diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9568 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9950467804072645,
+  "eval_steps": 500,
+  "global_step": 1362,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.002201430930104568,
+      "grad_norm": 0.7037750482559204,
+      "learning_rate": 0.0,
+      "loss": 0.308,
+      "step": 1
+    },
+    {
+      "epoch": 0.004402861860209136,
+      "grad_norm": 1.0037487745285034,
+      "learning_rate": 1.4598540145985402e-06,
+      "loss": 0.2687,
+      "step": 2
+    },
+    {
+      "epoch": 0.006604292790313704,
+      "grad_norm": 2.8608241081237793,
+      "learning_rate": 2.9197080291970804e-06,
+      "loss": 0.391,
+      "step": 3
+    },
+    {
+      "epoch": 0.008805723720418272,
+      "grad_norm": 0.7382011413574219,
+      "learning_rate": 4.379562043795621e-06,
+      "loss": 0.2279,
+      "step": 4
+    },
+    {
+      "epoch": 0.01100715465052284,
+      "grad_norm": 2.2300755977630615,
+      "learning_rate": 5.839416058394161e-06,
+      "loss": 0.425,
+      "step": 5
+    },
+    {
+      "epoch": 0.013208585580627407,
+      "grad_norm": 1.0738085508346558,
+      "learning_rate": 7.2992700729927e-06,
+      "loss": 0.386,
+      "step": 6
+    },
+    {
+      "epoch": 0.015410016510731976,
+      "grad_norm": 1.1076860427856445,
+      "learning_rate": 8.759124087591241e-06,
+      "loss": 0.4133,
+      "step": 7
+    },
+    {
+      "epoch": 0.017611447440836543,
+      "grad_norm": 0.9959158897399902,
+      "learning_rate": 1.0218978102189781e-05,
+      "loss": 0.2856,
+      "step": 8
+    },
+    {
+      "epoch": 0.01981287837094111,
+      "grad_norm": 1.159873604774475,
+      "learning_rate": 1.1678832116788322e-05,
+      "loss": 0.302,
+      "step": 9
+    },
+    {
+      "epoch": 0.02201430930104568,
+      "grad_norm": 1.0309239625930786,
+      "learning_rate": 1.3138686131386862e-05,
+      "loss": 0.2593,
+      "step": 10
+    },
+    {
+      "epoch": 0.024215740231150248,
+      "grad_norm": 0.944840669631958,
+      "learning_rate": 1.45985401459854e-05,
+      "loss": 0.3306,
+      "step": 11
+    },
+    {
+      "epoch": 0.026417171161254815,
+      "grad_norm": 5.096871376037598,
+      "learning_rate": 1.605839416058394e-05,
+      "loss": 0.5597,
+      "step": 12
+    },
+    {
+      "epoch": 0.028618602091359385,
+      "grad_norm": 0.9964978694915771,
+      "learning_rate": 1.7518248175182482e-05,
+      "loss": 0.2529,
+      "step": 13
+    },
+    {
+      "epoch": 0.030820033021463952,
+      "grad_norm": 1.1680715084075928,
+      "learning_rate": 1.897810218978102e-05,
+      "loss": 0.2551,
+      "step": 14
+    },
+    {
+      "epoch": 0.03302146395156852,
+      "grad_norm": 1.3312640190124512,
+      "learning_rate": 2.0437956204379563e-05,
+      "loss": 0.2783,
+      "step": 15
+    },
+    {
+      "epoch": 0.035222894881673086,
+      "grad_norm": 1.0938522815704346,
+      "learning_rate": 2.1897810218978105e-05,
+      "loss": 0.2196,
+      "step": 16
+    },
+    {
+      "epoch": 0.03742432581177765,
+      "grad_norm": 0.844025731086731,
+      "learning_rate": 2.3357664233576643e-05,
+      "loss": 0.223,
+      "step": 17
+    },
+    {
+      "epoch": 0.03962575674188222,
+      "grad_norm": 1.0622016191482544,
+      "learning_rate": 2.4817518248175185e-05,
+      "loss": 0.4032,
+      "step": 18
+    },
+    {
+      "epoch": 0.041827187671986794,
+      "grad_norm": 1.267992377281189,
+      "learning_rate": 2.6277372262773724e-05,
+      "loss": 0.3885,
+      "step": 19
+    },
+    {
+      "epoch": 0.04402861860209136,
+      "grad_norm": 0.6925700902938843,
+      "learning_rate": 2.7737226277372262e-05,
+      "loss": 0.282,
+      "step": 20
+    },
+    {
+      "epoch": 0.04623004953219593,
+      "grad_norm": 0.49505650997161865,
+      "learning_rate": 2.91970802919708e-05,
+      "loss": 0.1894,
+      "step": 21
+    },
+    {
+      "epoch": 0.048431480462300495,
+      "grad_norm": 1.3833006620407104,
+      "learning_rate": 3.0656934306569346e-05,
+      "loss": 0.2406,
+      "step": 22
+    },
+    {
+      "epoch": 0.05063291139240506,
+      "grad_norm": 5.850524425506592,
+      "learning_rate": 3.211678832116788e-05,
+      "loss": 0.3697,
+      "step": 23
+    },
+    {
+      "epoch": 0.05283434232250963,
+      "grad_norm": 0.6721429824829102,
+      "learning_rate": 3.357664233576642e-05,
+      "loss": 0.1444,
+      "step": 24
+    },
+    {
+      "epoch": 0.0550357732526142,
+      "grad_norm": 1.379242181777954,
+      "learning_rate": 3.5036496350364965e-05,
+      "loss": 0.188,
+      "step": 25
+    },
+    {
+      "epoch": 0.05723720418271877,
+      "grad_norm": 1.0648667812347412,
+      "learning_rate": 3.649635036496351e-05,
+      "loss": 0.1628,
+      "step": 26
+    },
+    {
+      "epoch": 0.05943863511282334,
+      "grad_norm": 1.553063154220581,
+      "learning_rate": 3.795620437956204e-05,
+      "loss": 0.4074,
+      "step": 27
+    },
+    {
+      "epoch": 0.061640066042927905,
+      "grad_norm": 2.283454418182373,
+      "learning_rate": 3.941605839416059e-05,
+      "loss": 0.3169,
+      "step": 28
+    },
+    {
+      "epoch": 0.06384149697303247,
+      "grad_norm": 2.135236978530884,
+      "learning_rate": 4.0875912408759126e-05,
+      "loss": 0.3472,
+      "step": 29
+    },
+    {
+      "epoch": 0.06604292790313704,
+      "grad_norm": 7.040278911590576,
+      "learning_rate": 4.233576642335767e-05,
+      "loss": 0.2676,
+      "step": 30
+    },
+    {
+      "epoch": 0.0682443588332416,
+      "grad_norm": 1.1241406202316284,
+      "learning_rate": 4.379562043795621e-05,
+      "loss": 0.4671,
+      "step": 31
+    },
+    {
+      "epoch": 0.07044578976334617,
+      "grad_norm": 1.736489176750183,
+      "learning_rate": 4.5255474452554745e-05,
+      "loss": 0.2511,
+      "step": 32
+    },
+    {
+      "epoch": 0.07264722069345074,
+      "grad_norm": 1.144145131111145,
+      "learning_rate": 4.6715328467153287e-05,
+      "loss": 0.226,
+      "step": 33
+    },
+    {
+      "epoch": 0.0748486516235553,
+      "grad_norm": 0.8320156931877136,
+      "learning_rate": 4.817518248175183e-05,
+      "loss": 0.1128,
+      "step": 34
+    },
+    {
+      "epoch": 0.07705008255365987,
+      "grad_norm": 1.2651524543762207,
+      "learning_rate": 4.963503649635037e-05,
+      "loss": 0.1757,
+      "step": 35
+    },
+    {
+      "epoch": 0.07925151348376444,
+      "grad_norm": 0.9096120595932007,
+      "learning_rate": 5.109489051094891e-05,
+      "loss": 0.1775,
+      "step": 36
+    },
+    {
+      "epoch": 0.08145294441386902,
+      "grad_norm": 0.9276086091995239,
+      "learning_rate": 5.255474452554745e-05,
+      "loss": 0.1733,
+      "step": 37
