epoch": 0.8637011690903906,
+      "grad_norm": 1.2575862109268776,
+      "learning_rate": 4.795291356828335e-07,
+      "loss": 0.0228,
+      "step": 3029
+    },
+    {
+      "epoch": 0.8639863130881095,
+      "grad_norm": 0.6138345409799086,
+      "learning_rate": 4.775573699462926e-07,
+      "loss": 0.0098,
+      "step": 3030
+    },
+    {
+      "epoch": 0.8642714570858283,
+      "grad_norm": 2.4032353134017397,
+      "learning_rate": 4.7558946307414645e-07,
+      "loss": 0.0509,
+      "step": 3031
+    },
+    {
+      "epoch": 0.8645566010835471,
+      "grad_norm": 1.3870212191516793,
+      "learning_rate": 4.736254167455473e-07,
+      "loss": 0.0358,
+      "step": 3032
+    },
+    {
+      "epoch": 0.864841745081266,
+      "grad_norm": 1.4321688335485183,
+      "learning_rate": 4.7166523263635743e-07,
+      "loss": 0.0353,
+      "step": 3033
+    },
+    {
+      "epoch": 0.8651268890789849,
+      "grad_norm": 0.452148177107619,
+      "learning_rate": 4.697089124191434e-07,
+      "loss": 0.0071,
+      "step": 3034
+    },
+    {
+      "epoch": 0.8654120330767038,
+      "grad_norm": 2.022511254722525,
+      "learning_rate": 4.677564577631749e-07,
+      "loss": 0.0269,
+      "step": 3035
+    },
+    {
+      "epoch": 0.8656971770744226,
+      "grad_norm": 1.1264957765805845,
+      "learning_rate": 4.6580787033442376e-07,
+      "loss": 0.0252,
+      "step": 3036
+    },
+    {
+      "epoch": 0.8659823210721415,
+      "grad_norm": 0.9989388112218847,
+      "learning_rate": 4.6386315179555883e-07,
+      "loss": 0.0241,
+      "step": 3037
+    },
+    {
+      "epoch": 0.8662674650698603,
+      "grad_norm": 1.9978868244980388,
+      "learning_rate": 4.6192230380595004e-07,
+      "loss": 0.0216,
+      "step": 3038
+    },
+    {
+      "epoch": 0.8665526090675791,
+      "grad_norm": 0.4630351383686262,
+      "learning_rate": 4.599853280216665e-07,
+      "loss": 0.0103,
+      "step": 3039
+    },
+    {
+      "epoch": 0.866837753065298,
+      "grad_norm": 0.7910478687505361,
+      "learning_rate": 4.580522260954706e-07,
+      "loss": 0.0133,
+      "step": 3040
+    },
+    {
+      "epoch": 0.8671228970630168,
+      "grad_norm": 0.905830864917694,
+      "learning_rate": 4.561229996768196e-07,
+      "loss": 0.0294,
+      "step": 3041
+    },
+    {
+      "epoch": 0.8674080410607357,
+      "grad_norm": 0.8213970821241439,
+      "learning_rate": 4.5419765041186556e-07,
+      "loss": 0.0219,
+      "step": 3042
+    },
+    {
+      "epoch": 0.8676931850584545,
+      "grad_norm": 1.1748421184539621,
+      "learning_rate": 4.5227617994345053e-07,
+      "loss": 0.0163,
+      "step": 3043
+    },
+    {
+      "epoch": 0.8679783290561733,
+      "grad_norm": 0.43672133470671437,
+      "learning_rate": 4.503585899111068e-07,
+      "loss": 0.007,
+      "step": 3044
+    },
+    {
+      "epoch": 0.8682634730538922,
+      "grad_norm": 2.0611265738885183,
+      "learning_rate": 4.4844488195105784e-07,
+      "loss": 0.048,
+      "step": 3045
+    },
+    {
+      "epoch": 0.868548617051611,
+      "grad_norm": 1.3462808320153008,
+      "learning_rate": 4.4653505769621073e-07,
+      "loss": 0.0322,
+      "step": 3046
+    },
+    {
+      "epoch": 0.86883376104933,
+      "grad_norm": 1.1019872968537088,
+      "learning_rate": 4.446291187761648e-07,
+      "loss": 0.0257,
+      "step": 3047
+    },
+    {
+      "epoch": 0.8691189050470488,
+      "grad_norm": 0.4768401486598245,
+      "learning_rate": 4.4272706681719737e-07,
+      "loss": 0.0138,
+      "step": 3048
+    },
+    {
+      "epoch": 0.8694040490447676,
+      "grad_norm": 1.2680969131683553,
+      "learning_rate": 4.408289034422736e-07,
+      "loss": 0.035,
+      "step": 3049
+    },
+    {
+      "epoch": 0.8696891930424865,
+      "grad_norm": 1.0552008198715335,
+      "learning_rate": 4.3893463027103735e-07,
+      "loss": 0.0191,
+      "step": 3050
+    },
+    {
+      "epoch": 0.8699743370402053,
+      "grad_norm": 0.7865993113772903,
+      "learning_rate": 4.370442489198179e-07,
+      "loss": 0.017,
+      "step": 3051
+    },
+    {
+      "epoch": 0.8702594810379242,
+      "grad_norm": 0.7490127620354204,
+      "learning_rate": 4.351577610016189e-07,
+      "loss": 0.0132,
+      "step": 3052
+    },
+    {
+      "epoch": 0.870544625035643,
+      "grad_norm": 0.5028891048262714,
+      "learning_rate": 4.3327516812612545e-07,
+      "loss": 0.0068,
+      "step": 3053
+    },
+    {
+      "epoch": 0.8708297690333618,
+      "grad_norm": 0.7876344772007207,
+      "learning_rate": 4.313964718996949e-07,
+      "loss": 0.0195,
+      "step": 3054
+    },
+    {
+      "epoch": 0.8711149130310807,
+      "grad_norm": 0.46302658918296347,
+      "learning_rate": 4.2952167392536436e-07,
+      "loss": 0.0081,
+      "step": 3055
+    },
+    {
+      "epoch": 0.8714000570287995,
+      "grad_norm": 0.5289145118332526,
+      "learning_rate": 4.2765077580284197e-07,
+      "loss": 0.0065,
+      "step": 3056
+    },
+    {
+      "epoch": 0.8716852010265184,
+      "grad_norm": 1.5234422334108315,
+      "learning_rate": 4.257837791285091e-07,
+      "loss": 0.016,
+      "step": 3057
+    },
+    {
+      "epoch": 0.8719703450242372,
+      "grad_norm": 1.6244487998081059,
+      "learning_rate": 4.2392068549541755e-07,
+      "loss": 0.0346,
+      "step": 3058
+    },
+    {
+      "epoch": 0.872255489021956,
+      "grad_norm": 1.141957719991174,
+      "learning_rate": 4.22061496493289e-07,
+      "loss": 0.0178,
+      "step": 3059
+    },
+    {
+      "epoch": 0.8725406330196749,
+      "grad_norm": 1.6278771393473568,
+      "learning_rate": 4.202062137085139e-07,
+      "loss": 0.0276,
+      "step": 3060
+    },
+    {
+      "epoch": 0.8728257770173938,
+      "grad_norm": 0.514051654602654,
+      "learning_rate": 4.183548387241498e-07,
+      "loss": 0.0086,
+      "step": 3061
+    },
+    {
+      "epoch": 0.8731109210151127,
+      "grad_norm": 0.8895694324937317,
+      "learning_rate": 4.1650737311991765e-07,
+      "loss": 0.0163,
+      "step": 3062
+    },
+    {
+      "epoch": 0.8733960650128315,
+      "grad_norm": 1.7832475192370916,
+      "learning_rate": 4.146638184722057e-07,
+      "loss": 0.0443,
+      "step": 3063
+    },
+    {
+      "epoch": 0.8736812090105504,
+      "grad_norm": 1.1145459448952413,
+      "learning_rate": 4.1282417635406525e-07,
+      "loss": 0.0241,
+      "step": 3064
+    },
+    {
+      "epoch": 0.8739663530082692,
+      "grad_norm": 1.3530250524727738,
+      "learning_rate": 4.109884483352045e-07,
+      "loss": 0.0204,
+      "step": 3065
+    },
+    {
+      "epoch": 0.874251497005988,
+      "grad_norm": 1.0947124737811844,
+      "learning_rate": 4.091566359819954e-07,
+      "loss": 0.0418,
+      "step": 3066
+    },
+    {
+      "epoch": 0.8745366410037069,
+      "grad_norm": 0.9920153082074292,
+      "learning_rate": 4.0732874085747053e-07,
+      "loss": 0.0183,
+      "step": 3067
+    },
+    {
+      "epoch": 0.8748217850014257,
+      "grad_norm": 1.065157293408351,
+      "learning_rate": 4.055047645213167e-07,
+      "loss": 0.0162,
+      "step": 3068
+    },
+    {
+      "epoch": 0.8751069289991446,
+      "grad_norm": 0.8965691490104757,
+      "learning_rate": 4.0368470852987893e-07,
+      "loss": 0.024,
+      "step": 3069
+    },
+    {
+      "epoch": 0.8753920729968634,
+      "grad_norm": 2.346622705630544,
+      "learning_rate": 4.018685744361539e-07,
+      "loss": 0.0497,
+      "step": 3070
+    },
+    {
+      "epoch": 0.8756772169945822,
+      "grad_norm": 0.9532499289760871,
+      "learning_rate": 4.000563637897964e-07,
+      "loss": 0.0121,
+      "step": 3071
+    },
+    {
+      "epoch": 0.8759623609923011,
+      "grad_norm": 0.4104488966130429,
+      "learning_rate": 3.982480781371106e-07,
+      "loss": 0.0082,
+      "step": 3072
+    },
+    {
+      "epoch": 0.8762475049900199,
+      "grad_norm": 1.5437624866770545,
+      "learning_rate": 3.9644371902105296e-07,
+      "loss": 0.0588,
+      "step": 3073
+    },
+    {
+      "epoch": 0.8765326489877389,
+      "grad_norm": 1.0598825987060185,
+      "learning_rate": 3.9464328798122843e-07,
+      "loss": 0.011,
+      "step": 3074
+    },
+    {
+      "epoch": 0.8768177929854577,
+      "grad_norm": 0.7745631171488191,
+      "learning_rate": 3.928467865538904e-07,
+      "loss": 0.0277,
+      "step": 3075
+    },
+    {
+      "epoch": 0.8771029369831765,
+      "grad_norm": 1.1461659365443477,
+      "learning_rate": 3.910542162719394e-07,
+      "loss": 0.0379,
+      "step": 3076
+    },
+    {
+      "epoch": 0.8773880809808954,
+      "grad_norm": 1.0689248460984624,
+      "learning_rate": 3.8926557866492297e-07,
+      "loss": 0.0189,
+      "step": 3077
+    },
+    {
+      "epoch": 0.8776732249786142,
+      "grad_norm": 1.7797425568134084,
+      "learning_rate": 3.874808752590298e-07,
+      "loss": 0.0512,
+      "step": 3078
+    },
+    {
+      "epoch": 0.8779583689763331,
+      "grad_norm": 1.313282350888063,
+      "learning_rate": 3.8570010757709555e-07,
+      "loss": 0.0285,
+      "step": 3079
+    },
+    {
+      "epoch": 0.8782435129740519,
+      "grad_norm": 1.3537466884000247,
+      "learning_rate": 3.8392327713859644e-07,
+      "loss": 0.0192,
+      "step": 3080
+    },
+    {
+      "epoch": 0.8785286569717707,
+      "grad_norm": 1.1739115495562644,
+      "learning_rate": 3.821503854596459e-07,
+      "loss": 0.0154,
+      "step": 3081
+    },
+    {
+      "epoch": 0.8788138009694896,
+      "grad_norm": 0.6978895204370438,
+      "learning_rate": 3.803814340529999e-07,
+      "loss": 0.0205,
+      "step": 3082
+    },
+    {
+      "epoch": 0.8790989449672084,
+      "grad_norm": 2.472278902571292,
+      "learning_rate": 3.786164244280532e-07,
+      "loss": 0.0816,
+      "step": 3083
+    },
+    {
+      "epoch": 0.8793840889649273,
+      "grad_norm": 0.42124346849130617,
+      "learning_rate": 3.7685535809083406e-07,
+      "loss": 0.0061,
+      "step": 3084
+    },
+    {
+      "epoch": 0.8796692329626461,
+      "grad_norm": 0.641144962594694,
+      "learning_rate": 3.7509823654400757e-07,
+      "loss": 0.0065,
+      "step": 3085
+    },
+    {
+      "epoch": 0.8799543769603649,
+      "grad_norm": 0.9745992756194799,
+      "learning_rate": 3.7334506128687277e-07,
+      "loss": 0.0127,
+      "step": 3086
+    },
+    {
+      "epoch": 0.8802395209580839,
+      "grad_norm": 1.0045168845028045,
+      "learning_rate": 3.715958338153619e-07,
+      "loss": 0.0247,
+      "step": 3087
+    },
+    {
+      "epoch": 0.8805246649558027,
+      "grad_norm": 0.49384156816707714,
+      "learning_rate": 3.698505556220372e-07,
+      "loss": 0.0134,
+      "step": 3088
+    },
+    {
+      "epoch": 0.8808098089535216,
+      "grad_norm": 1.153521953839027,
+      "learning_rate": 3.681092281960935e-07,
+      "loss": 0.0098,
+      "step": 3089
+    },
+    {
+      "epoch": 0.8810949529512404,
+      "grad_norm": 1.7322788165810779,
+      "learning_rate": 3.6637185302335234e-07,
+      "loss": 0.035,
+      "step": 3090
+    },
+    {
+      "epoch": 0.8813800969489592,
+      "grad_norm": 1.129612121303509,
+      "learning_rate": 3.646384315862633e-07,
+      "loss": 0.0126,
+      "step": 3091
+    },
+    {
+      "epoch": 0.8816652409466781,
+      "grad_norm": 0.602950178845579,
+      "learning_rate": 3.629089653639034e-07,
+      "loss": 0.0165,
+      "step": 3092
+    },
+    {
+      "epoch": 0.8819503849443969,
+      "grad_norm": 0.6273478691828218,
+      "learning_rate": 3.6118345583197434e-07,
+      "loss": 0.0087,
+      "step": 3093
+    },
+    {
+      "epoch": 0.8822355289421158,
+      "grad_norm": 0.7988486703644966,
+      "learning_rate": 3.594619044628017e-07,
+      "loss": 0.0235,
+      "step": 3094
+    },
+    {
+      "epoch": 0.8825206729398346,
+      "grad_norm": 1.1404611832652243,
+      "learning_rate": 3.577443127253316e-07,
+      "loss": 0.0208,
+      "step": 3095
+    },
+    {
+      "epoch": 0.8828058169375534,
+      "grad_norm": 0.673777489436252,
+      "learning_rate": 3.5603068208513616e-07,
+      "loss": 0.0139,
+      "step": 3096
+    },
+    {
+      "epoch": 0.8830909609352723,
+      "grad_norm": 1.0996498578604454,
+      "learning_rate": 3.5432101400440456e-07,
+      "loss": 0.0188,
+      "step": 3097
+    },
+    {
+      "epoch": 0.8833761049329911,
+      "grad_norm": 1.7534824016132662,
+      "learning_rate": 3.526153099419427e-07,
+      "loss": 0.0332,
+      "step": 3098
+    },
+    {
+      "epoch": 0.88366124893071,
+      "grad_norm": 1.5098939443616157,
+      "learning_rate": 3.5091357135317917e-07,
+      "loss": 0.0269,
+      "step": 3099
+    },
+    {
+      "epoch": 0.8839463929284288,
+      "grad_norm": 0.8049820887786194,
+      "learning_rate": 3.492157996901552e-07,
+      "loss": 0.0082,
+      "step": 3100
+    },
+    {
+      "epoch": 0.8842315369261478,
+      "grad_norm": 0.9473485933582722,
+      "learning_rate": 3.4752199640152916e-07,
+      "loss": 0.0215,
+      "step": 3101
+    },
+    {
+      "epoch": 0.8845166809238666,
+      "grad_norm": 1.2883061493145709,
+      "learning_rate": 3.458321629325717e-07,
+      "loss": 0.0306,
+      "step": 3102
+    },
+    {
+      "epoch": 0.8848018249215854,
+      "grad_norm": 0.9187716869904959,
+      "learning_rate": 3.441463007251672e-07,
+      "loss": 0.0132,
+      "step": 3103
+    },
+    {
+      "epoch": 0.8850869689193043,
+      "grad_norm": 0.6349246018941566,
+      "learning_rate": 3.4246441121781105e-07,
+      "loss": 0.0059,
+      "step": 3104
+    },
+    {
+      "epoch": 0.8853721129170231,
+      "grad_norm": 1.3681391815563542,
+      "learning_rate": 3.407864958456092e-07,
+      "loss": 0.016,
+      "step": 3105
+    },
+    {
+      "epoch": 0.885657256914742,
+      "grad_norm": 1.3040792173489308,
+      "learning_rate": 3.391125560402764e-07,
+      "loss": 0.0424,
+      "step": 3106
+    },
+    {
+      "epoch": 0.8859424009124608,
+      "grad_norm": 1.2333440427099385,
+      "learning_rate": 3.3744259323013453e-07,
+      "loss": 0.0154,
+      "step": 3107
+    },
+    {
+      "epoch": 0.8862275449101796,
+      "grad_norm": 1.150041429446506,
+      "learning_rate": 3.357766088401149e-07,
+      "loss": 0.0355,
+      "step": 3108
+    },
+    {
+      "epoch": 0.8865126889078985,
+      "grad_norm": 0.5551063042416896,
+      "learning_rate": 3.3411460429174927e-07,
+      "loss": 0.0093,
+      "step": 3109
+    },
+    {
+      "epoch": 0.8867978329056173,
+      "grad_norm": 1.2252348783292581,
+      "learning_rate": 3.324565810031777e-07,
+      "loss": 0.0247,
+      "step": 3110
+    },
+    {
+      "epoch": 0.8870829769033362,
+      "grad_norm": 0.7969957705002801,
+      "learning_rate": 3.3080254038914014e-07,
+      "loss": 0.0128,
+      "step": 3111
+    },
+    {
+      "epoch": 0.887368120901055,
+      "grad_norm": 0.5694148413447736,
+      "learning_rate": 3.2915248386098163e-07,
+      "loss": 0.0049,
+      "step": 3112
+    },
+    {
+      "epoch": 0.8876532648987738,
+      "grad_norm": 0.5807428864056374,
+      "learning_rate": 3.275064128266453e-07,
+      "loss": 0.0108,
+      "step": 3113
+    },
+    {
+      "epoch": 0.8879384088964928,
+      "grad_norm": 0.9409478396226437,
+      "learning_rate": 3.2586432869067263e-07,
+      "loss": 0.0291,
+      "step": 3114
+    },
+    {
+      "epoch": 0.8882235528942116,
+      "grad_norm": 1.9926399501252572,
+      "learning_rate": 3.242262328542067e-07,
+      "loss": 0.0305,
+      "step": 3115
+    },
+    {
+      "epoch": 0.8885086968919305,
+      "grad_norm": 0.9596467871013998,
+      "learning_rate": 3.225921267149845e-07,
+      "loss": 0.0175,
+      "step": 3116
+    },
+    {
+      "epoch": 0.8887938408896493,
+      "grad_norm": 1.307498365238962,
+      "learning_rate": 3.2096201166734007e-07,
+      "loss": 0.0352,
+      "step": 3117
+    },
+    {
+      "epoch": 0.8890789848873681,
+      "grad_norm": 1.5303253426806933,
+      "learning_rate": 3.193358891022008e-07,
+      "loss": 0.0146,
+      "step": 3118
+    },
+    {
+      "epoch": 0.889364128885087,
+      "grad_norm": 0.33325411716229314,
+      "learning_rate": 3.17713760407089e-07,
+      "loss": 0.0056,
+      "step": 3119
+    },
+    {
+      "epoch": 0.8896492728828058,
+      "grad_norm": 1.538780520413921,
+      "learning_rate": 3.160956269661175e-07,
+      "loss": 0.0188,
+      "step": 3120
+    },
+    {
+      "epoch": 0.8899344168805247,
+      "grad_norm": 0.7446784766283835,
+      "learning_rate": 3.1448149015999187e-07,
+      "loss": 0.013,
+      "step": 3121
+    },
+    {
+      "epoch": 0.8902195608782435,
+      "grad_norm": 1.3141671766830856,
+      "learning_rate": 3.1287135136600643e-07,
+      "loss": 0.0254,
+      "step": 3122
+    },
+    {
+      "epoch": 0.8905047048759623,
+      "grad_norm": 1.4567278490282125,
+      "learning_rate": 3.112652119580428e-07,
+      "loss": 0.0329,
+      "step": 3123
+    },
+    {
+      "epoch": 0.8907898488736812,
+      "grad_norm": 1.6857368705721907,
+      "learning_rate": 3.096630733065742e-07,
+      "loss": 0.0288,
+      "step": 3124
+    },
+    {
+      "epoch": 0.8910749928714,
+      "grad_norm": 2.0692748305620747,
+      "learning_rate": 3.0806493677865534e-07,
+      "loss": 0.0278,
+      "step": 3125
+    },
+    {
+      "epoch": 0.8913601368691189,
+      "grad_norm": 0.7567155805940998,
+      "learning_rate": 3.0647080373792824e-07,
+      "loss": 0.0153,
+      "step": 3126
+    },
+    {
+      "epoch": 0.8916452808668378,
+      "grad_norm": 2.5094863114875863,
+      "learning_rate": 3.048806755446182e-07,
+      "loss": 0.0432,
+      "step": 3127
+    },
+    {
+      "epoch": 0.8919304248645566,
+      "grad_norm": 1.2661236090125079,
+      "learning_rate": 3.032945535555354e-07,
+      "loss": 0.0616,
+      "step": 3128
+    },
+    {
+      "epoch": 0.8922155688622755,
+      "grad_norm": 0.377347821334773,
+      "learning_rate": 3.0171243912406944e-07,
+      "loss": 0.0064,
+      "step": 3129
+    },
+    {
+      "epoch": 0.8925007128599943,
+      "grad_norm": 0.7116646851338002,
+      "learning_rate": 3.0013433360019105e-07,
+      "loss": 0.0234,
+      "step": 3130
+    },
+    {
+      "epoch": 0.8927858568577132,
+      "grad_norm": 1.280990787410855,
+      "learning_rate": 2.9856023833045033e-07,
+      "loss": 0.0218,
+      "step": 3131
+    },
+    {
+      "epoch": 0.893071000855432,
+      "grad_norm": 0.7586342697071281,
+      "learning_rate": 2.969901546579751e-07,
+      "loss": 0.0211,
+      "step": 3132
+    },
+    {
+      "epoch": 0.8933561448531508,
+      "grad_norm": 0.8495441832875039,
+      "learning_rate": 2.9542408392247036e-07,
+      "loss": 0.0106,
+      "step": 3133
+    },
+    {
+      "epoch": 0.8936412888508697,
+      "grad_norm": 1.7650449817458465,
+      "learning_rate": 2.9386202746021773e-07,
+      "loss": 0.0278,
+      "step": 3134
+    },
+    {
+      "epoch": 0.8939264328485885,
+      "grad_norm": 0.7986477079444706,
+      "learning_rate": 2.9230398660407277e-07,
+      "loss": 0.0136,
+      "step": 3135
+    },
+    {
+      "epoch": 0.8942115768463074,
+      "grad_norm": 0.8463514580140247,
+      "learning_rate": 2.9074996268346533e-07,
+      "loss": 0.0252,
+      "step": 3136
+    },
+    {
+      "epoch": 0.8944967208440262,
+      "grad_norm": 1.3930496907278234,
+      "learning_rate": 2.8919995702439696e-07,
+      "loss": 0.0329,
+      "step": 3137
+    },
+    {
+      "epoch": 0.894781864841745,
+      "grad_norm": 1.0464784961095484,
+      "learning_rate": 2.8765397094944083e-07,
+      "loss": 0.0249,
+      "step": 3138
+    },
+    {
+      "epoch": 0.8950670088394639,
+      "grad_norm": 1.3983247885259522,
+      "learning_rate": 2.8611200577774e-07,
+      "loss": 0.0182,
+      "step": 3139
+    },
+    {
+      "epoch": 0.8953521528371827,
+      "grad_norm": 1.0011589985184415,
+      "learning_rate": 2.845740628250082e-07,
+      "loss": 0.0155,
+      "step": 3140
+    },
+    {
+      "epoch": 0.8956372968349017,
+      "grad_norm": 0.41127888337629276,
+      "learning_rate": 2.8304014340352625e-07,
+      "loss": 0.0091,
+      "step": 3141
+    },
+    {
+      "epoch": 0.8959224408326205,
+      "grad_norm": 0.7008237115353383,
+      "learning_rate": 2.815102488221394e-07,
+      "loss": 0.0092,
+      "step": 3142
+    },
+    {
+      "epoch": 0.8962075848303394,
+      "grad_norm": 1.6731644282839946,
+      "learning_rate": 2.7998438038626174e-07,
+      "loss": 0.0176,
+      "step": 3143
+    },
+    {
+      "epoch": 0.8964927288280582,
+      "grad_norm": 1.606249319903414,
+      "learning_rate": 2.7846253939787125e-07,
+      "loss": 0.0415,
+      "step": 3144
+    },
+    {
+      "epoch": 0.896777872825777,
+      "grad_norm": 1.300144651734939,
+      "learning_rate": 2.7694472715550925e-07,
+      "loss": 0.0182,
+      "step": 3145
+    },
+    {
+      "epoch": 0.8970630168234959,
+      "grad_norm": 0.8102382687014947,
+      "learning_rate": 2.7543094495427913e-07,
+      "loss": 0.0164,
+      "step": 3146
+    },
+    {
+      "epoch": 0.8973481608212147,
+      "grad_norm": 0.4796368067100617,
+      "learning_rate": 2.7392119408584493e-07,
+      "loss": 0.0071,
+      "step": 3147
+    },
+    {
+      "epoch": 0.8976333048189336,
+      "grad_norm": 0.4708253859579334,
+      "learning_rate": 2.7241547583843286e-07,
+      "loss": 0.0096,
+      "step": 3148
+    },
+    {
+      "epoch": 0.8979184488166524,