+    },
+    {
+      "epoch": 0.08365437534397359,
+      "grad_norm": 0.9676797389984131,
+      "learning_rate": 5.401459854014599e-05,
+      "loss": 0.1706,
+      "step": 38
+    },
+    {
+      "epoch": 0.08585580627407816,
+      "grad_norm": 1.0610243082046509,
+      "learning_rate": 5.5474452554744524e-05,
+      "loss": 0.1549,
+      "step": 39
+    },
+    {
+      "epoch": 0.08805723720418272,
+      "grad_norm": 1.3586833477020264,
+      "learning_rate": 5.6934306569343066e-05,
+      "loss": 0.1767,
+      "step": 40
+    },
+    {
+      "epoch": 0.09025866813428729,
+      "grad_norm": 1.7117362022399902,
+      "learning_rate": 5.83941605839416e-05,
+      "loss": 0.2045,
+      "step": 41
+    },
+    {
+      "epoch": 0.09246009906439186,
+      "grad_norm": 1.0564417839050293,
+      "learning_rate": 5.985401459854016e-05,
+      "loss": 0.2057,
+      "step": 42
+    },
+    {
+      "epoch": 0.09466152999449642,
+      "grad_norm": 0.84938645362854,
+      "learning_rate": 6.131386861313869e-05,
+      "loss": 0.1064,
+      "step": 43
+    },
+    {
+      "epoch": 0.09686296092460099,
+      "grad_norm": 1.660563349723816,
+      "learning_rate": 6.277372262773723e-05,
+      "loss": 0.154,
+      "step": 44
+    },
+    {
+      "epoch": 0.09906439185470556,
+      "grad_norm": 2.250649929046631,
+      "learning_rate": 6.423357664233576e-05,
+      "loss": 0.29,
+      "step": 45
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 1.0987548828125,
+      "learning_rate": 6.56934306569343e-05,
+      "loss": 0.1161,
+      "step": 46
+    },
+    {
+      "epoch": 0.10346725371491469,
+      "grad_norm": 1.6264435052871704,
+      "learning_rate": 6.715328467153285e-05,
+      "loss": 0.3335,
+      "step": 47
+    },
+    {
+      "epoch": 0.10566868464501926,
+      "grad_norm": 1.36955726146698,
+      "learning_rate": 6.86131386861314e-05,
+      "loss": 0.1541,
+      "step": 48
+    },
+    {
+      "epoch": 0.10787011557512383,
+      "grad_norm": 1.5003987550735474,
+      "learning_rate": 7.007299270072993e-05,
+      "loss": 0.4296,
+      "step": 49
+    },
+    {
+      "epoch": 0.1100715465052284,
+      "grad_norm": 0.9643859267234802,
+      "learning_rate": 7.153284671532847e-05,
+      "loss": 0.1276,
+      "step": 50
+    },
+    {
+      "epoch": 0.11227297743533296,
+      "grad_norm": 0.8790437579154968,
+      "learning_rate": 7.299270072992701e-05,
+      "loss": 0.0857,
+      "step": 51
+    },
+    {
+      "epoch": 0.11447440836543754,
+      "grad_norm": 1.8895280361175537,
+      "learning_rate": 7.445255474452556e-05,
+      "loss": 0.3803,
+      "step": 52
+    },
+    {
+      "epoch": 0.11667583929554211,
+      "grad_norm": 0.6855382323265076,
+      "learning_rate": 7.591240875912408e-05,
+      "loss": 0.1926,
+      "step": 53
+    },
+    {
+      "epoch": 0.11887727022564668,
+      "grad_norm": 1.257523536682129,
+      "learning_rate": 7.737226277372264e-05,
+      "loss": 0.1345,
+      "step": 54
+    },
+    {
+      "epoch": 0.12107870115575124,
+      "grad_norm": 1.2312171459197998,
+      "learning_rate": 7.883211678832118e-05,
+      "loss": 0.3421,
+      "step": 55
+    },
+    {
+      "epoch": 0.12328013208585581,
+      "grad_norm": 1.0815132856369019,
+      "learning_rate": 8.029197080291971e-05,
+      "loss": 0.2859,
+      "step": 56
+    },
+    {
+      "epoch": 0.12548156301596036,
+      "grad_norm": 0.5844975709915161,
+      "learning_rate": 8.175182481751825e-05,
+      "loss": 0.1157,
+      "step": 57
+    },
+    {
+      "epoch": 0.12768299394606494,
+      "grad_norm": 1.176499605178833,
+      "learning_rate": 8.32116788321168e-05,
+      "loss": 0.2674,
+      "step": 58
+    },
+    {
+      "epoch": 0.1298844248761695,
+      "grad_norm": 0.8862079977989197,
+      "learning_rate": 8.467153284671534e-05,
+      "loss": 0.1591,
+      "step": 59
+    },
+    {
+      "epoch": 0.13208585580627408,
+      "grad_norm": 0.5914232134819031,
+      "learning_rate": 8.613138686131386e-05,
+      "loss": 0.0925,
+      "step": 60
+    },
+    {
+      "epoch": 0.13428728673637866,
+      "grad_norm": 0.5085222125053406,
+      "learning_rate": 8.759124087591242e-05,
+      "loss": 0.1197,
+      "step": 61
+    },
+    {
+      "epoch": 0.1364887176664832,
+      "grad_norm": 0.5677929520606995,
+      "learning_rate": 8.905109489051096e-05,
+      "loss": 0.1964,
+      "step": 62
+    },
+    {
+      "epoch": 0.1386901485965878,
+      "grad_norm": 0.7557106614112854,
+      "learning_rate": 9.051094890510949e-05,
+      "loss": 0.1455,
+      "step": 63
+    },
+    {
+      "epoch": 0.14089157952669235,
+      "grad_norm": 0.962935745716095,
+      "learning_rate": 9.197080291970803e-05,
+      "loss": 0.3851,
+      "step": 64
+    },
+    {
+      "epoch": 0.14309301045679693,
+      "grad_norm": 1.9366933107376099,
+      "learning_rate": 9.343065693430657e-05,
+      "loss": 0.2542,
+      "step": 65
+    },
+    {
+      "epoch": 0.14529444138690148,
+      "grad_norm": 1.4374990463256836,
+      "learning_rate": 9.489051094890511e-05,
+      "loss": 0.2277,
+      "step": 66
+    },
+    {
+      "epoch": 0.14749587231700606,
+      "grad_norm": 1.0540611743927002,
+      "learning_rate": 9.635036496350366e-05,
+      "loss": 0.2025,
+      "step": 67
+    },
+    {
+      "epoch": 0.1496973032471106,
+      "grad_norm": 1.1396342515945435,
+      "learning_rate": 9.78102189781022e-05,
+      "loss": 0.1476,
+      "step": 68
+    },
+    {
+      "epoch": 0.1518987341772152,
+      "grad_norm": 0.7722952365875244,
+      "learning_rate": 9.927007299270074e-05,
+      "loss": 0.0759,
+      "step": 69
+    },
+    {
+      "epoch": 0.15410016510731975,
+      "grad_norm": 1.0545134544372559,
+      "learning_rate": 0.00010072992700729928,
+      "loss": 0.1877,
+      "step": 70
+    },
+    {
+      "epoch": 0.15630159603742433,
+      "grad_norm": 0.7256941795349121,
+      "learning_rate": 0.00010218978102189782,
+      "loss": 0.2213,
+      "step": 71
+    },
+    {
+      "epoch": 0.15850302696752888,
+      "grad_norm": 0.8643118143081665,
+      "learning_rate": 0.00010364963503649637,
+      "loss": 0.1217,
+      "step": 72
+    },
+    {
+      "epoch": 0.16070445789763346,
+      "grad_norm": 1.3518829345703125,
+      "learning_rate": 0.0001051094890510949,
+      "loss": 0.1657,
+      "step": 73
+    },
+    {
+      "epoch": 0.16290588882773804,
+      "grad_norm": 1.2510859966278076,
+      "learning_rate": 0.00010656934306569344,
+      "loss": 0.3364,
+      "step": 74
+    },
+    {
+      "epoch": 0.1651073197578426,
+      "grad_norm": 0.7721868753433228,
+      "learning_rate": 0.00010802919708029198,
+      "loss": 0.1296,
+      "step": 75
+    },
+    {
+      "epoch": 0.16730875068794718,