+      "grad_norm": 1.6685451122274635,
+      "learning_rate": 2.7091379149682683e-07,
+      "loss": 0.0186,
+      "step": 3149
+    },
+    {
+      "epoch": 0.8982035928143712,
+      "grad_norm": 1.3143101566649746,
+      "learning_rate": 2.6941614234236857e-07,
+      "loss": 0.0192,
+      "step": 3150
+    },
+    {
+      "epoch": 0.8984887368120901,
+      "grad_norm": 0.6686035692072104,
+      "learning_rate": 2.679225296529564e-07,
+      "loss": 0.0135,
+      "step": 3151
+    },
+    {
+      "epoch": 0.8987738808098089,
+      "grad_norm": 0.8841163748576049,
+      "learning_rate": 2.664329547030475e-07,
+      "loss": 0.0315,
+      "step": 3152
+    },
+    {
+      "epoch": 0.8990590248075278,
+      "grad_norm": 0.9004178197360575,
+      "learning_rate": 2.649474187636492e-07,
+      "loss": 0.0154,
+      "step": 3153
+    },
+    {
+      "epoch": 0.8993441688052467,
+      "grad_norm": 2.2456953621383815,
+      "learning_rate": 2.634659231023251e-07,
+      "loss": 0.0293,
+      "step": 3154
+    },
+    {
+      "epoch": 0.8996293128029655,
+      "grad_norm": 0.6666588261244653,
+      "learning_rate": 2.619884689831909e-07,
+      "loss": 0.0109,
+      "step": 3155
+    },
+    {
+      "epoch": 0.8999144568006844,
+      "grad_norm": 0.8776064374516797,
+      "learning_rate": 2.6051505766691464e-07,
+      "loss": 0.0144,
+      "step": 3156
+    },
+    {
+      "epoch": 0.9001996007984032,
+      "grad_norm": 1.1268349745247128,
+      "learning_rate": 2.5904569041071417e-07,
+      "loss": 0.0245,
+      "step": 3157
+    },
+    {
+      "epoch": 0.9004847447961221,
+      "grad_norm": 0.4970618468994337,
+      "learning_rate": 2.5758036846835476e-07,
+      "loss": 0.0081,
+      "step": 3158
+    },
+    {
+      "epoch": 0.9007698887938409,
+      "grad_norm": 1.0343525333492938,
+      "learning_rate": 2.561190930901519e-07,
+      "loss": 0.0217,
+      "step": 3159
+    },
+    {
+      "epoch": 0.9010550327915597,
+      "grad_norm": 1.1597997219850231,
+      "learning_rate": 2.5466186552296933e-07,
+      "loss": 0.029,
+      "step": 3160
+    },
+    {
+      "epoch": 0.9013401767892786,
+      "grad_norm": 0.5518918587988524,
+      "learning_rate": 2.532086870102146e-07,
+      "loss": 0.0107,
+      "step": 3161
+    },
+    {
+      "epoch": 0.9016253207869974,
+      "grad_norm": 1.555015490765001,
+      "learning_rate": 2.5175955879184146e-07,
+      "loss": 0.0201,
+      "step": 3162
+    },
+    {
+      "epoch": 0.9019104647847163,
+      "grad_norm": 0.6795609328872417,
+      "learning_rate": 2.503144821043474e-07,
+      "loss": 0.0142,
+      "step": 3163
+    },
+    {
+      "epoch": 0.9021956087824351,
+      "grad_norm": 0.6884525201103593,
+      "learning_rate": 2.488734581807728e-07,
+      "loss": 0.0185,
+      "step": 3164
+    },
+    {
+      "epoch": 0.9024807527801539,
+      "grad_norm": 0.5375639386409782,
+      "learning_rate": 2.474364882507002e-07,
+      "loss": 0.0059,
+      "step": 3165
+    },
+    {
+      "epoch": 0.9027658967778728,
+      "grad_norm": 0.5080847121580953,
+      "learning_rate": 2.4600357354025275e-07,
+      "loss": 0.0072,
+      "step": 3166
+    },
+    {
+      "epoch": 0.9030510407755917,
+      "grad_norm": 0.806467203465794,
+      "learning_rate": 2.4457471527209343e-07,
+      "loss": 0.0246,
+      "step": 3167
+    },
+    {
+      "epoch": 0.9033361847733106,
+      "grad_norm": 1.700259776983858,
+      "learning_rate": 2.431499146654243e-07,
+      "loss": 0.0195,
+      "step": 3168
+    },
+    {
+      "epoch": 0.9036213287710294,
+      "grad_norm": 1.066316674315371,
+      "learning_rate": 2.4172917293598607e-07,
+      "loss": 0.0332,
+      "step": 3169
+    },
+    {
+      "epoch": 0.9039064727687482,
+      "grad_norm": 1.2640803766957995,
+      "learning_rate": 2.4031249129605305e-07,
+      "loss": 0.0148,
+      "step": 3170
+    },
+    {
+      "epoch": 0.9041916167664671,
+      "grad_norm": 2.356961541868472,
+      "learning_rate": 2.3889987095443657e-07,
+      "loss": 0.0305,
+      "step": 3171
+    },
+    {
+      "epoch": 0.9044767607641859,
+      "grad_norm": 0.8215136443424736,
+      "learning_rate": 2.3749131311648576e-07,
+      "loss": 0.0111,
+      "step": 3172
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 1.21999981493346,
+      "learning_rate": 2.360868189840787e-07,
+      "loss": 0.0162,
+      "step": 3173
+    },
+    {
+      "epoch": 0.9050470487596236,
+      "grad_norm": 1.955946628179618,
+      "learning_rate": 2.346863897556295e-07,
+      "loss": 0.033,
+      "step": 3174
+    },
+    {
+      "epoch": 0.9053321927573424,
+      "grad_norm": 1.3460408883251782,
+      "learning_rate": 2.3329002662608068e-07,
+      "loss": 0.0461,
+      "step": 3175
+    },
+    {
+      "epoch": 0.9056173367550613,
+      "grad_norm": 1.1009867392611814,
+      "learning_rate": 2.3189773078690748e-07,
+      "loss": 0.0119,
+      "step": 3176
+    },
+    {
+      "epoch": 0.9059024807527801,
+      "grad_norm": 0.6976449459445861,
+      "learning_rate": 2.305095034261151e-07,
+      "loss": 0.0104,
+      "step": 3177
+    },
+    {
+      "epoch": 0.906187624750499,
+      "grad_norm": 1.050109166261638,
+      "learning_rate": 2.2912534572823498e-07,
+      "loss": 0.0226,
+      "step": 3178
+    },
+    {
+      "epoch": 0.9064727687482178,
+      "grad_norm": 0.6801788595908712,
+      "learning_rate": 2.2774525887432786e-07,
+      "loss": 0.0185,
+      "step": 3179
+    },
+    {
+      "epoch": 0.9067579127459366,
+      "grad_norm": 0.6869437524292872,
+      "learning_rate": 2.2636924404198014e-07,
+      "loss": 0.0104,
+      "step": 3180
+    },
+    {
+      "epoch": 0.9070430567436556,
+      "grad_norm": 1.4278818202476977,
+      "learning_rate": 2.2499730240530426e-07,
+      "loss": 0.0239,
+      "step": 3181
+    },
+    {
+      "epoch": 0.9073282007413744,
+      "grad_norm": 0.6426812316033088,
+      "learning_rate": 2.2362943513493662e-07,
+      "loss": 0.0164,
+      "step": 3182
+    },
+    {
+      "epoch": 0.9076133447390933,
+      "grad_norm": 0.7907561871363531,
+      "learning_rate": 2.2226564339803636e-07,
+      "loss": 0.0096,
+      "step": 3183
+    },
+    {
+      "epoch": 0.9078984887368121,
+      "grad_norm": 0.9117928880337626,
+      "learning_rate": 2.2090592835828817e-07,
+      "loss": 0.017,
+      "step": 3184
+    },
+    {
+      "epoch": 0.908183632734531,
+      "grad_norm": 0.9775102371730449,
+      "learning_rate": 2.1955029117589454e-07,
+      "loss": 0.0145,
+      "step": 3185
+    },
+    {
+      "epoch": 0.9084687767322498,
+      "grad_norm": 0.6325452457433002,
+      "learning_rate": 2.1819873300758022e-07,
+      "loss": 0.0088,
+      "step": 3186
+    },
+    {
+      "epoch": 0.9087539207299686,
+      "grad_norm": 0.8958197615264045,
+      "learning_rate": 2.168512550065882e-07,
+      "loss": 0.0312,
+      "step": 3187
+    },
+    {
+      "epoch": 0.9090390647276875,
+      "grad_norm": 0.6725945959693137,
+      "learning_rate": 2.1550785832268217e-07,
+      "loss": 0.0147,
+      "step": 3188
+    },
+    {
+      "epoch": 0.9093242087254063,
+      "grad_norm": 0.7200310122010458,
+      "learning_rate": 2.1416854410214183e-07,
+      "loss": 0.009,
+      "step": 3189
+    },
+    {
+      "epoch": 0.9096093527231252,
+      "grad_norm": 1.3177205015746256,
+      "learning_rate": 2.1283331348776414e-07,
+      "loss": 0.0207,
+      "step": 3190
+    },
+    {
+      "epoch": 0.909894496720844,
+      "grad_norm": 1.6289098120134233,
+      "learning_rate": 2.115021676188611e-07,
+      "loss": 0.0245,
+      "step": 3191
+    },
+    {
+      "epoch": 0.9101796407185628,
+      "grad_norm": 1.9024020760670197,
+      "learning_rate": 2.101751076312586e-07,
+      "loss": 0.0367,
+      "step": 3192
+    },
+    {
+      "epoch": 0.9104647847162817,
+      "grad_norm": 1.1389262118067653,
+      "learning_rate": 2.0885213465729802e-07,
+      "loss": 0.0238,
+      "step": 3193
+    },
+    {
+      "epoch": 0.9107499287140006,
+      "grad_norm": 2.0058186055618106,
+      "learning_rate": 2.0753324982583202e-07,
+      "loss": 0.0419,
+      "step": 3194
+    },
+    {
+      "epoch": 0.9110350727117195,
+      "grad_norm": 1.2169233475560073,
+      "learning_rate": 2.0621845426222587e-07,
+      "loss": 0.0326,
+      "step": 3195
+    },
+    {
+      "epoch": 0.9113202167094383,
+      "grad_norm": 0.3754420977930853,
+      "learning_rate": 2.0490774908835442e-07,
+      "loss": 0.0057,
+      "step": 3196
+    },
+    {
+      "epoch": 0.9116053607071571,
+      "grad_norm": 2.673609902784827,
+      "learning_rate": 2.0360113542260307e-07,
+      "loss": 0.0349,
+      "step": 3197
+    },
+    {
+      "epoch": 0.911890504704876,
+      "grad_norm": 1.0286839819383613,
+      "learning_rate": 2.0229861437986665e-07,
+      "loss": 0.0309,
+      "step": 3198
+    },
+    {
+      "epoch": 0.9121756487025948,
+      "grad_norm": 1.100305443067917,
+      "learning_rate": 2.0100018707154612e-07,
+      "loss": 0.0127,
+      "step": 3199
+    },
+    {
+      "epoch": 0.9124607927003137,
+      "grad_norm": 0.9752425027943912,
+      "learning_rate": 1.9970585460555193e-07,
+      "loss": 0.0114,
+      "step": 3200
+    },
+    {
+      "epoch": 0.9127459366980325,
+      "grad_norm": 1.5804332245975838,
+      "learning_rate": 1.984156180862984e-07,
+      "loss": 0.036,
+      "step": 3201
+    },
+    {
+      "epoch": 0.9130310806957513,
+      "grad_norm": 1.2405391166369548,
+      "learning_rate": 1.97129478614706e-07,
+      "loss": 0.0198,
+      "step": 3202
+    },
+    {
+      "epoch": 0.9133162246934702,
+      "grad_norm": 1.4988667995021159,
+      "learning_rate": 1.958474372881969e-07,
+      "loss": 0.0165,
+      "step": 3203
+    },
+    {
+      "epoch": 0.913601368691189,
+      "grad_norm": 2.6758434050413924,
+      "learning_rate": 1.94569495200701e-07,
+      "loss": 0.0473,
+      "step": 3204
+    },
+    {
+      "epoch": 0.9138865126889079,
+      "grad_norm": 0.37648975473173746,
+      "learning_rate": 1.9329565344264666e-07,
+      "loss": 0.009,
+      "step": 3205
+    },
+    {
+      "epoch": 0.9141716566866267,
+      "grad_norm": 1.4270330071786936,
+      "learning_rate": 1.9202591310096495e-07,
+      "loss": 0.0318,
+      "step": 3206
+    },
+    {
+      "epoch": 0.9144568006843456,
+      "grad_norm": 1.1284641851442696,
+      "learning_rate": 1.9076027525908704e-07,
+      "loss": 0.0169,
+      "step": 3207
+    },
+    {
+      "epoch": 0.9147419446820645,
+      "grad_norm": 0.22940010951660872,
+      "learning_rate": 1.8949874099694344e-07,
+      "loss": 0.0054,
+      "step": 3208
+    },
+    {
+      "epoch": 0.9150270886797833,
+      "grad_norm": 0.472117646501281,
+      "learning_rate": 1.8824131139096424e-07,
+      "loss": 0.013,
+      "step": 3209
+    },
+    {
+      "epoch": 0.9153122326775022,
+      "grad_norm": 1.0145391952231864,
+      "learning_rate": 1.8698798751407566e-07,
+      "loss": 0.0163,
+      "step": 3210
+    },
+    {
+      "epoch": 0.915597376675221,
+      "grad_norm": 1.6566652751095565,
+      "learning_rate": 1.8573877043570166e-07,
+      "loss": 0.0293,
+      "step": 3211
+    },
+    {
+      "epoch": 0.9158825206729398,
+      "grad_norm": 0.8730480588302432,
+      "learning_rate": 1.8449366122176072e-07,
+      "loss": 0.0211,
+      "step": 3212
+    },
+    {
+      "epoch": 0.9161676646706587,
+      "grad_norm": 0.762722802208298,
+      "learning_rate": 1.8325266093466908e-07,
+      "loss": 0.0167,
+      "step": 3213
+    },
+    {
+      "epoch": 0.9164528086683775,
+      "grad_norm": 1.8519175215823354,
+      "learning_rate": 1.820157706333331e-07,
+      "loss": 0.042,
+      "step": 3214
+    },
+    {
+      "epoch": 0.9167379526660964,
+      "grad_norm": 0.6178345429669876,
+      "learning_rate": 1.8078299137315457e-07,
+      "loss": 0.0115,
+      "step": 3215
+    },
+    {
+      "epoch": 0.9170230966638152,
+      "grad_norm": 1.8389032255626088,
+      "learning_rate": 1.7955432420602714e-07,
+      "loss": 0.0449,
+      "step": 3216
+    },
+    {
+      "epoch": 0.917308240661534,
+      "grad_norm": 0.8976456881308423,
+      "learning_rate": 1.7832977018033604e-07,
+      "loss": 0.0146,
+      "step": 3217
+    },
+    {
+      "epoch": 0.9175933846592529,
+      "grad_norm": 1.1641398023584142,
+      "learning_rate": 1.7710933034095658e-07,
+      "loss": 0.0167,
+      "step": 3218
+    },
+    {
+      "epoch": 0.9178785286569717,
+      "grad_norm": 0.34100280170072866,
+      "learning_rate": 1.7589300572925184e-07,
+      "loss": 0.0044,
+      "step": 3219
+    },
+    {
+      "epoch": 0.9181636726546906,
+      "grad_norm": 1.7683542198735762,
+      "learning_rate": 1.7468079738307608e-07,
+      "loss": 0.0254,
+      "step": 3220
+    },
+    {
+      "epoch": 0.9184488166524095,
+      "grad_norm": 0.9779660998455367,
+      "learning_rate": 1.7347270633677082e-07,
+      "loss": 0.0247,
+      "step": 3221
+    },
+    {
+      "epoch": 0.9187339606501284,
+      "grad_norm": 0.7604628137905527,
+      "learning_rate": 1.722687336211626e-07,
+      "loss": 0.016,
+      "step": 3222
+    },
+    {
+      "epoch": 0.9190191046478472,
+      "grad_norm": 1.9616843025886015,
+      "learning_rate": 1.7106888026356626e-07,
+      "loss": 0.0218,
+      "step": 3223
+    },
+    {
+      "epoch": 0.919304248645566,
+      "grad_norm": 1.4389747508794408,
+      "learning_rate": 1.6987314728778014e-07,
+      "loss": 0.1148,
+      "step": 3224
+    },
+    {
+      "epoch": 0.9195893926432849,
+      "grad_norm": 1.3027980674511903,
+      "learning_rate": 1.6868153571408695e-07,
+      "loss": 0.0248,
+      "step": 3225
+    },
+    {
+      "epoch": 0.9198745366410037,
+      "grad_norm": 0.4343153104128488,
+      "learning_rate": 1.6749404655925338e-07,
+      "loss": 0.009,
+      "step": 3226
+    },
+    {
+      "epoch": 0.9201596806387226,
+      "grad_norm": 0.47044929809354835,
+      "learning_rate": 1.663106808365289e-07,
+      "loss": 0.0077,
+      "step": 3227
+    },
+    {
+      "epoch": 0.9204448246364414,
+      "grad_norm": 1.7245884546851775,
+      "learning_rate": 1.6513143955564192e-07,
+      "loss": 0.0386,
+      "step": 3228
+    },
+    {
+      "epoch": 0.9207299686341602,
+      "grad_norm": 1.2986893334701157,
+      "learning_rate": 1.6395632372280646e-07,
+      "loss": 0.0219,
+      "step": 3229
+    },
+    {
+      "epoch": 0.9210151126318791,
+      "grad_norm": 1.1072209359887417,
+      "learning_rate": 1.62785334340711e-07,
+      "loss": 0.0162,
+      "step": 3230
+    },
+    {
+      "epoch": 0.9213002566295979,
+      "grad_norm": 1.116053157390405,
+      "learning_rate": 1.6161847240852624e-07,
+      "loss": 0.0133,
+      "step": 3231
+    },
+    {
+      "epoch": 0.9215854006273168,
+      "grad_norm": 0.6591215243411573,
+      "learning_rate": 1.6045573892190136e-07,
+      "loss": 0.0161,
+      "step": 3232
+    },
+    {
+      "epoch": 0.9218705446250356,
+      "grad_norm": 0.44109595747664676,
+      "learning_rate": 1.5929713487296162e-07,
+      "loss": 0.0055,
+      "step": 3233
+    },
+    {
+      "epoch": 0.9221556886227545,
+      "grad_norm": 1.3398968208505428,
+      "learning_rate": 1.581426612503084e-07,
+      "loss": 0.0225,
+      "step": 3234
+    },
+    {
+      "epoch": 0.9224408326204734,
+      "grad_norm": 1.4579154403674046,
+      "learning_rate": 1.5699231903901934e-07,
+      "loss": 0.0216,
+      "step": 3235
+    },
+    {
+      "epoch": 0.9227259766181922,
+      "grad_norm": 2.4188092429370704,
+      "learning_rate": 1.5584610922064759e-07,
+      "loss": 0.0523,
+      "step": 3236
+    },
+    {
+      "epoch": 0.9230111206159111,
+      "grad_norm": 0.5884136082750657,
+      "learning_rate": 1.547040327732191e-07,
+      "loss": 0.0091,
+      "step": 3237
+    },
+    {
+      "epoch": 0.9232962646136299,
+      "grad_norm": 1.108590752242791,
+      "learning_rate": 1.535660906712333e-07,
+      "loss": 0.009,
+      "step": 3238
+    },
+    {
+      "epoch": 0.9235814086113487,
+      "grad_norm": 1.3748359488240862,
+      "learning_rate": 1.5243228388566233e-07,
+      "loss": 0.0145,
+      "step": 3239
+    },
+    {
+      "epoch": 0.9238665526090676,
+      "grad_norm": 0.20338331633106474,
+      "learning_rate": 1.5130261338394904e-07,
+      "loss": 0.004,
+      "step": 3240
+    },
+    {
+      "epoch": 0.9241516966067864,
+      "grad_norm": 0.6798693978374936,
+      "learning_rate": 1.5017708013000787e-07,
+      "loss": 0.02,
+      "step": 3241
+    },
+    {
+      "epoch": 0.9244368406045053,
+      "grad_norm": 1.2945381953335924,
+      "learning_rate": 1.4905568508422173e-07,
+      "loss": 0.0221,
+      "step": 3242
+    },
+    {
+      "epoch": 0.9247219846022241,
+      "grad_norm": 1.1088855335116197,
+      "learning_rate": 1.4793842920344358e-07,
+      "loss": 0.0173,
+      "step": 3243
+    },
+    {
+      "epoch": 0.9250071285999429,
+      "grad_norm": 0.6902543763473945,
+      "learning_rate": 1.468253134409947e-07,
+      "loss": 0.0078,
+      "step": 3244
+    },
+    {
+      "epoch": 0.9252922725976618,
+      "grad_norm": 0.36426685758935945,
+      "learning_rate": 1.4571633874666313e-07,
+      "loss": 0.0068,
+      "step": 3245
+    },
+    {
+      "epoch": 0.9255774165953806,
+      "grad_norm": 1.089534630630725,
+      "learning_rate": 1.4461150606670414e-07,
+      "loss": 0.0283,
+      "step": 3246
+    },
+    {
+      "epoch": 0.9258625605930996,
+      "grad_norm": 1.1642004537200386,
+      "learning_rate": 1.4351081634383647e-07,
+      "loss": 0.0257,
+      "step": 3247
+    },
+    {
+      "epoch": 0.9261477045908184,
+      "grad_norm": 0.7868438826662598,
+      "learning_rate": 1.4241427051724765e-07,
+      "loss": 0.0103,
+      "step": 3248
+    },
+    {
+      "epoch": 0.9264328485885372,
+      "grad_norm": 0.9770444382275083,
+      "learning_rate": 1.4132186952258653e-07,
+      "loss": 0.0153,
+      "step": 3249
+    },
+    {
+      "epoch": 0.9267179925862561,
+      "grad_norm": 1.785958725620067,
+      "learning_rate": 1.402336142919658e-07,
+      "loss": 0.0264,
+      "step": 3250
+    },
+    {
+      "epoch": 0.9270031365839749,
+      "grad_norm": 1.2964838487983503,
+      "learning_rate": 1.3914950575396102e-07,
+      "loss": 0.0267,
+      "step": 3251
+    },
+    {
+      "epoch": 0.9272882805816938,
+      "grad_norm": 1.6603947739477958,
+      "learning_rate": 1.3806954483361002e-07,
+      "loss": 0.0208,
+      "step": 3252
+    },
+    {
+      "epoch": 0.9275734245794126,
+      "grad_norm": 1.9307212298456977,
+      "learning_rate": 1.369937324524101e-07,
+      "loss": 0.0458,
+      "step": 3253
+    },
+    {
+      "epoch": 0.9278585685771314,
+      "grad_norm": 3.4936463480149653,
+      "learning_rate": 1.3592206952832031e-07,
+      "loss": 0.0591,
+      "step": 3254
+    },
+    {
+      "epoch": 0.9281437125748503,
+      "grad_norm": 1.3343590595906798,
+      "learning_rate": 1.3485455697575755e-07,
+      "loss": 0.0246,
+      "step": 3255
+    },
+    {
+      "epoch": 0.9284288565725691,
+      "grad_norm": 1.4183852372661856,
+      "learning_rate": 1.3379119570559872e-07,
+      "loss": 0.0362,
+      "step": 3256
+    },
+    {
+      "epoch": 0.928714000570288,
+      "grad_norm": 1.540850659243285,
+      "learning_rate": 1.3273198662517917e-07,
+      "loss": 0.0269,
+      "step": 3257
+    },
+    {
+      "epoch": 0.9289991445680068,
+      "grad_norm": 1.5560860076422636,
+      "learning_rate": 1.3167693063828867e-07,
+      "loss": 0.0163,
+      "step": 3258
+    },
+    {
+      "epoch": 0.9292842885657256,
+      "grad_norm": 1.484976547707544,
+      "learning_rate": 1.3062602864517548e-07,
+      "loss": 0.0472,
+      "step": 3259
+    },
+    {
+      "epoch": 0.9295694325634445,
+      "grad_norm": 0.6381737168749684,
+      "learning_rate": 1.2957928154254174e-07,
+      "loss": 0.0107,
+      "step": 3260
+    },
+    {
+      "epoch": 0.9298545765611634,
+      "grad_norm": 1.1036012557283326,
+      "learning_rate": 1.285366902235463e-07,
+      "loss": 0.0258,
+      "step": 3261
+    },
+    {
+      "epoch": 0.9301397205588823,
+      "grad_norm": 1.9037929843957349,
+      "learning_rate": 1.2749825557780148e-07,
+      "loss": 0.0647,
+      "step": 3262
+    },
+    {
+      "epoch": 0.9304248645566011,
+      "grad_norm": 1.1279142271655538,
+      "learning_rate": 1.264639784913707e-07,
+      "loss": 0.0206,
+      "step": 3263
+    },
+    {
+      "epoch": 0.93071000855432,
+      "grad_norm": 1.3538375710084851,
+      "learning_rate": 1.2543385984677249e-07,
+      "loss": 0.0294,
+      "step": 3264
+    },
+    {
+      "epoch": 0.9309951525520388,
+      "grad_norm": 1.5857087537071777,
+      "learning_rate": 1.2440790052297648e-07,
+      "loss": 0.0262,
+      "step": 3265
+    },
+    {
+      "epoch": 0.9312802965497576,
+      "grad_norm": 0.8358164956623412,
+      "learning_rate": 1.233861013954024e-07,
+      "loss": 0.0128,
+      "step": 3266
+    },
+    {
+      "epoch": 0.9315654405474765,
+      "grad_norm": 1.0889195632514523,
+      "learning_rate": 1.2236846333592068e-07,
+      "loss": 0.0161,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9318505845451953,
+      "grad_norm": 1.3316230123061,
+      "learning_rate": 1.2135498721285167e-07,
+      "loss": 0.0568,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9321357285429142,
+      "grad_norm": 1.1761990589238678,