+      "grad_norm": 1.2092798948287964,
+      "learning_rate": 0.00010948905109489052,
+      "loss": 0.183,
+      "step": 76
+    },
+    {
+      "epoch": 0.16951018161805173,
+      "grad_norm": 0.6521322727203369,
+      "learning_rate": 0.00011094890510948905,
+      "loss": 0.129,
+      "step": 77
+    },
+    {
+      "epoch": 0.1717116125481563,
+      "grad_norm": 1.9378713369369507,
+      "learning_rate": 0.00011240875912408759,
+      "loss": 0.2796,
+      "step": 78
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 1.120902180671692,
+      "learning_rate": 0.00011386861313868613,
+      "loss": 0.1491,
+      "step": 79
+    },
+    {
+      "epoch": 0.17611447440836545,
+      "grad_norm": 1.9726943969726562,
+      "learning_rate": 0.00011532846715328467,
+      "loss": 0.3284,
+      "step": 80
+    },
+    {
+      "epoch": 0.17831590533847,
+      "grad_norm": 0.799227237701416,
+      "learning_rate": 0.0001167883211678832,
+      "loss": 0.1668,
+      "step": 81
+    },
+    {
+      "epoch": 0.18051733626857458,
+      "grad_norm": 1.8196526765823364,
+      "learning_rate": 0.00011824817518248177,
+      "loss": 0.4528,
+      "step": 82
+    },
+    {
+      "epoch": 0.18271876719867913,
+      "grad_norm": 0.9238607883453369,
+      "learning_rate": 0.00011970802919708031,
+      "loss": 0.2334,
+      "step": 83
+    },
+    {
+      "epoch": 0.1849201981287837,
+      "grad_norm": 0.9001494646072388,
+      "learning_rate": 0.00012116788321167884,
+      "loss": 0.1732,
+      "step": 84
+    },
+    {
+      "epoch": 0.18712162905888827,
+      "grad_norm": 0.7415959239006042,
+      "learning_rate": 0.00012262773722627738,
+      "loss": 0.0985,
+      "step": 85
+    },
+    {
+      "epoch": 0.18932305998899285,
+      "grad_norm": 0.9463704824447632,
+      "learning_rate": 0.0001240875912408759,
+      "loss": 0.151,
+      "step": 86
+    },
+    {
+      "epoch": 0.1915244909190974,
+      "grad_norm": 0.735754132270813,
+      "learning_rate": 0.00012554744525547447,
+      "loss": 0.095,
+      "step": 87
+    },
+    {
+      "epoch": 0.19372592184920198,
+      "grad_norm": 0.9355161786079407,
+      "learning_rate": 0.000127007299270073,
+      "loss": 0.2339,
+      "step": 88
+    },
+    {
+      "epoch": 0.19592735277930656,
+      "grad_norm": 1.4002137184143066,
+      "learning_rate": 0.00012846715328467152,
+      "loss": 0.2057,
+      "step": 89
+    },
+    {
+      "epoch": 0.19812878370941112,
+      "grad_norm": 1.038177251815796,
+      "learning_rate": 0.00012992700729927008,
+      "loss": 0.1742,
+      "step": 90
+    },
+    {
+      "epoch": 0.2003302146395157,
+      "grad_norm": 0.9062851071357727,
+      "learning_rate": 0.0001313868613138686,
+      "loss": 0.134,
+      "step": 91
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 1.0446114540100098,
+      "learning_rate": 0.00013284671532846716,
+      "loss": 0.1967,
+      "step": 92
+    },
+    {
+      "epoch": 0.20473307649972483,
+      "grad_norm": 1.0713139772415161,
+      "learning_rate": 0.0001343065693430657,
+      "loss": 0.2289,
+      "step": 93
+    },
+    {
+      "epoch": 0.20693450742982938,
+      "grad_norm": 0.8900029063224792,
+      "learning_rate": 0.00013576642335766422,
+      "loss": 0.1469,
+      "step": 94
+    },
+    {
+      "epoch": 0.20913593835993396,
+      "grad_norm": 0.615839958190918,
+      "learning_rate": 0.0001372262773722628,
+      "loss": 0.2123,
+      "step": 95
+    },
+    {
+      "epoch": 0.21133736929003852,
+      "grad_norm": 0.9755569696426392,
+      "learning_rate": 0.00013868613138686133,
+      "loss": 0.1719,
+      "step": 96
+    },
+    {
+      "epoch": 0.2135388002201431,
+      "grad_norm": 0.7173188328742981,
+      "learning_rate": 0.00014014598540145986,
+      "loss": 0.2311,
+      "step": 97
+    },
+    {
+      "epoch": 0.21574023115024765,
+      "grad_norm": 2.7052557468414307,
+      "learning_rate": 0.00014160583941605841,
+      "loss": 0.2516,
+      "step": 98
+    },
+    {
+      "epoch": 0.21794166208035223,
+      "grad_norm": 1.7500773668289185,
+      "learning_rate": 0.00014306569343065694,
+      "loss": 0.2488,
+      "step": 99
+    },
+    {
+      "epoch": 0.2201430930104568,
+      "grad_norm": 1.7758418321609497,
+      "learning_rate": 0.00014452554744525547,
+      "loss": 0.4666,
+      "step": 100
+    },
+    {
+      "epoch": 0.22234452394056137,
+      "grad_norm": 1.0867117643356323,
+      "learning_rate": 0.00014598540145985403,
+      "loss": 0.3035,
+      "step": 101
+    },
+    {
+      "epoch": 0.22454595487066592,
+      "grad_norm": 1.184135913848877,
+      "learning_rate": 0.00014744525547445256,
+      "loss": 0.175,
+      "step": 102
+    },
+    {
+      "epoch": 0.2267473858007705,
+      "grad_norm": 0.8942990303039551,
+      "learning_rate": 0.0001489051094890511,
+      "loss": 0.1176,
+      "step": 103
+    },
+    {
+      "epoch": 0.22894881673087508,
+      "grad_norm": 1.145623803138733,
+      "learning_rate": 0.00015036496350364964,
+      "loss": 0.1915,
+      "step": 104
+    },
+    {
+      "epoch": 0.23115024766097964,
+      "grad_norm": 1.5253881216049194,
+      "learning_rate": 0.00015182481751824817,
+      "loss": 0.3467,
+      "step": 105
+    },
+    {
+      "epoch": 0.23335167859108422,
+      "grad_norm": 1.0943421125411987,
+      "learning_rate": 0.00015328467153284672,
+      "loss": 0.2571,
+      "step": 106
+    },
+    {
+      "epoch": 0.23555310952118877,
+      "grad_norm": 1.0738698244094849,
+      "learning_rate": 0.00015474452554744528,
+      "loss": 0.1974,
+      "step": 107
+    },
+    {
+      "epoch": 0.23775454045129335,
+      "grad_norm": 0.8418899178504944,
+      "learning_rate": 0.0001562043795620438,
+      "loss": 0.1585,
+      "step": 108
+    },
+    {
+      "epoch": 0.2399559713813979,
+      "grad_norm": 2.6217808723449707,
+      "learning_rate": 0.00015766423357664236,
+      "loss": 0.4359,
+      "step": 109
+    },
+    {
+      "epoch": 0.24215740231150248,
+      "grad_norm": 0.8705646395683289,
+      "learning_rate": 0.0001591240875912409,
+      "loss": 0.2176,
+      "step": 110
+    },
+    {
+      "epoch": 0.24435883324160704,
+      "grad_norm": 0.8313264846801758,
+      "learning_rate": 0.00016058394160583942,
+      "loss": 0.1573,
+      "step": 111
+    },
+    {
+      "epoch": 0.24656026417171162,
+      "grad_norm": 1.1230180263519287,
+      "learning_rate": 0.00016204379562043797,
+      "loss": 0.2631,
+      "step": 112
+    },
+    {
+      "epoch": 0.24876169510181617,
+      "grad_norm": 1.0018892288208008,
+      "learning_rate": 0.0001635036496350365,
+      "loss": 0.1711,
+      "step": 113
+    },
+    {
+      "epoch": 0.2509631260319207,