+      "learning_rate": 1.2034567389096364e-07,
+      "loss": 0.0179,
+      "step": 3269
+    },
+    {
+      "epoch": 0.932420872540633,
+      "grad_norm": 1.460756188712811,
+      "learning_rate": 1.193405242314738e-07,
+      "loss": 0.0242,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9327060165383518,
+      "grad_norm": 1.4799019491857686,
+      "learning_rate": 1.1833953909204554e-07,
+      "loss": 0.0227,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9329911605360707,
+      "grad_norm": 1.500459177807598,
+      "learning_rate": 1.1734271932679008e-07,
+      "loss": 0.0475,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9332763045337895,
+      "grad_norm": 1.4260297811235727,
+      "learning_rate": 1.1635006578626374e-07,
+      "loss": 0.0249,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9335614485315085,
+      "grad_norm": 2.2936699937794636,
+      "learning_rate": 1.1536157931746728e-07,
+      "loss": 0.049,
+      "step": 3274
+    },
+    {
+      "epoch": 0.9338465925292273,
+      "grad_norm": 1.3381636184621566,
+      "learning_rate": 1.1437726076384715e-07,
+      "loss": 0.0273,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9341317365269461,
+      "grad_norm": 2.3124956924737776,
+      "learning_rate": 1.1339711096529149e-07,
+      "loss": 0.0484,
+      "step": 3276
+    },
+    {
+      "epoch": 0.934416880524665,
+      "grad_norm": 1.3552422519009062,
+      "learning_rate": 1.1242113075813466e-07,
+      "loss": 0.0544,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9347020245223838,
+      "grad_norm": 0.4791576527546003,
+      "learning_rate": 1.1144932097515048e-07,
+      "loss": 0.0083,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9349871685201027,
+      "grad_norm": 1.1383690407288758,
+      "learning_rate": 1.1048168244555513e-07,
+      "loss": 0.0321,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9352723125178215,
+      "grad_norm": 0.5018323683744155,
+      "learning_rate": 1.0951821599500423e-07,
+      "loss": 0.0099,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9355574565155403,
+      "grad_norm": 0.38194627856193947,
+      "learning_rate": 1.0855892244559573e-07,
+      "loss": 0.0058,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9358426005132592,
+      "grad_norm": 1.7992899459503027,
+      "learning_rate": 1.0760380261586656e-07,
+      "loss": 0.022,
+      "step": 3282
+    },
+    {
+      "epoch": 0.936127744510978,
+      "grad_norm": 1.608919056354979,
+      "learning_rate": 1.0665285732079145e-07,
+      "loss": 0.0373,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9364128885086969,
+      "grad_norm": 0.5295751980400589,
+      "learning_rate": 1.0570608737178245e-07,
+      "loss": 0.0143,
+      "step": 3284
+    },
+    {
+      "epoch": 0.9366980325064157,
+      "grad_norm": 1.0884180574425513,
+      "learning_rate": 1.0476349357669113e-07,
+      "loss": 0.0176,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9369831765041345,
+      "grad_norm": 0.8642293290298916,
+      "learning_rate": 1.0382507673980358e-07,
+      "loss": 0.0239,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9372683205018535,
+      "grad_norm": 1.0721029415636114,
+      "learning_rate": 1.0289083766184371e-07,
+      "loss": 0.0125,
+      "step": 3287
+    },
+    {
+      "epoch": 0.9375534644995723,
+      "grad_norm": 1.8058621968187927,
+      "learning_rate": 1.0196077713996777e-07,
+      "loss": 0.0268,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9378386084972912,
+      "grad_norm": 1.6236022683941318,
+      "learning_rate": 1.0103489596777094e-07,
+      "loss": 0.0507,
+      "step": 3289
+    },
+    {
+      "epoch": 0.93812375249501,
+      "grad_norm": 0.3847928622518018,
+      "learning_rate": 1.0011319493527849e-07,
+      "loss": 0.0048,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9384088964927289,
+      "grad_norm": 0.5354601853769945,
+      "learning_rate": 9.919567482894965e-08,
+      "loss": 0.0138,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9386940404904477,
+      "grad_norm": 1.707818158964234,
+      "learning_rate": 9.828233643167762e-08,
+      "loss": 0.0157,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9389791844881665,
+      "grad_norm": 0.8286681216996449,
+      "learning_rate": 9.737318052278622e-08,
+      "loss": 0.0227,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9392643284858854,
+      "grad_norm": 1.638770990611571,
+      "learning_rate": 9.646820787803102e-08,
+      "loss": 0.0238,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9395494724836042,
+      "grad_norm": 1.4844955571277005,
+      "learning_rate": 9.556741926959878e-08,
+      "loss": 0.0251,
+      "step": 3295
+    },
+    {
+      "epoch": 0.939834616481323,
+      "grad_norm": 1.2669550894482697,
+      "learning_rate": 9.467081546610357e-08,
+      "loss": 0.0279,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9401197604790419,
+      "grad_norm": 0.607099774823454,
+      "learning_rate": 9.377839723259175e-08,
+      "loss": 0.0121,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9404049044767607,
+      "grad_norm": 0.8781917053000157,
+      "learning_rate": 9.289016533053696e-08,
+      "loss": 0.0104,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9406900484744796,
+      "grad_norm": 1.3542338084425016,
+      "learning_rate": 9.200612051784019e-08,
+      "loss": 0.0167,
+      "step": 3299
+    },
+    {
+      "epoch": 0.9409751924721984,
+      "grad_norm": 1.1515026257845251,
+      "learning_rate": 9.112626354883025e-08,
+      "loss": 0.0166,
+      "step": 3300
+    },
+    {
+      "epoch": 0.9412603364699174,
+      "grad_norm": 1.2561544765269976,
+      "learning_rate": 9.025059517426383e-08,
+      "loss": 0.0202,
+      "step": 3301
+    },
+    {
+      "epoch": 0.9415454804676362,
+      "grad_norm": 0.8869786585507469,
+      "learning_rate": 8.937911614132155e-08,
+      "loss": 0.0113,
+      "step": 3302
+    },
+    {
+      "epoch": 0.941830624465355,
+      "grad_norm": 1.1028012235771263,
+      "learning_rate": 8.851182719361029e-08,
+      "loss": 0.0213,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9421157684630739,
+      "grad_norm": 0.8016773289830269,
+      "learning_rate": 8.764872907116084e-08,
+      "loss": 0.0126,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9424009124607927,
+      "grad_norm": 2.1285540435739647,
+      "learning_rate": 8.678982251043078e-08,
+      "loss": 0.0291,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9426860564585116,
+      "grad_norm": 0.5550319327426171,
+      "learning_rate": 8.59351082442983e-08,
+      "loss": 0.0102,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9429712004562304,
+      "grad_norm": 0.8827415276408771,
+      "learning_rate": 8.508458700206501e-08,
+      "loss": 0.0167,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9432563444539492,
+      "grad_norm": 1.0858933176242238,
+      "learning_rate": 8.423825950945541e-08,
+      "loss": 0.0139,
+      "step": 3308
+    },
+    {
+      "epoch": 0.9435414884516681,
+      "grad_norm": 0.9312052457091772,
+      "learning_rate": 8.339612648861573e-08,
+      "loss": 0.0189,
+      "step": 3309
+    },
+    {
+      "epoch": 0.9438266324493869,
+      "grad_norm": 0.8767853437349821,
+      "learning_rate": 8.255818865811226e-08,
+      "loss": 0.0209,
+      "step": 3310
+    },
+    {
+      "epoch": 0.9441117764471058,
+      "grad_norm": 1.1912037158829953,
+      "learning_rate": 8.172444673293201e-08,
+      "loss": 0.0154,
+      "step": 3311
+    },
+    {
+      "epoch": 0.9443969204448246,
+      "grad_norm": 1.1446724760051077,
+      "learning_rate": 8.089490142448254e-08,
+      "loss": 0.0159,
+      "step": 3312
+    },
+    {
+      "epoch": 0.9446820644425434,
+      "grad_norm": 1.254371535428258,
+      "learning_rate": 8.006955344058986e-08,
+      "loss": 0.0203,
+      "step": 3313
+    },
+    {
+      "epoch": 0.9449672084402624,
+      "grad_norm": 0.9243205335182174,
+      "learning_rate": 7.92484034854979e-08,
+      "loss": 0.0147,
+      "step": 3314
+    },
+    {
+      "epoch": 0.9452523524379812,
+      "grad_norm": 0.5842500345183507,
+      "learning_rate": 7.843145225987003e-08,
+      "loss": 0.0075,
+      "step": 3315
+    },
+    {
+      "epoch": 0.9455374964357001,
+      "grad_norm": 1.493712093765859,
+      "learning_rate": 7.761870046078534e-08,
+      "loss": 0.0231,
+      "step": 3316
+    },
+    {
+      "epoch": 0.9458226404334189,
+      "grad_norm": 1.162136015617278,
+      "learning_rate": 7.681014878174187e-08,
+      "loss": 0.0321,
+      "step": 3317
+    },
+    {
+      "epoch": 0.9461077844311377,
+      "grad_norm": 0.5474839381848375,
+      "learning_rate": 7.600579791265161e-08,
+      "loss": 0.0117,
+      "step": 3318
+    },
+    {
+      "epoch": 0.9463929284288566,
+      "grad_norm": 1.3755166848139597,
+      "learning_rate": 7.52056485398428e-08,
+      "loss": 0.0244,
+      "step": 3319
+    },
+    {
+      "epoch": 0.9466780724265754,
+      "grad_norm": 1.3943608791399622,
+      "learning_rate": 7.440970134605819e-08,
+      "loss": 0.0342,
+      "step": 3320
+    },
+    {
+      "epoch": 0.9469632164242943,
+      "grad_norm": 0.9763179433503778,
+      "learning_rate": 7.361795701045726e-08,
+      "loss": 0.0233,
+      "step": 3321
+    },
+    {
+      "epoch": 0.9472483604220131,
+      "grad_norm": 0.8089246096238166,
+      "learning_rate": 7.283041620861131e-08,
+      "loss": 0.0127,
+      "step": 3322
+    },
+    {
+      "epoch": 0.9475335044197319,
+      "grad_norm": 0.7679809907075772,
+      "learning_rate": 7.204707961250446e-08,
+      "loss": 0.0156,
+      "step": 3323
+    },
+    {
+      "epoch": 0.9478186484174508,
+      "grad_norm": 0.9473437484277629,
+      "learning_rate": 7.126794789053426e-08,
+      "loss": 0.0147,
+      "step": 3324
+    },
+    {
+      "epoch": 0.9481037924151696,
+      "grad_norm": 1.38358551291656,
+      "learning_rate": 7.049302170751115e-08,
+      "loss": 0.0567,
+      "step": 3325
+    },
+    {
+      "epoch": 0.9483889364128885,
+      "grad_norm": 1.2621054129425233,
+      "learning_rate": 6.972230172465567e-08,
+      "loss": 0.0231,
+      "step": 3326
+    },
+    {
+      "epoch": 0.9486740804106074,
+      "grad_norm": 0.3297509743111443,
+      "learning_rate": 6.895578859960062e-08,
+      "loss": 0.0072,
+      "step": 3327
+    },
+    {
+      "epoch": 0.9489592244083263,
+      "grad_norm": 0.2484790392335108,
+      "learning_rate": 6.819348298638839e-08,
+      "loss": 0.0045,
+      "step": 3328
+    },
+    {
+      "epoch": 0.9492443684060451,
+      "grad_norm": 0.5873609332531889,
+      "learning_rate": 6.743538553547091e-08,
+      "loss": 0.0107,
+      "step": 3329
+    },
+    {
+      "epoch": 0.9495295124037639,
+      "grad_norm": 1.7018257018683238,
+      "learning_rate": 6.668149689371074e-08,
+      "loss": 0.0432,
+      "step": 3330
+    },
+    {
+      "epoch": 0.9498146564014828,
+      "grad_norm": 0.5246171733900967,
+      "learning_rate": 6.593181770437829e-08,
+      "loss": 0.0068,
+      "step": 3331
+    },
+    {
+      "epoch": 0.9500998003992016,
+      "grad_norm": 0.7763822843561252,
+      "learning_rate": 6.518634860715134e-08,
+      "loss": 0.0162,
+      "step": 3332
+    },
+    {
+      "epoch": 0.9503849443969205,
+      "grad_norm": 1.0168579745893171,
+      "learning_rate": 6.444509023811773e-08,
+      "loss": 0.0209,
+      "step": 3333
+    },
+    {
+      "epoch": 0.9506700883946393,
+      "grad_norm": 1.075777400723064,
+      "learning_rate": 6.370804322977042e-08,
+      "loss": 0.0189,
+      "step": 3334
+    },
+    {
+      "epoch": 0.9509552323923581,
+      "grad_norm": 1.5206330880639494,
+      "learning_rate": 6.297520821100911e-08,
+      "loss": 0.0462,
+      "step": 3335
+    },
+    {
+      "epoch": 0.951240376390077,
+      "grad_norm": 1.5897941053837084,
+      "learning_rate": 6.224658580713971e-08,
+      "loss": 0.0238,
+      "step": 3336
+    },
+    {
+      "epoch": 0.9515255203877958,
+      "grad_norm": 2.1685547816279316,
+      "learning_rate": 6.152217663987437e-08,
+      "loss": 0.0302,
+      "step": 3337
+    },
+    {
+      "epoch": 0.9518106643855146,
+      "grad_norm": 0.41072290294514274,
+      "learning_rate": 6.080198132732917e-08,
+      "loss": 0.0062,
+      "step": 3338
+    },
+    {
+      "epoch": 0.9520958083832335,
+      "grad_norm": 1.6595471882531296,
+      "learning_rate": 6.008600048402647e-08,
+      "loss": 0.0236,
+      "step": 3339
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 1.3876995960785814,
+      "learning_rate": 5.937423472088866e-08,
+      "loss": 0.0173,
+      "step": 3340
+    },
+    {
+      "epoch": 0.9526660963786713,
+      "grad_norm": 0.95914211608916,
+      "learning_rate": 5.866668464524661e-08,
+      "loss": 0.0146,
+      "step": 3341
+    },
+    {
+      "epoch": 0.9529512403763901,
+      "grad_norm": 2.367073977432817,
+      "learning_rate": 5.796335086083016e-08,
+      "loss": 0.0281,
+      "step": 3342
+    },
+    {
+      "epoch": 0.953236384374109,
+      "grad_norm": 0.8209213096210055,
+      "learning_rate": 5.7264233967773696e-08,
+      "loss": 0.0127,
+      "step": 3343
+    },
+    {
+      "epoch": 0.9535215283718278,
+      "grad_norm": 1.8947375988470225,
+      "learning_rate": 5.6569334562611714e-08,
+      "loss": 0.0205,
+      "step": 3344
+    },
+    {
+      "epoch": 0.9538066723695466,
+      "grad_norm": 1.043061161634943,
+      "learning_rate": 5.5878653238281564e-08,
+      "loss": 0.0209,
+      "step": 3345
+    },
+    {
+      "epoch": 0.9540918163672655,
+      "grad_norm": 0.8863361955005523,
+      "learning_rate": 5.519219058412129e-08,
+      "loss": 0.0098,
+      "step": 3346
+    },
+    {
+      "epoch": 0.9543769603649843,
+      "grad_norm": 0.5971170239016343,
+      "learning_rate": 5.4509947185867883e-08,
+      "loss": 0.0162,
+      "step": 3347
+    },
+    {
+      "epoch": 0.9546621043627032,
+      "grad_norm": 1.117922239866681,
+      "learning_rate": 5.3831923625659034e-08,
+      "loss": 0.0235,
+      "step": 3348
+    },
+    {
+      "epoch": 0.954947248360422,
+      "grad_norm": 0.4746383532243566,
+      "learning_rate": 5.315812048203306e-08,
+      "loss": 0.0081,
+      "step": 3349
+    },
+    {
+      "epoch": 0.9552323923581408,
+      "grad_norm": 0.7993643810754686,
+      "learning_rate": 5.2488538329926175e-08,
+      "loss": 0.0214,
+      "step": 3350
+    },
+    {
+      "epoch": 0.9555175363558597,
+      "grad_norm": 1.4852695501324207,
+      "learning_rate": 5.18231777406708e-08,
+      "loss": 0.032,
+      "step": 3351
+    },
+    {
+      "epoch": 0.9558026803535785,
+      "grad_norm": 2.15399813077587,
+      "learning_rate": 5.116203928200003e-08,
+      "loss": 0.056,
+      "step": 3352
+    },
+    {
+      "epoch": 0.9560878243512974,
+      "grad_norm": 1.4790577170889294,
+      "learning_rate": 5.050512351804371e-08,
+      "loss": 0.035,
+      "step": 3353
+    },
+    {
+      "epoch": 0.9563729683490163,
+      "grad_norm": 1.5138221982520688,
+      "learning_rate": 4.9852431009328464e-08,
+      "loss": 0.0206,
+      "step": 3354
+    },
+    {
+      "epoch": 0.9566581123467351,
+      "grad_norm": 1.2976796226213914,
+      "learning_rate": 4.920396231277713e-08,
+      "loss": 0.0185,
+      "step": 3355
+    },
+    {
+      "epoch": 0.956943256344454,
+      "grad_norm": 0.8940591579415054,
+      "learning_rate": 4.855971798170822e-08,
+      "loss": 0.0161,
+      "step": 3356
+    },
+    {
+      "epoch": 0.9572284003421728,
+      "grad_norm": 1.3689820258335235,
+      "learning_rate": 4.7919698565835894e-08,
+      "loss": 0.0346,
+      "step": 3357
+    },
+    {
+      "epoch": 0.9575135443398917,
+      "grad_norm": 0.6602976007376902,
+      "learning_rate": 4.728390461126997e-08,
+      "loss": 0.0135,
+      "step": 3358
+    },
+    {
+      "epoch": 0.9577986883376105,
+      "grad_norm": 1.7405971970346032,
+      "learning_rate": 4.6652336660514275e-08,
+      "loss": 0.0391,
+      "step": 3359
+    },
+    {
+      "epoch": 0.9580838323353293,
+      "grad_norm": 0.7948682203377249,
+      "learning_rate": 4.602499525246606e-08,
+      "loss": 0.0095,
+      "step": 3360
+    },
+    {
+      "epoch": 0.9583689763330482,
+      "grad_norm": 2.246065889190937,
+      "learning_rate": 4.5401880922418264e-08,
+      "loss": 0.0295,
+      "step": 3361
+    },
+    {
+      "epoch": 0.958654120330767,
+      "grad_norm": 0.998355740167951,
+      "learning_rate": 4.478299420205445e-08,
+      "loss": 0.0134,
+      "step": 3362
+    },
+    {
+      "epoch": 0.9589392643284859,
+      "grad_norm": 1.4618361724398996,
+      "learning_rate": 4.416833561945222e-08,
+      "loss": 0.0361,
+      "step": 3363
+    },
+    {
+      "epoch": 0.9592244083262047,
+      "grad_norm": 1.3947578420400226,
+      "learning_rate": 4.355790569908147e-08,
+      "loss": 0.0316,
+      "step": 3364
+    },
+    {
+      "epoch": 0.9595095523239235,
+      "grad_norm": 1.092207434346036,
+      "learning_rate": 4.29517049618039e-08,
+      "loss": 0.0205,
+      "step": 3365
+    },
+    {
+      "epoch": 0.9597946963216424,
+      "grad_norm": 1.2842978731112948,
+      "learning_rate": 4.2349733924872406e-08,
+      "loss": 0.0261,
+      "step": 3366
+    },
+    {
+      "epoch": 0.9600798403193613,
+      "grad_norm": 1.9319381855432591,
+      "learning_rate": 4.1751993101930565e-08,
+      "loss": 0.0313,
+      "step": 3367
+    },
+    {
+      "epoch": 0.9603649843170802,
+      "grad_norm": 1.9881438705501773,
+      "learning_rate": 4.1158483003012614e-08,
+      "loss": 0.0338,
+      "step": 3368
+    },
+    {
+      "epoch": 0.960650128314799,
+      "grad_norm": 1.5737829057522759,
+      "learning_rate": 4.056920413454291e-08,
+      "loss": 0.0237,
+      "step": 3369
+    },
+    {
+      "epoch": 0.9609352723125179,
+      "grad_norm": 1.000096415972,
+      "learning_rate": 3.9984156999335334e-08,
+      "loss": 0.0145,
+      "step": 3370
+    },
+    {
+      "epoch": 0.9612204163102367,
+      "grad_norm": 1.35746116703504,
+      "learning_rate": 3.94033420965928e-08,
+      "loss": 0.015,
+      "step": 3371
+    },
+    {
+      "epoch": 0.9615055603079555,
+      "grad_norm": 2.149928406053472,
+      "learning_rate": 3.882675992190832e-08,
+      "loss": 0.0354,
+      "step": 3372
+    },
+    {
+      "epoch": 0.9617907043056744,
+      "grad_norm": 0.7299966913484716,
+      "learning_rate": 3.825441096726057e-08,
+      "loss": 0.008,
+      "step": 3373
+    },
+    {
+      "epoch": 0.9620758483033932,
+      "grad_norm": 0.3263502956977538,
+      "learning_rate": 3.76862957210189e-08,
+      "loss": 0.004,
+      "step": 3374
+    },
+    {
+      "epoch": 0.962360992301112,
+      "grad_norm": 1.2745281137121607,
+      "learning_rate": 3.712241466793887e-08,
+      "loss": 0.0148,
+      "step": 3375
+    },
+    {
+      "epoch": 0.9626461362988309,
+      "grad_norm": 1.0802932953346063,
+      "learning_rate": 3.6562768289162834e-08,
+      "loss": 0.0155,
+      "step": 3376
+    },
+    {
+      "epoch": 0.9629312802965497,
+      "grad_norm": 0.8278153218304807,
+      "learning_rate": 3.6007357062219914e-08,
+      "loss": 0.0122,
+      "step": 3377
+    },
+    {
+      "epoch": 0.9632164242942686,
+      "grad_norm": 0.22430709604137844,
+      "learning_rate": 3.5456181461026585e-08,
+      "loss": 0.0042,
+      "step": 3378
+    },
+    {
+      "epoch": 0.9635015682919874,
+      "grad_norm": 0.8451639389040133,
+      "learning_rate": 3.4909241955883853e-08,
+      "loss": 0.014,
+      "step": 3379
+    },
+    {
+      "epoch": 0.9637867122897062,
+      "grad_norm": 1.3080849693612018,
+      "learning_rate": 3.4366539013478975e-08,
+      "loss": 0.024,
+      "step": 3380
+    },
+    {
+      "epoch": 0.9640718562874252,
+      "grad_norm": 1.4925602998353593,
+      "learning_rate": 3.3828073096884294e-08,
+      "loss": 0.0229,
+      "step": 3381
+    },
+    {
+      "epoch": 0.964357000285144,
+      "grad_norm": 0.9866730078294964,
+      "learning_rate": 3.329384466555619e-08,
+      "loss": 0.0332,
+      "step": 3382
+    },
+    {
+      "epoch": 0.9646421442828629,
+      "grad_norm": 0.53963382342648,
+      "learning_rate": 3.276385417533612e-08,
+      "loss": 0.0076,
+      "step": 3383
+    },
+    {
+      "epoch": 0.9649272882805817,
+      "grad_norm": 2.028169739046422,
+      "learning_rate": 3.2238102078448466e-08,
+      "loss": 0.0525,
+      "step": 3384
+    },
+    {
+      "epoch": 0.9652124322783006,
+      "grad_norm": 0.70221035179003,
+      "learning_rate": 3.17165888235027e-08,
+      "loss": 0.0147,
+      "step": 3385
+    },
+    {
+      "epoch": 0.9654975762760194,
+      "grad_norm": 1.594502446966948,
+      "learning_rate": 3.1199314855489547e-08,
+      "loss": 0.0386,
+      "step": 3386
+    },
+    {
+      "epoch": 0.9657827202737382,
+      "grad_norm": 0.6155279599436443,
+      "learning_rate": 3.0686280615783156e-08,
+      "loss": 0.007,
+      "step": 3387
+    },
+    {
+      "epoch": 0.9660678642714571,
+      "grad_norm": 0.9940196228788715,
+      "learning_rate": 3.0177486542141144e-08,
+      "loss": 0.0181,
+      "step": 3388
+    },
+    {
+      "epoch": 0.9663530082691759,
+      "grad_norm": 0.9662957584867693,
+      "learning_rate": 2.9672933068701227e-08,
+      "loss": 0.0191,
+      "step": 3389
+    },
+    {
+      "epoch": 0.9666381522668948,