+      "grad_norm": 0.5920968055725098,
+      "learning_rate": 0.00016496350364963503,
+      "loss": 0.1189,
+      "step": 114
+    },
+    {
+      "epoch": 0.25316455696202533,
+      "grad_norm": 1.4356651306152344,
+      "learning_rate": 0.0001664233576642336,
+      "loss": 0.2634,
+      "step": 115
+    },
+    {
+      "epoch": 0.2553659878921299,
+      "grad_norm": 1.3919761180877686,
+      "learning_rate": 0.00016788321167883211,
+      "loss": 0.2505,
+      "step": 116
+    },
+    {
+      "epoch": 0.25756741882223444,
+      "grad_norm": 0.9886398911476135,
+      "learning_rate": 0.00016934306569343067,
+      "loss": 0.1908,
+      "step": 117
+    },
+    {
+      "epoch": 0.259768849752339,
+      "grad_norm": 0.9664818644523621,
+      "learning_rate": 0.0001708029197080292,
+      "loss": 0.1563,
+      "step": 118
+    },
+    {
+      "epoch": 0.2619702806824436,
+      "grad_norm": 0.9564926624298096,
+      "learning_rate": 0.00017226277372262773,
+      "loss": 0.1542,
+      "step": 119
+    },
+    {
+      "epoch": 0.26417171161254815,
+      "grad_norm": 0.613195538520813,
+      "learning_rate": 0.00017372262773722628,
+      "loss": 0.0748,
+      "step": 120
+    },
+    {
+      "epoch": 0.2663731425426527,
+      "grad_norm": 1.116240382194519,
+      "learning_rate": 0.00017518248175182484,
+      "loss": 0.2469,
+      "step": 121
+    },
+    {
+      "epoch": 0.2685745734727573,
+      "grad_norm": 0.6556792259216309,
+      "learning_rate": 0.00017664233576642337,
+      "loss": 0.1449,
+      "step": 122
+    },
+    {
+      "epoch": 0.27077600440286187,
+      "grad_norm": 1.6238877773284912,
+      "learning_rate": 0.00017810218978102192,
+      "loss": 0.2584,
+      "step": 123
+    },
+    {
+      "epoch": 0.2729774353329664,
+      "grad_norm": 1.42080819606781,
+      "learning_rate": 0.00017956204379562045,
+      "loss": 0.3169,
+      "step": 124
+    },
+    {
+      "epoch": 0.275178866263071,
+      "grad_norm": 0.721822202205658,
+      "learning_rate": 0.00018102189781021898,
+      "loss": 0.1188,
+      "step": 125
+    },
+    {
+      "epoch": 0.2773802971931756,
+      "grad_norm": 1.00337553024292,
+      "learning_rate": 0.00018248175182481753,
+      "loss": 0.2328,
+      "step": 126
+    },
+    {
+      "epoch": 0.27958172812328014,
+      "grad_norm": 1.1204198598861694,
+      "learning_rate": 0.00018394160583941606,
+      "loss": 0.191,
+      "step": 127
+    },
+    {
+      "epoch": 0.2817831590533847,
+      "grad_norm": 0.7608072757720947,
+      "learning_rate": 0.0001854014598540146,
+      "loss": 0.1397,
+      "step": 128
+    },
+    {
+      "epoch": 0.28398458998348924,
+      "grad_norm": 0.9406429529190063,
+      "learning_rate": 0.00018686131386861315,
+      "loss": 0.3071,
+      "step": 129
+    },
+    {
+      "epoch": 0.28618602091359385,
+      "grad_norm": 0.8172856569290161,
+      "learning_rate": 0.00018832116788321167,
+      "loss": 0.1572,
+      "step": 130
+    },
+    {
+      "epoch": 0.2883874518436984,
+      "grad_norm": 0.7377148866653442,
+      "learning_rate": 0.00018978102189781023,
+      "loss": 0.2005,
+      "step": 131
+    },
+    {
+      "epoch": 0.29058888277380296,
+      "grad_norm": 1.500381588935852,
+      "learning_rate": 0.00019124087591240879,
+      "loss": 0.2424,
+      "step": 132
+    },
+    {
+      "epoch": 0.29279031370390757,
+      "grad_norm": 1.5137453079223633,
+      "learning_rate": 0.00019270072992700731,
+      "loss": 0.1776,
+      "step": 133
+    },
+    {
+      "epoch": 0.2949917446340121,
+      "grad_norm": 1.6952030658721924,
+      "learning_rate": 0.00019416058394160587,
+      "loss": 0.1664,
+      "step": 134
+    },
+    {
+      "epoch": 0.2971931755641167,
+      "grad_norm": 1.0178738832473755,
+      "learning_rate": 0.0001956204379562044,
+      "loss": 0.1929,
+      "step": 135
+    },
+    {
+      "epoch": 0.2993946064942212,
+      "grad_norm": 2.562960624694824,
+      "learning_rate": 0.00019708029197080293,
+      "loss": 0.2782,
+      "step": 136
+    },
+    {
+      "epoch": 0.30159603742432584,
+      "grad_norm": 1.6147712469100952,
+      "learning_rate": 0.00019854014598540148,
+      "loss": 0.2905,
+      "step": 137
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 1.1269806623458862,
+      "learning_rate": 0.0002,
+      "loss": 0.2205,
+      "step": 138
+    },
+    {
+      "epoch": 0.30599889928453494,
+      "grad_norm": 1.709218978881836,
+      "learning_rate": 0.00019999967115038762,
+      "loss": 0.203,
+      "step": 139
+    },
+    {
+      "epoch": 0.3082003302146395,
+      "grad_norm": 1.1186352968215942,
+      "learning_rate": 0.00019999868460371332,
+      "loss": 0.2359,
+      "step": 140
+    },
+    {
+      "epoch": 0.3104017611447441,
+      "grad_norm": 2.0815446376800537,
+      "learning_rate": 0.00019999704036646558,
+      "loss": 0.7654,
+      "step": 141
+    },
+    {
+      "epoch": 0.31260319207484866,
+      "grad_norm": 1.123903751373291,
+      "learning_rate": 0.00019999473844945856,
+      "loss": 0.1531,
+      "step": 142
+    },
+    {
+      "epoch": 0.3148046230049532,
+      "grad_norm": 1.1775410175323486,
+      "learning_rate": 0.00019999177886783194,
+      "loss": 0.2142,
+      "step": 143
+    },
+    {
+      "epoch": 0.31700605393505776,
+      "grad_norm": 1.0821821689605713,
+      "learning_rate": 0.00019998816164105087,
+      "loss": 0.194,
+      "step": 144
+    },
+    {
+      "epoch": 0.31920748486516237,
+      "grad_norm": 1.846646785736084,
+      "learning_rate": 0.00019998388679290583,
+      "loss": 0.421,
+      "step": 145
+    },
+    {
+      "epoch": 0.3214089157952669,
+      "grad_norm": 0.9930664300918579,
+      "learning_rate": 0.00019997895435151246,
+      "loss": 0.2811,
+      "step": 146
+    },
+    {
+      "epoch": 0.3236103467253715,
+      "grad_norm": 0.9030304551124573,
+      "learning_rate": 0.00019997336434931137,
+      "loss": 0.1815,
+      "step": 147
+    },
+    {
+      "epoch": 0.3258117776554761,
+      "grad_norm": 0.6974136829376221,
+      "learning_rate": 0.000199967116823068,
+      "loss": 0.1949,
+      "step": 148
+    },
+    {
+      "epoch": 0.32801320858558064,
+      "grad_norm": 0.8216617107391357,
+      "learning_rate": 0.00019996021181387224,
+      "loss": 0.2127,
+      "step": 149
+    },
+    {
+      "epoch": 0.3302146395156852,
+      "grad_norm": 1.0203906297683716,
+      "learning_rate": 0.00019995264936713828,
+      "loss": 0.2603,
+      "step": 150
+    },
+    {
+      "epoch": 0.33241607044578975,
+      "grad_norm": 1.0693780183792114,
+      "learning_rate": 0.00019994442953260433,
+      "loss": 0.258,
+      "step": 151
+    },
+    {
+      "epoch": 0.33461750137589436,