+      "grad_norm": 1.3090922190101484,
+      "learning_rate": 2.9172620625984583e-08,
+      "loss": 0.0244,
+      "step": 3390
+    },
+    {
+      "epoch": 0.9669232962646136,
+      "grad_norm": 0.8123264288075666,
+      "learning_rate": 2.8676549640892502e-08,
+      "loss": 0.015,
+      "step": 3391
+    },
+    {
+      "epoch": 0.9672084402623324,
+      "grad_norm": 0.8232507163935322,
+      "learning_rate": 2.8184720536706956e-08,
+      "loss": 0.0161,
+      "step": 3392
+    },
+    {
+      "epoch": 0.9674935842600513,
+      "grad_norm": 0.6151000895965828,
+      "learning_rate": 2.7697133733091686e-08,
+      "loss": 0.01,
+      "step": 3393
+    },
+    {
+      "epoch": 0.9677787282577702,
+      "grad_norm": 1.5924017390252279,
+      "learning_rate": 2.7213789646088896e-08,
+      "loss": 0.055,
+      "step": 3394
+    },
+    {
+      "epoch": 0.9680638722554891,
+      "grad_norm": 1.4169645143655845,
+      "learning_rate": 2.673468868812312e-08,
+      "loss": 0.0179,
+      "step": 3395
+    },
+    {
+      "epoch": 0.9683490162532079,
+      "grad_norm": 0.5948053783440101,
+      "learning_rate": 2.625983126799514e-08,
+      "loss": 0.0083,
+      "step": 3396
+    },
+    {
+      "epoch": 0.9686341602509267,
+      "grad_norm": 1.575445770817084,
+      "learning_rate": 2.57892177908875e-08,
+      "loss": 0.0313,
+      "step": 3397
+    },
+    {
+      "epoch": 0.9689193042486456,
+      "grad_norm": 1.500329123622438,
+      "learning_rate": 2.53228486583601e-08,
+      "loss": 0.0241,
+      "step": 3398
+    },
+    {
+      "epoch": 0.9692044482463644,
+      "grad_norm": 1.311203579000811,
+      "learning_rate": 2.4860724268351845e-08,
+      "loss": 0.0378,
+      "step": 3399
+    },
+    {
+      "epoch": 0.9694895922440833,
+      "grad_norm": 1.7134696799995803,
+      "learning_rate": 2.4402845015180088e-08,
+      "loss": 0.0345,
+      "step": 3400
+    },
+    {
+      "epoch": 0.9697747362418021,
+      "grad_norm": 0.6710349220659272,
+      "learning_rate": 2.3949211289538975e-08,
+      "loss": 0.0073,
+      "step": 3401
+    },
+    {
+      "epoch": 0.9700598802395209,
+      "grad_norm": 1.5504891835178136,
+      "learning_rate": 2.3499823478499995e-08,
+      "loss": 0.0237,
+      "step": 3402
+    },
+    {
+      "epoch": 0.9703450242372398,
+      "grad_norm": 0.7081799782835921,
+      "learning_rate": 2.305468196551308e-08,
+      "loss": 0.0148,
+      "step": 3403
+    },
+    {
+      "epoch": 0.9706301682349586,
+      "grad_norm": 0.9281303546894957,
+      "learning_rate": 2.2613787130403854e-08,
+      "loss": 0.0192,
+      "step": 3404
+    },
+    {
+      "epoch": 0.9709153122326775,
+      "grad_norm": 1.27109390632808,
+      "learning_rate": 2.217713934937471e-08,
+      "loss": 0.0453,
+      "step": 3405
+    },
+    {
+      "epoch": 0.9712004562303963,
+      "grad_norm": 0.6464573130091568,
+      "learning_rate": 2.1744738995003733e-08,
+      "loss": 0.0253,
+      "step": 3406
+    },
+    {
+      "epoch": 0.9714856002281153,
+      "grad_norm": 1.119205251723142,
+      "learning_rate": 2.131658643624579e-08,
+      "loss": 0.0356,
+      "step": 3407
+    },
+    {
+      "epoch": 0.9717707442258341,
+      "grad_norm": 1.4078034602604481,
+      "learning_rate": 2.0892682038429758e-08,
+      "loss": 0.0391,
+      "step": 3408
+    },
+    {
+      "epoch": 0.9720558882235529,
+      "grad_norm": 0.6417882749586009,
+      "learning_rate": 2.0473026163261302e-08,
+      "loss": 0.0155,
+      "step": 3409
+    },
+    {
+      "epoch": 0.9723410322212718,
+      "grad_norm": 0.5777849338634182,
+      "learning_rate": 2.0057619168819544e-08,
+      "loss": 0.0132,
+      "step": 3410
+    },
+    {
+      "epoch": 0.9726261762189906,
+      "grad_norm": 0.7877201908002133,
+      "learning_rate": 1.964646140955928e-08,
+      "loss": 0.0282,
+      "step": 3411
+    },
+    {
+      "epoch": 0.9729113202167095,
+      "grad_norm": 1.1195581875615601,
+      "learning_rate": 1.923955323630877e-08,
+      "loss": 0.0124,
+      "step": 3412
+    },
+    {
+      "epoch": 0.9731964642144283,
+      "grad_norm": 1.1199818626019644,
+      "learning_rate": 1.883689499627084e-08,
+      "loss": 0.0156,
+      "step": 3413
+    },
+    {
+      "epoch": 0.9734816082121471,
+      "grad_norm": 0.600014710134955,
+      "learning_rate": 1.84384870330212e-08,
+      "loss": 0.0217,
+      "step": 3414
+    },
+    {
+      "epoch": 0.973766752209866,
+      "grad_norm": 0.980931450691564,
+      "learning_rate": 1.8044329686509598e-08,
+      "loss": 0.0199,
+      "step": 3415
+    },
+    {
+      "epoch": 0.9740518962075848,
+      "grad_norm": 0.820500426709812,
+      "learning_rate": 1.7654423293058666e-08,
+      "loss": 0.0099,
+      "step": 3416
+    },
+    {
+      "epoch": 0.9743370402053037,
+      "grad_norm": 0.7809793057569531,
+      "learning_rate": 1.726876818536394e-08,
+      "loss": 0.0146,
+      "step": 3417
+    },
+    {
+      "epoch": 0.9746221842030225,
+      "grad_norm": 1.0867941197474569,
+      "learning_rate": 1.6887364692493303e-08,
+      "loss": 0.0171,
+      "step": 3418
+    },
+    {
+      "epoch": 0.9749073282007413,
+      "grad_norm": 1.128048720541707,
+      "learning_rate": 1.6510213139886987e-08,
+      "loss": 0.0558,
+      "step": 3419
+    },
+    {
+      "epoch": 0.9751924721984602,
+      "grad_norm": 1.548358358713432,
+      "learning_rate": 1.613731384935702e-08,
+      "loss": 0.0403,
+      "step": 3420
+    },
+    {
+      "epoch": 0.9754776161961791,
+      "grad_norm": 1.50416589986925,
+      "learning_rate": 1.5768667139086645e-08,
+      "loss": 0.0432,
+      "step": 3421
+    },
+    {
+      "epoch": 0.975762760193898,
+      "grad_norm": 1.3525365106907985,
+      "learning_rate": 1.540427332363148e-08,
+      "loss": 0.0193,
+      "step": 3422
+    },
+    {
+      "epoch": 0.9760479041916168,
+      "grad_norm": 2.136152304647279,
+      "learning_rate": 1.5044132713917803e-08,
+      "loss": 0.0434,
+      "step": 3423
+    },
+    {
+      "epoch": 0.9763330481893356,
+      "grad_norm": 0.703158736191946,
+      "learning_rate": 1.4688245617243135e-08,
+      "loss": 0.0141,
+      "step": 3424
+    },
+    {
+      "epoch": 0.9766181921870545,
+      "grad_norm": 1.7291368031174819,
+      "learning_rate": 1.4336612337274014e-08,
+      "loss": 0.0177,
+      "step": 3425
+    },
+    {
+      "epoch": 0.9769033361847733,
+      "grad_norm": 0.7954994827635179,
+      "learning_rate": 1.3989233174050431e-08,
+      "loss": 0.0233,
+      "step": 3426
+    },
+    {
+      "epoch": 0.9771884801824922,
+      "grad_norm": 0.5057022729587526,
+      "learning_rate": 1.3646108423978621e-08,
+      "loss": 0.0083,
+      "step": 3427
+    },
+    {
+      "epoch": 0.977473624180211,
+      "grad_norm": 1.395543096080145,
+      "learning_rate": 1.3307238379838273e-08,
+      "loss": 0.0238,
+      "step": 3428
+    },
+    {
+      "epoch": 0.9777587681779298,
+      "grad_norm": 0.5812031981674576,
+      "learning_rate": 1.2972623330775869e-08,
+      "loss": 0.0112,
+      "step": 3429
+    },
+    {
+      "epoch": 0.9780439121756487,
+      "grad_norm": 1.9594929069324818,
+      "learning_rate": 1.2642263562309131e-08,
+      "loss": 0.0526,
+      "step": 3430
+    },
+    {
+      "epoch": 0.9783290561733675,
+      "grad_norm": 0.6563286778890409,
+      "learning_rate": 1.2316159356323132e-08,
+      "loss": 0.0078,
+      "step": 3431
+    },
+    {
+      "epoch": 0.9786142001710864,
+      "grad_norm": 0.5324232039060124,
+      "learning_rate": 1.1994310991074177e-08,
+      "loss": 0.0108,
+      "step": 3432
+    },
+    {
+      "epoch": 0.9788993441688052,
+      "grad_norm": 1.2259761094193657,
+      "learning_rate": 1.1676718741184812e-08,
+      "loss": 0.033,
+      "step": 3433
+    },
+    {
+      "epoch": 0.9791844881665241,
+      "grad_norm": 1.365541853973848,
+      "learning_rate": 1.1363382877647155e-08,
+      "loss": 0.0183,
+      "step": 3434
+    },
+    {
+      "epoch": 0.979469632164243,
+      "grad_norm": 1.8621831220471732,
+      "learning_rate": 1.1054303667821232e-08,
+      "loss": 0.0299,
+      "step": 3435
+    },
+    {
+      "epoch": 0.9797547761619618,
+      "grad_norm": 0.6347679936689795,
+      "learning_rate": 1.0749481375434966e-08,
+      "loss": 0.0099,
+      "step": 3436
+    },
+    {
+      "epoch": 0.9800399201596807,
+      "grad_norm": 1.1434239902853023,
+      "learning_rate": 1.0448916260584752e-08,
+      "loss": 0.0215,
+      "step": 3437
+    },
+    {
+      "epoch": 0.9803250641573995,
+      "grad_norm": 0.7640056037278686,
+      "learning_rate": 1.0152608579733214e-08,
+      "loss": 0.0187,
+      "step": 3438
+    },
+    {
+      "epoch": 0.9806102081551183,
+      "grad_norm": 0.9356956746133379,
+      "learning_rate": 9.860558585710334e-09,
+      "loss": 0.0122,
+      "step": 3439
+    },
+    {
+      "epoch": 0.9808953521528372,
+      "grad_norm": 1.1473267478745108,
+      "learning_rate": 9.57276652771455e-09,
+      "loss": 0.0166,
+      "step": 3440
+    },
+    {
+      "epoch": 0.981180496150556,
+      "grad_norm": 1.9021257652076382,
+      "learning_rate": 9.289232651309432e-09,
+      "loss": 0.0264,
+      "step": 3441
+    },
+    {
+      "epoch": 0.9814656401482749,
+      "grad_norm": 0.630448042704614,
+      "learning_rate": 9.009957198426455e-09,
+      "loss": 0.0092,
+      "step": 3442
+    },
+    {
+      "epoch": 0.9817507841459937,
+      "grad_norm": 1.1654682091314135,
+      "learning_rate": 8.73494040736278e-09,
+      "loss": 0.0467,
+      "step": 3443
+    },
+    {
+      "epoch": 0.9820359281437125,
+      "grad_norm": 0.8094012015584077,
+      "learning_rate": 8.46418251278236e-09,
+      "loss": 0.0122,
+      "step": 3444
+    },
+    {
+      "epoch": 0.9823210721414314,
+      "grad_norm": 0.43072831856408855,
+      "learning_rate": 8.197683745713725e-09,
+      "loss": 0.0099,
+      "step": 3445
+    },
+    {
+      "epoch": 0.9826062161391502,
+      "grad_norm": 0.9313780942022813,
+      "learning_rate": 7.935444333552756e-09,
+      "loss": 0.0375,
+      "step": 3446
+    },
+    {
+      "epoch": 0.9828913601368692,
+      "grad_norm": 1.86814045946998,
+      "learning_rate": 7.677464500061015e-09,
+      "loss": 0.0354,
+      "step": 3447
+    },
+    {
+      "epoch": 0.983176504134588,
+      "grad_norm": 0.9242833759734096,
+      "learning_rate": 7.423744465364091e-09,
+      "loss": 0.017,
+      "step": 3448
+    },
+    {
+      "epoch": 0.9834616481323069,
+      "grad_norm": 1.004654257677267,
+      "learning_rate": 7.1742844459543605e-09,
+      "loss": 0.0225,
+      "step": 3449
+    },
+    {
+      "epoch": 0.9837467921300257,