+      "grad_norm": 1.3966225385665894,
+      "learning_rate": 0.00019993555236433213,
+      "loss": 0.2827,
+      "step": 152
+    },
+    {
+      "epoch": 0.3368189323059989,
+      "grad_norm": 0.5770395994186401,
+      "learning_rate": 0.00019992601792070679,
+      "loss": 0.1481,
+      "step": 153
+    },
+    {
+      "epoch": 0.33902036323610346,
+      "grad_norm": 1.5406584739685059,
+      "learning_rate": 0.00019991582626443618,
+      "loss": 0.1926,
+      "step": 154
+    },
+    {
+      "epoch": 0.341221794166208,
+      "grad_norm": 1.5380587577819824,
+      "learning_rate": 0.00019990497746255086,
+      "loss": 0.2582,
+      "step": 155
+    },
+    {
+      "epoch": 0.3434232250963126,
+      "grad_norm": 0.7996829748153687,
+      "learning_rate": 0.00019989347158640323,
+      "loss": 0.2201,
+      "step": 156
+    },
+    {
+      "epoch": 0.3456246560264172,
+      "grad_norm": 1.2687666416168213,
+      "learning_rate": 0.00019988130871166737,
+      "loss": 0.2027,
+      "step": 157
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.9393393993377686,
+      "learning_rate": 0.00019986848891833845,
+      "loss": 0.2079,
+      "step": 158
+    },
+    {
+      "epoch": 0.3500275178866263,
+      "grad_norm": 1.0372052192687988,
+      "learning_rate": 0.0001998550122907321,
+      "loss": 0.1452,
+      "step": 159
+    },
+    {
+      "epoch": 0.3522289488167309,
+      "grad_norm": 1.1515039205551147,
+      "learning_rate": 0.00019984087891748402,
+      "loss": 0.2363,
+      "step": 160
+    },
+    {
+      "epoch": 0.35443037974683544,
+      "grad_norm": 2.2971878051757812,
+      "learning_rate": 0.0001998260888915493,
+      "loss": 0.5014,
+      "step": 161
+    },
+    {
+      "epoch": 0.35663181067694,
+      "grad_norm": 1.187995433807373,
+      "learning_rate": 0.00019981064231020182,
+      "loss": 0.2729,
+      "step": 162
+    },
+    {
+      "epoch": 0.3588332416070446,
+      "grad_norm": 1.2573753595352173,
+      "learning_rate": 0.00019979453927503364,
+      "loss": 0.2384,
+      "step": 163
+    },
+    {
+      "epoch": 0.36103467253714916,
+      "grad_norm": 1.0439549684524536,
+      "learning_rate": 0.0001997777798919543,
+      "loss": 0.2956,
+      "step": 164
+    },
+    {
+      "epoch": 0.3632361034672537,
+      "grad_norm": 1.4184736013412476,
+      "learning_rate": 0.00019976036427119006,
+      "loss": 0.2207,
+      "step": 165
+    },
+    {
+      "epoch": 0.36543753439735827,
+      "grad_norm": 1.2747489213943481,
+      "learning_rate": 0.00019974229252728342,
+      "loss": 0.2199,
+      "step": 166
+    },
+    {
+      "epoch": 0.3676389653274629,
+      "grad_norm": 0.513881266117096,
+      "learning_rate": 0.00019972356477909205,
+      "loss": 0.113,
+      "step": 167
+    },
+    {
+      "epoch": 0.3698403962575674,
+      "grad_norm": 1.0329495668411255,
+      "learning_rate": 0.0001997041811497882,
+      "loss": 0.2586,
+      "step": 168
+    },
+    {
+      "epoch": 0.372041827187672,
+      "grad_norm": 0.6814864277839661,
+      "learning_rate": 0.00019968414176685788,
+      "loss": 0.1375,
+      "step": 169
+    },
+    {
+      "epoch": 0.37424325811777653,
+      "grad_norm": 1.2967740297317505,
+      "learning_rate": 0.0001996634467620999,
+      "loss": 0.201,
+      "step": 170
+    },
+    {
+      "epoch": 0.37644468904788114,
+      "grad_norm": 2.24857759475708,
+      "learning_rate": 0.00019964209627162526,
+      "loss": 0.2661,
+      "step": 171
+    },
+    {
+      "epoch": 0.3786461199779857,
+      "grad_norm": 1.3992667198181152,
+      "learning_rate": 0.00019962009043585584,
+      "loss": 0.0903,
+      "step": 172
+    },
+    {
+      "epoch": 0.38084755090809025,
+      "grad_norm": 2.2193541526794434,
+      "learning_rate": 0.00019959742939952392,
+      "loss": 0.2786,
+      "step": 173
+    },
+    {
+      "epoch": 0.3830489818381948,
+      "grad_norm": 1.2357624769210815,
+      "learning_rate": 0.00019957411331167094,
+      "loss": 0.3981,
+      "step": 174
+    },
+    {
+      "epoch": 0.3852504127682994,
+      "grad_norm": 0.8800520300865173,
+      "learning_rate": 0.00019955014232564663,
+      "loss": 0.1451,
+      "step": 175
+    },
+    {
+      "epoch": 0.38745184369840396,
+      "grad_norm": 0.7195889949798584,
+      "learning_rate": 0.000199525516599108,
+      "loss": 0.0959,
+      "step": 176
+    },
+    {
+      "epoch": 0.3896532746285085,
+      "grad_norm": 2.2375988960266113,
+      "learning_rate": 0.00019950023629401823,
+      "loss": 0.3118,
+      "step": 177
+    },
+    {
+      "epoch": 0.3918547055586131,
+      "grad_norm": 1.8316059112548828,
+      "learning_rate": 0.00019947430157664576,
+      "loss": 0.2448,
+      "step": 178
+    },
+    {
+      "epoch": 0.3940561364887177,
+      "grad_norm": 1.2774708271026611,
+      "learning_rate": 0.00019944771261756292,
+      "loss": 0.1959,
+      "step": 179
+    },
+    {
+      "epoch": 0.39625756741882223,
+      "grad_norm": 1.0134649276733398,
+      "learning_rate": 0.00019942046959164515,
+      "loss": 0.2163,
+      "step": 180
+    },
+    {
+      "epoch": 0.3984589983489268,
+      "grad_norm": 1.4235607385635376,
+      "learning_rate": 0.00019939257267806963,
+      "loss": 0.3726,
+      "step": 181
+    },
+    {
+      "epoch": 0.4006604292790314,
+      "grad_norm": 0.4046984910964966,
+      "learning_rate": 0.00019936402206031413,
+      "loss": 0.0713,
+      "step": 182
+    },
+    {
+      "epoch": 0.40286186020913595,
+      "grad_norm": 1.2674638032913208,
+      "learning_rate": 0.00019933481792615583,
+      "loss": 0.2618,
+      "step": 183
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 1.141114592552185,
+      "learning_rate": 0.00019930496046767008,
+      "loss": 0.3159,
+      "step": 184
+    },
+    {
+      "epoch": 0.40726472206934505,
+      "grad_norm": 0.9752854704856873,
+      "learning_rate": 0.00019927444988122918,
+      "loss": 0.3026,
+      "step": 185
+    },
+    {
+      "epoch": 0.40946615299944966,
+      "grad_norm": 1.7940340042114258,
+      "learning_rate": 0.00019924328636750102,
+      "loss": 0.3452,
+      "step": 186
+    },
+    {
+      "epoch": 0.4116675839295542,
+      "grad_norm": 1.4885001182556152,
+      "learning_rate": 0.0001992114701314478,
+      "loss": 0.2027,
+      "step": 187
+    },
+    {
+      "epoch": 0.41386901485965877,
+      "grad_norm": 1.3337723016738892,
+      "learning_rate": 0.0001991790013823246,
+      "loss": 0.2417,
+      "step": 188
+    },
+    {
+      "epoch": 0.4160704457897633,
+      "grad_norm": 1.3942782878875732,
+      "learning_rate": 0.00019914588033367823,
+      "loss": 0.3172,
+      "step": 189
+    },
+    {
+      "epoch": 0.41827187671986793,
+      "grad_norm": 0.6452035307884216,
+      "learning_rate": 0.00019911210720334547,
+      "loss": 0.1793,
+      "step": 190
+    },
+    {
+      "epoch": 0.4204733076499725,
+      "grad_norm": 1.146669864654541,
+      "learning_rate": 0.000199077682213452,
+      "loss": 0.2297,
+      "step": 191
+    },
+    {
+      "epoch": 0.42267473858007704,
+      "grad_norm": 1.0184296369552612,
+      "learning_rate": 0.0001990426055904107,
+      "loss": 0.1563,
+      "step": 192
+    },
+    {
+      "epoch": 0.42487616951018164,
+      "grad_norm": 1.3531533479690552,
+      "learning_rate": 0.0001990068775649202,
+      "loss": 0.194,
+      "step": 193
+    },
+    {
+      "epoch": 0.4270776004402862,
+      "grad_norm": 1.0374277830123901,
+      "learning_rate": 0.0001989704983719635,
+      "loss": 0.1652,
+      "step": 194
+    },
+    {
+      "epoch": 0.42927903137039075,
+      "grad_norm": 0.8380899429321289,
+      "learning_rate": 0.00019893346825080625,
+      "loss": 0.195,
+      "step": 195
+    },
+    {
+      "epoch": 0.4314804623004953,
+      "grad_norm": 1.239431619644165,
+      "learning_rate": 0.00019889578744499532,
+      "loss": 0.2463,
+      "step": 196
+    },
+    {
+      "epoch": 0.4336818932305999,
+      "grad_norm": 1.1668486595153809,
+      "learning_rate": 0.000198857456202357,
+      "loss": 0.237,
+      "step": 197
+    },
+    {
+      "epoch": 0.43588332416070447,
+      "grad_norm": 1.1140515804290771,
+      "learning_rate": 0.00019881847477499557,
+      "loss": 0.2848,
+      "step": 198
+    },
+    {
+      "epoch": 0.438084755090809,
+      "grad_norm": 1.3733386993408203,
+      "learning_rate": 0.00019877884341929164,
+      "loss": 0.2022,
+      "step": 199
+    },
+    {
+      "epoch": 0.4402861860209136,
+      "grad_norm": 1.389896273612976,
+      "learning_rate": 0.00019873856239590034,
+      "loss": 0.267,
+      "step": 200
+    },
+    {
+      "epoch": 0.4424876169510182,
+      "grad_norm": 1.0052469968795776,
+      "learning_rate": 0.00019869763196974957,
+      "loss": 0.2277,
+      "step": 201
+    },
+    {
+      "epoch": 0.44468904788112273,
+      "grad_norm": 1.034943699836731,
+      "learning_rate": 0.00019865605241003848,
+      "loss": 0.1446,
+      "step": 202
+    },
+    {
+      "epoch": 0.4468904788112273,
+      "grad_norm": 0.5385359525680542,
+      "learning_rate": 0.0001986138239902355,
+      "loss": 0.0705,
+      "step": 203
+    },
+    {
+      "epoch": 0.44909190974133184,
+      "grad_norm": 0.9618946313858032,
+      "learning_rate": 0.00019857094698807662,
+      "loss": 0.2932,
+      "step": 204
+    },
+    {
+      "epoch": 0.45129334067143645,
+      "grad_norm": 1.1805518865585327,
+      "learning_rate": 0.00019852742168556355,
+      "loss": 0.2388,
+      "step": 205
+    },
+    {
+      "epoch": 0.453494771601541,
+      "grad_norm": 1.1751959323883057,
+      "learning_rate": 0.00019848324836896185,
+      "loss": 0.2859,
+      "step": 206
+    },
+    {
+      "epoch": 0.45569620253164556,
+      "grad_norm": 1.0479226112365723,
+      "learning_rate": 0.00019843842732879914,
+      "loss": 0.2722,
+      "step": 207
+    },
+    {
+      "epoch": 0.45789763346175016,
+      "grad_norm": 1.0439608097076416,
+      "learning_rate": 0.00019839295885986296,
+      "loss": 0.1869,
+      "step": 208
+    },
+    {
+      "epoch": 0.4600990643918547,
+      "grad_norm": 1.0632809400558472,
+      "learning_rate": 0.00019834684326119916,
+      "loss": 0.2354,
+      "step": 209
+    },
+    {
+      "epoch": 0.46230049532195927,
+      "grad_norm": 0.8700536489486694,
+      "learning_rate": 0.00019830008083610964,
+      "loss": 0.1734,
+      "step": 210
+    },
+    {
+      "epoch": 0.4645019262520638,
+      "grad_norm": 1.8080542087554932,
+      "learning_rate": 0.0001982526718921505,
+      "loss": 0.411,
+      "step": 211
+    },
+    {
+      "epoch": 0.46670335718216843,
+      "grad_norm": 1.38103449344635,
+      "learning_rate": 0.00019820461674113,
+      "loss": 0.2625,
+      "step": 212
+    },
+    {
+      "epoch": 0.468904788112273,
+      "grad_norm": 1.205725073814392,
+      "learning_rate": 0.00019815591569910654,
+      "loss": 0.3933,
+      "step": 213
+    },
+    {
+      "epoch": 0.47110621904237754,
+      "grad_norm": 2.5588293075561523,
+      "learning_rate": 0.00019810656908638644,
+      "loss": 0.3953,
+      "step": 214
+    },
+    {
+      "epoch": 0.4733076499724821,
+      "grad_norm": 1.0547009706497192,
+      "learning_rate": 0.00019805657722752202,
+      "loss": 0.1817,
+      "step": 215
+    },
+    {
+      "epoch": 0.4755090809025867,
+      "grad_norm": 1.2319700717926025,
+      "learning_rate": 0.00019800594045130932,
+      "loss": 0.2348,
+      "step": 216
+    },
+    {
+      "epoch": 0.47771051183269125,
+      "grad_norm": 1.148016333580017,
+      "learning_rate": 0.00019795465909078607,
+      "loss": 0.201,
+      "step": 217
+    },
+    {
+      "epoch": 0.4799119427627958,
+      "grad_norm": 1.0850389003753662,
+      "learning_rate": 0.0001979027334832293,
+      "loss": 0.2078,
+      "step": 218
+    },
+    {
+      "epoch": 0.48211337369290036,
+      "grad_norm": 1.154354214668274,
+      "learning_rate": 0.0001978501639701534,
+      "loss": 0.3007,
+      "step": 219
+    },
+    {
+      "epoch": 0.48431480462300497,
+      "grad_norm": 1.7488455772399902,
+      "learning_rate": 0.00019779695089730765,
+      "loss": 0.2684,
+      "step": 220
+    },
+    {
+      "epoch": 0.4865162355531095,
+      "grad_norm": 2.215618848800659,
+      "learning_rate": 0.00019774309461467399,
+      "loss": 0.1546,
+      "step": 221
+    },
+    {
+      "epoch": 0.4887176664832141,
+      "grad_norm": 1.2282485961914062,
+      "learning_rate": 0.00019768859547646478,
+      "loss": 0.2675,
+      "step": 222
+    },
+    {
+      "epoch": 0.4909190974133187,
+      "grad_norm": 0.7270124554634094,
+      "learning_rate": 0.00019763345384112043,
+      "loss": 0.1491,
+      "step": 223
+    },
+    {
+      "epoch": 0.49312052834342324,
+      "grad_norm": 1.3781054019927979,
+      "learning_rate": 0.00019757767007130706,
+      "loss": 0.3164,
+      "step": 224
+    },
+    {
+      "epoch": 0.4953219592735278,
+      "grad_norm": 1.1072008609771729,
+      "learning_rate": 0.00019752124453391404,
+      "loss": 0.2054,
+      "step": 225
+    },
+    {
+      "epoch": 0.49752339020363234,
+      "grad_norm": 1.453762412071228,
+      "learning_rate": 0.00019746417760005178,
+      "loss": 0.1959,
+      "step": 226
+    },
+    {
+      "epoch": 0.49972482113373695,
+      "grad_norm": 0.9458669424057007,
+      "learning_rate": 0.00019740646964504902,
+      "loss": 0.195,
+      "step": 227
+    },
+    {