+      "grad_norm": 0.5132997575591574,
+      "learning_rate": 6.929084654688223e-09,
+      "loss": 0.0076,
+      "step": 3450
+    },
+    {
+      "epoch": 0.9840319361277445,
+      "grad_norm": 1.2537143990748647,
+      "learning_rate": 6.688145300787208e-09,
+      "loss": 0.0174,
+      "step": 3451
+    },
+    {
+      "epoch": 0.9843170801254634,
+      "grad_norm": 0.856916087539409,
+      "learning_rate": 6.451466589837974e-09,
+      "loss": 0.0076,
+      "step": 3452
+    },
+    {
+      "epoch": 0.9846022241231822,
+      "grad_norm": 1.2874699118028397,
+      "learning_rate": 6.219048723790644e-09,
+      "loss": 0.0161,
+      "step": 3453
+    },
+    {
+      "epoch": 0.984887368120901,
+      "grad_norm": 0.7191572461319289,
+      "learning_rate": 5.990891900961582e-09,
+      "loss": 0.0143,
+      "step": 3454
+    },
+    {
+      "epoch": 0.9851725121186199,
+      "grad_norm": 1.5720947624205361,
+      "learning_rate": 5.766996316029505e-09,
+      "loss": 0.0312,
+      "step": 3455
+    },
+    {
+      "epoch": 0.9854576561163387,
+      "grad_norm": 1.5965820870491516,
+      "learning_rate": 5.547362160037151e-09,
+      "loss": 0.0483,
+      "step": 3456
+    },
+    {
+      "epoch": 0.9857428001140576,
+      "grad_norm": 2.324661905639617,
+      "learning_rate": 5.331989620392386e-09,
+      "loss": 0.0569,
+      "step": 3457
+    },
+    {
+      "epoch": 0.9860279441117764,
+      "grad_norm": 1.1089308013214734,
+      "learning_rate": 5.120878880866542e-09,
+      "loss": 0.0142,
+      "step": 3458
+    },
+    {
+      "epoch": 0.9863130881094953,
+      "grad_norm": 0.8282424350000585,
+      "learning_rate": 4.914030121593305e-09,
+      "loss": 0.025,
+      "step": 3459
+    },
+    {
+      "epoch": 0.9865982321072141,
+      "grad_norm": 1.2887917701143625,
+      "learning_rate": 4.711443519070935e-09,
+      "loss": 0.0357,
+      "step": 3460
+    },
+    {
+      "epoch": 0.986883376104933,
+      "grad_norm": 0.948215811306342,
+      "learning_rate": 4.513119246160602e-09,
+      "loss": 0.0131,
+      "step": 3461
+    },
+    {
+      "epoch": 0.9871685201026519,
+      "grad_norm": 1.220943268524718,
+      "learning_rate": 4.3190574720858305e-09,
+      "loss": 0.0326,
+      "step": 3462
+    },
+    {
+      "epoch": 0.9874536641003707,
+      "grad_norm": 0.41899450985279857,
+      "learning_rate": 4.129258362434163e-09,
+      "loss": 0.0054,
+      "step": 3463
+    },
+    {
+      "epoch": 0.9877388080980896,
+      "grad_norm": 0.647778329863355,
+      "learning_rate": 3.943722079155499e-09,
+      "loss": 0.0151,
+      "step": 3464
+    },
+    {
+      "epoch": 0.9880239520958084,
+      "grad_norm": 1.802544256763322,
+      "learning_rate": 3.762448780562089e-09,
+      "loss": 0.034,
+      "step": 3465
+    },
+    {
+      "epoch": 0.9883090960935272,
+      "grad_norm": 0.9733016714220596,
+      "learning_rate": 3.585438621329096e-09,
+      "loss": 0.0091,
+      "step": 3466
+    },
+    {
+      "epoch": 0.9885942400912461,
+      "grad_norm": 0.7507618500967945,
+      "learning_rate": 3.41269175249459e-09,
+      "loss": 0.0134,
+      "step": 3467
+    },
+    {
+      "epoch": 0.9888793840889649,
+      "grad_norm": 0.6719411044195046,
+      "learning_rate": 3.2442083214573316e-09,
+      "loss": 0.0074,
+      "step": 3468
+    },
+    {
+      "epoch": 0.9891645280866838,
+      "grad_norm": 2.0659767490048186,
+      "learning_rate": 3.0799884719795448e-09,
+      "loss": 0.0223,
+      "step": 3469
+    },
+    {
+      "epoch": 0.9894496720844026,
+      "grad_norm": 1.0486040613170182,
+      "learning_rate": 2.920032344185253e-09,
+      "loss": 0.0154,
+      "step": 3470
+    },
+    {
+      "epoch": 0.9897348160821214,
+      "grad_norm": 1.226563276780761,
+      "learning_rate": 2.7643400745602787e-09,
+      "loss": 0.0185,
+      "step": 3471
+    },
+    {
+      "epoch": 0.9900199600798403,
+      "grad_norm": 1.1703998955142922,
+      "learning_rate": 2.612911795951689e-09,
+      "loss": 0.0124,
+      "step": 3472
+    },
+    {
+      "epoch": 0.9903051040775591,
+      "grad_norm": 0.6924588334370636,
+      "learning_rate": 2.465747637568905e-09,
+      "loss": 0.0075,
+      "step": 3473
+    },
+    {
+      "epoch": 0.9905902480752781,
+      "grad_norm": 0.9080422538502102,
+      "learning_rate": 2.322847724982591e-09,
+      "loss": 0.0115,
+      "step": 3474
+    },
+    {
+      "epoch": 0.9908753920729969,
+      "grad_norm": 0.8362530707299618,
+      "learning_rate": 2.1842121801257666e-09,
+      "loss": 0.0132,
+      "step": 3475
+    },
+    {
+      "epoch": 0.9911605360707157,
+      "grad_norm": 0.4226196218790714,
+      "learning_rate": 2.0498411212904746e-09,
+      "loss": 0.0058,
+      "step": 3476
+    },
+    {
+      "epoch": 0.9914456800684346,
+      "grad_norm": 1.8399346633917293,
+      "learning_rate": 1.9197346631327774e-09,
+      "loss": 0.0283,
+      "step": 3477
+    },
+    {
+      "epoch": 0.9917308240661534,
+      "grad_norm": 0.7267003973721158,
+      "learning_rate": 1.7938929166683161e-09,
+      "loss": 0.0095,
+      "step": 3478
+    },
+    {
+      "epoch": 0.9920159680638723,
+      "grad_norm": 1.8275007401191372,
+      "learning_rate": 1.6723159892734209e-09,
+      "loss": 0.0189,
+      "step": 3479
+    },
+    {
+      "epoch": 0.9923011120615911,
+      "grad_norm": 0.9793326780518042,
+      "learning_rate": 1.5550039846867758e-09,
+      "loss": 0.0273,
+      "step": 3480
+    },
+    {
+      "epoch": 0.99258625605931,
+      "grad_norm": 1.3801801510684655,
+      "learning_rate": 1.4419570030071995e-09,
+      "loss": 0.0356,
+      "step": 3481
+    },
+    {
+      "epoch": 0.9928714000570288,
+      "grad_norm": 1.5847937802859948,
+      "learning_rate": 1.3331751406936433e-09,
+      "loss": 0.0445,
+      "step": 3482
+    },
+    {
+      "epoch": 0.9931565440547476,
+      "grad_norm": 0.8944423135053186,
+      "learning_rate": 1.2286584905668587e-09,
+      "loss": 0.0219,
+      "step": 3483
+    },
+    {
+      "epoch": 0.9934416880524665,
+      "grad_norm": 1.4836875534178315,
+      "learning_rate": 1.1284071418077303e-09,
+      "loss": 0.0213,
+      "step": 3484
+    },
+    {
+      "epoch": 0.9937268320501853,
+      "grad_norm": 1.3626285135130423,
+      "learning_rate": 1.0324211799578321e-09,
+      "loss": 0.0259,
+      "step": 3485
+    },
+    {
+      "epoch": 0.9940119760479041,
+      "grad_norm": 0.47759862221040755,
+      "learning_rate": 9.407006869188718e-10,
+      "loss": 0.0055,
+      "step": 3486
+    },
+    {
+      "epoch": 0.9942971200456231,
+      "grad_norm": 0.5968701456927105,
+      "learning_rate": 8.532457409532457e-10,
+      "loss": 0.0082,
+      "step": 3487
+    },
+    {
+      "epoch": 0.9945822640433419,
+      "grad_norm": 0.5594488628362712,
+      "learning_rate": 7.700564166834845e-10,
+      "loss": 0.0097,
+      "step": 3488
+    },
+    {
+      "epoch": 0.9948674080410608,
+      "grad_norm": 1.836502535790717,
+      "learning_rate": 6.911327850928074e-10,
+      "loss": 0.0312,
+      "step": 3489
+    },
+    {
+      "epoch": 0.9951525520387796,
+      "grad_norm": 0.7196938824374762,
+      "learning_rate": 6.164749135240122e-10,
+      "loss": 0.0261,
+      "step": 3490
+    },
+    {
+      "epoch": 0.9954376960364985,
+      "grad_norm": 0.9928051386279392,
+      "learning_rate": 5.460828656811412e-10,
+      "loss": 0.0251,
+      "step": 3491
+    },
+    {
+      "epoch": 0.9957228400342173,
+      "grad_norm": 0.9905040540471107,
+      "learning_rate": 4.799567016267048e-10,
+      "loss": 0.0167,
+      "step": 3492
+    },
+    {
+      "epoch": 0.9960079840319361,
+      "grad_norm": 1.4772976262817499,
+      "learning_rate": 4.1809647778501274e-10,
+      "loss": 0.0144,
+      "step": 3493
+    },
+    {
+      "epoch": 0.996293128029655,
+      "grad_norm": 1.617884852382969,
+      "learning_rate": 3.605022469388431e-10,
+      "loss": 0.0367,
+      "step": 3494
+    },
+    {
+      "epoch": 0.9965782720273738,
+      "grad_norm": 1.9532885950196877,
+      "learning_rate": 3.07174058232218e-10,
+      "loss": 0.0321,
+      "step": 3495
+    },
+    {
+      "epoch": 0.9968634160250927,
+      "grad_norm": 1.21126863713605,
+      "learning_rate": 2.5811195716762827e-10,
+      "loss": 0.0241,
+      "step": 3496
+    },
+    {
+      "epoch": 0.9971485600228115,
+      "grad_norm": 0.8578791491329285,
+      "learning_rate": 2.133159856093636e-10,
+      "loss": 0.0158,
+      "step": 3497
+    },
+    {
+      "epoch": 0.9974337040205303,
+      "grad_norm": 0.8397533407417973,
+      "learning_rate": 1.7278618177962726e-10,
+      "loss": 0.0104,
+      "step": 3498
+    },
+    {
+      "epoch": 0.9977188480182492,
+      "grad_norm": 0.8670159276111888,
+      "learning_rate": 1.3652258026186638e-10,
+      "loss": 0.0264,
+      "step": 3499
+    },
+    {
+      "epoch": 0.998003992015968,
+      "grad_norm": 1.4392317009722384,
+      "learning_rate": 1.045252119979967e-10,
+      "loss": 0.0371,
+      "step": 3500
+    },
+    {
+      "epoch": 0.998289136013687,
+      "grad_norm": 0.8578595317936397,
+      "learning_rate": 7.679410429117795e-11,
+      "loss": 0.0353,
+      "step": 3501
+    },
+    {
+      "epoch": 0.9985742800114058,
+      "grad_norm": 0.43156184395377595,
+      "learning_rate": 5.332928080359345e-11,
+      "loss": 0.01,
+      "step": 3502
+    },
+    {
+      "epoch": 0.9988594240091246,
+      "grad_norm": 1.35487670264924,
+      "learning_rate": 3.413076155645012e-11,
+      "loss": 0.0233,
+      "step": 3503
+    },
+    {
+      "epoch": 0.9991445680068435,
+      "grad_norm": 0.8191645811893169,
+      "learning_rate": 1.9198562931088682e-11,
+      "loss": 0.0199,
+      "step": 3504
+    },
+    {
+      "epoch": 0.9994297120045623,
+      "grad_norm": 1.885688610242415,
+      "learning_rate": 8.532697669538791e-12,
+      "loss": 0.0381,
+      "step": 3505
+    },
+    {
+      "epoch": 0.9997148560022812,
+      "grad_norm": 1.563181683513951,
+      "learning_rate": 2.1331748722985823e-12,
+      "loss": 0.0318,
+      "step": 3506
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8918846626454352,
+      "learning_rate": 0.0,
+      "loss": 0.0097,
+      "step": 3507
+    },
+    {
+      "epoch": 1.0,
+      "step": 3507,
+      "total_flos": 28156764236288.0,
+      "train_loss": 0.031115454253810983,
+      "train_runtime": 8471.7741,
+      "train_samples_per_second": 3.312,
+      "train_steps_per_second": 0.414
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 3507,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1365,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 28156764236288.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}