+      "epoch": 0.5019262520638414,
+      "grad_norm": 0.9705711603164673,
+      "learning_rate": 0.00019734812104845047,
+      "loss": 0.1799,
+      "step": 228
+    },
+    {
+      "epoch": 0.5041276829939461,
+      "grad_norm": 1.1919350624084473,
+      "learning_rate": 0.0001972891321940145,
+      "loss": 0.1922,
+      "step": 229
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 1.200522541999817,
+      "learning_rate": 0.0001972295034697103,
+      "loss": 0.2551,
+      "step": 230
+    },
+    {
+      "epoch": 0.5085305448541552,
+      "grad_norm": 1.3816964626312256,
+      "learning_rate": 0.00019716923526771552,
+      "loss": 0.3134,
+      "step": 231
+    },
+    {
+      "epoch": 0.5107319757842598,
+      "grad_norm": 0.6731604933738708,
+      "learning_rate": 0.00019710832798441365,
+      "loss": 0.1294,
+      "step": 232
+    },
+    {
+      "epoch": 0.5129334067143644,
+      "grad_norm": 0.9393294453620911,
+      "learning_rate": 0.0001970467820203915,
+      "loss": 0.1441,
+      "step": 233
+    },
+    {
+      "epoch": 0.5151348376444689,
+      "grad_norm": 0.743293821811676,
+      "learning_rate": 0.00019698459778043627,
+      "loss": 0.1028,
+      "step": 234
+    },
+    {
+      "epoch": 0.5173362685745735,
+      "grad_norm": 1.600050449371338,
+      "learning_rate": 0.00019692177567353332,
+      "loss": 0.2226,
+      "step": 235
+    },
+    {
+      "epoch": 0.519537699504678,
+      "grad_norm": 1.7203608751296997,
+      "learning_rate": 0.0001968583161128631,
+      "loss": 0.3033,
+      "step": 236
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 1.3202494382858276,
+      "learning_rate": 0.00019679421951579874,
+      "loss": 0.0937,
+      "step": 237
+    },
+    {
+      "epoch": 0.5239405613648872,
+      "grad_norm": 1.847307801246643,
+      "learning_rate": 0.00019672948630390294,
+      "loss": 0.2683,
+      "step": 238
+    },
+    {
+      "epoch": 0.5261419922949917,
+      "grad_norm": 1.322928786277771,
+      "learning_rate": 0.00019666411690292564,
+      "loss": 0.2265,
+      "step": 239
+    },
+    {
+      "epoch": 0.5283434232250963,
+      "grad_norm": 1.9180538654327393,
+      "learning_rate": 0.0001965981117428008,
+      "loss": 0.2164,
+      "step": 240
+    },
+    {
+      "epoch": 0.5305448541552009,
+      "grad_norm": 2.2924304008483887,
+      "learning_rate": 0.00019653147125764392,
+      "loss": 0.2647,
+      "step": 241
+    },
+    {
+      "epoch": 0.5327462850853054,
+      "grad_norm": 0.9613167643547058,
+      "learning_rate": 0.0001964641958857489,
+      "loss": 0.1628,
+      "step": 242
+    },
+    {
+      "epoch": 0.53494771601541,
+      "grad_norm": 2.0845701694488525,
+      "learning_rate": 0.00019639628606958533,
+      "loss": 0.3281,
+      "step": 243
+    },
+    {
+      "epoch": 0.5371491469455146,
+      "grad_norm": 1.326499104499817,
+      "learning_rate": 0.0001963277422557956,
+      "loss": 0.1656,
+      "step": 244
+    },
+    {
+      "epoch": 0.5393505778756191,
+      "grad_norm": 0.9228304028511047,
+      "learning_rate": 0.00019625856489519184,
+      "loss": 0.19,
+      "step": 245
+    },
+    {
+      "epoch": 0.5415520088057237,
+      "grad_norm": 1.0690181255340576,
+      "learning_rate": 0.00019618875444275293,
+      "loss": 0.2124,
+      "step": 246
+    },
+    {
+      "epoch": 0.5437534397358282,
+      "grad_norm": 1.492824673652649,
+      "learning_rate": 0.00019611831135762174,
+      "loss": 0.1966,
+      "step": 247
+    },
+    {
+      "epoch": 0.5459548706659328,
+      "grad_norm": 1.7507821321487427,
+      "learning_rate": 0.00019604723610310194,
+      "loss": 0.4592,
+      "step": 248
+    },
+    {
+      "epoch": 0.5481563015960375,
+      "grad_norm": 1.523306965827942,
+      "learning_rate": 0.00019597552914665483,
+      "loss": 0.3597,
+      "step": 249
+    },
+    {
+      "epoch": 0.550357732526142,
+      "grad_norm": 1.8270065784454346,
+      "learning_rate": 0.0001959031909598966,
+      "loss": 0.314,
+      "step": 250
+    },
+    {
+      "epoch": 0.5525591634562466,
+      "grad_norm": 0.5709218382835388,
+      "learning_rate": 0.00019583022201859483,
+      "loss": 0.1128,
+      "step": 251
+    },
+    {
+      "epoch": 0.5547605943863512,
+      "grad_norm": 1.1685845851898193,
+      "learning_rate": 0.00019575662280266578,
+      "loss": 0.2384,
+      "step": 252
+    },
+    {
+      "epoch": 0.5569620253164557,
+      "grad_norm": 0.5500782132148743,
+      "learning_rate": 0.00019568239379617088,
+      "loss": 0.1463,
+      "step": 253
+    },
+    {
+      "epoch": 0.5591634562465603,
+      "grad_norm": 1.9069963693618774,
+      "learning_rate": 0.00019560753548731374,
+      "loss": 0.5131,
+      "step": 254
+    },
+    {
+      "epoch": 0.5613648871766649,
+      "grad_norm": 1.0922200679779053,
+      "learning_rate": 0.0001955320483684369,
+      "loss": 0.2108,
+      "step": 255
+    },
+    {
+      "epoch": 0.5635663181067694,
+      "grad_norm": 1.1500283479690552,
+      "learning_rate": 0.0001954559329360185,
+      "loss": 0.2229,
+      "step": 256
+    },
+    {
+      "epoch": 0.565767749036874,
+      "grad_norm": 1.501747488975525,
+      "learning_rate": 0.0001953791896906692,
+      "loss": 0.1476,
+      "step": 257
+    },
+    {
+      "epoch": 0.5679691799669785,
+      "grad_norm": 0.5918743014335632,
+      "learning_rate": 0.00019530181913712872,
+      "loss": 0.1434,
+      "step": 258
+    },
+    {
+      "epoch": 0.5701706108970831,
+      "grad_norm": 0.9154205322265625,
+      "learning_rate": 0.00019522382178426258,
+      "loss": 0.2219,
+      "step": 259
+    },
+    {
+      "epoch": 0.5723720418271877,
+      "grad_norm": 1.260572910308838,
+      "learning_rate": 0.00019514519814505877,
+      "loss": 0.3396,
+      "step": 260
+    },
+    {
+      "epoch": 0.5745734727572922,
+      "grad_norm": 0.9173365235328674,
+      "learning_rate": 0.00019506594873662433,
+      "loss": 0.1369,
+      "step": 261
+    },
+    {
+      "epoch": 0.5767749036873968,
+      "grad_norm": 1.759440302848816,
+      "learning_rate": 0.00019498607408018203,
+      "loss": 0.7267,
+      "step": 262
+    },
+    {
+      "epoch": 0.5789763346175014,
+      "grad_norm": 1.300220251083374,
+      "learning_rate": 0.00019490557470106686,
+      "loss": 0.3193,
+      "step": 263
+    },
+    {
+      "epoch": 0.5811777655476059,
+      "grad_norm": 1.3596434593200684,
+      "learning_rate": 0.00019482445112872264,
+      "loss": 0.3193,
+      "step": 264
+    },
+    {
+      "epoch": 0.5833791964777105,
+      "grad_norm": 0.9206734299659729,
+      "learning_rate": 0.00019474270389669844,
+      "loss": 0.173,
+      "step": 265
+    },
+    {
+      "epoch": 0.5855806274078151,
+      "grad_norm": 1.1209437847137451,
+      "learning_rate": 0.0001946603335426452,
+      "loss": 0.1857,
+      "step": 266
+    },
+    {
+      "epoch": 0.5877820583379196,
+      "grad_norm": 0.7680370211601257,
+      "learning_rate": 0.00019457734060831204,
+      "loss": 0.2384,
+      "step": 267
+    },
+    {
+      "epoch": 0.5899834892680242,
+      "grad_norm": 1.1541163921356201,
+      "learning_rate": 0.00019449372563954293,
+      "loss": 0.2907,
+      "step": 268
+    },
+    {
+      "epoch": 0.5921849201981287,
+      "grad_norm": 0.7708389759063721,
+      "learning_rate": 0.00019440948918627283,
+      "loss": 0.1215,
+      "step": 269
+    },
+    {
+      "epoch": 0.5943863511282333,
+      "grad_norm": 0.7209833860397339,
+      "learning_rate": 0.00019432463180252423,
+      "loss": 0.2057,
+      "step": 270
+    },
+    {
+      "epoch": 0.596587782058338,
+      "grad_norm": 1.0058317184448242,
+      "learning_rate": 0.0001942391540464035,
+      "loss": 0.316,
+      "step": 271
+    },
+    {
+      "epoch": 0.5987892129884425,
+      "grad_norm": 0.912278950214386,
+      "learning_rate": 0.00019415305648009715,
+      "loss": 0.2154,
+      "step": 272
+    },
+    {
+      "epoch": 0.6009906439185471,
+      "grad_norm": 0.9229905009269714,
+      "learning_rate": 0.00019406633966986828,
+      "loss": 0.1988,
+      "step": 273
+    },
+    {
+      "epoch": 0.6031920748486517,
+      "grad_norm": 0.9416797161102295,
+      "learning_rate": 0.00019397900418605257,
+      "loss": 0.2454,
+      "step": 274
+    },
+    {
+      "epoch": 0.6053935057787562,
+      "grad_norm": 0.7916048765182495,
+      "learning_rate": 0.0001938910506030549,
+      "loss": 0.1595,
+      "step": 275
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.7961133718490601,
+      "learning_rate": 0.00019380247949934534,
+      "loss": 0.3149,
+      "step": 276
+    },
+    {
+      "epoch": 0.6097963676389653,
+      "grad_norm": 2.367426633834839,
+      "learning_rate": 0.00019371329145745524,
+      "loss": 0.4522,
+      "step": 277
+    },
+    {
+      "epoch": 0.6119977985690699,
+      "grad_norm": 1.6261329650878906,
+      "learning_rate": 0.00019362348706397373,
+      "loss": 0.3005,
+      "step": 278
+    },
+    {
+      "epoch": 0.6141992294991745,
+      "grad_norm": 0.9070632457733154,
+      "learning_rate": 0.0001935330669095436,
+      "loss": 0.2101,
+      "step": 279
+    },
+    {
+      "epoch": 0.616400660429279,
+      "grad_norm": 1.2766557931900024,
+      "learning_rate": 0.0001934420315888575,
+      "loss": 0.2502,
+      "step": 280
+    },
+    {
+      "epoch": 0.6186020913593836,
+      "grad_norm": 1.20702064037323,
+      "learning_rate": 0.00019335038170065403,
+      "loss": 0.2563,
+      "step": 281
+    },
+    {
+      "epoch": 0.6208035222894882,
+      "grad_norm": 1.1549800634384155,
+      "learning_rate": 0.0001932581178477138,
+      "loss": 0.1833,
+      "step": 282
+    },
+    {
+      "epoch": 0.6230049532195927,
+      "grad_norm": 1.2302500009536743,
+      "learning_rate": 0.0001931652406368554,
+      "loss": 0.226,
+      "step": 283
+    },
+    {
+      "epoch": 0.6252063841496973,
+      "grad_norm": 1.1015033721923828,
+      "learning_rate": 0.00019307175067893165,
+      "loss": 0.3038,
+      "step": 284
+    },
+    {
+      "epoch": 0.6274078150798019,
+      "grad_norm": 1.1137440204620361,
+      "learning_rate": 0.00019297764858882514,
+      "loss": 0.2073,
+      "step": 285
+    },
+    {
+      "epoch": 0.6296092460099064,
+      "grad_norm": 1.1460460424423218,
+      "learning_rate": 0.00019288293498544469,
+      "loss": 0.4567,
+      "step": 286
+    },
+    {
+      "epoch": 0.631810676940011,
+      "grad_norm": 1.245483636856079,
+      "learning_rate": 0.00019278761049172088,
+      "loss": 0.3153,
+      "step": 287
+    },
+    {
+      "epoch": 0.6340121078701155,
+      "grad_norm": 0.587056040763855,
+      "learning_rate": 0.0001926916757346022,
+      "loss": 0.0915,
+      "step": 288
+    },
+    {
+      "epoch": 0.6362135388002201,
+      "grad_norm": 0.8686281442642212,
+      "learning_rate": 0.00019259513134505077,
+      "loss": 0.1085,
+      "step": 289
+    },
+    {
+      "epoch": 0.6384149697303247,
+      "grad_norm": 1.245270848274231,
+      "learning_rate": 0.0001924979779580383,
+      "loss": 0.2146,
+      "step": 290
+    },
+    {
+      "epoch": 0.6406164006604292,
+      "grad_norm": 0.9802015423774719,
+      "learning_rate": 0.00019240021621254186,
+      "loss": 0.3619,
+      "step": 291
+    },
+    {
+      "epoch": 0.6428178315905339,
+      "grad_norm": 1.0316482782363892,
+      "learning_rate": 0.00019230184675153976,
+      "loss": 0.5368,
+      "step": 292
+    },
+    {
+      "epoch": 0.6450192625206385,
+      "grad_norm": 1.010973334312439,
+      "learning_rate": 0.00019220287022200707,
+      "loss": 0.1715,
+      "step": 293
+    },
+    {
+      "epoch": 0.647220693450743,
+      "grad_norm": 1.4453409910202026,
+      "learning_rate": 0.00019210328727491173,
+      "loss": 0.1925,
+      "step": 294
+    },
+    {
+      "epoch": 0.6494221243808476,
+      "grad_norm": 1.1811559200286865,
+      "learning_rate": 0.00019200309856520997,
+      "loss": 0.2017,
+      "step": 295
+    },
+    {
+      "epoch": 0.6516235553109522,
+      "grad_norm": 0.8928061127662659,
+      "learning_rate": 0.00019190230475184223,
+      "loss": 0.2713,
+      "step": 296
+    },
+    {
+      "epoch": 0.6538249862410567,
+      "grad_norm": 1.332140326499939,
+      "learning_rate": 0.00019180090649772855,
+      "loss": 0.2451,
+      "step": 297
+    },
+    {
+      "epoch": 0.6560264171711613,
+      "grad_norm": 1.1139875650405884,
+      "learning_rate": 0.00019169890446976454,
+      "loss": 0.2006,
+      "step": 298
+    },
+    {
+      "epoch": 0.6582278481012658,
+      "grad_norm": 1.582215666770935,
+      "learning_rate": 0.00019159629933881666,
+      "loss": 0.2862,
+      "step": 299
+    },
+    {
+      "epoch": 0.6604292790313704,
+      "grad_norm": 0.647504448890686,
+      "learning_rate": 0.00019149309177971815,
+      "loss": 0.0764,
+      "step": 300
+    },
+    {
+      "epoch": 0.662630709961475,
+      "grad_norm": 2.3032169342041016,
+      "learning_rate": 0.00019138928247126424,
+      "loss": 0.3415,
+      "step": 301
+    },
+    {
+      "epoch": 0.6648321408915795,
+      "grad_norm": 1.4954179525375366,
+      "learning_rate": 0.00019128487209620803,
+      "loss": 0.3911,
+      "step": 302
+    },
+    {
+      "epoch": 0.6670335718216841,
+      "grad_norm": 1.450543999671936,
+      "learning_rate": 0.0001911798613412557,
+      "loss": 0.2685,
+      "step": 303
+    },
+    {
+      "epoch": 0.6692350027517887,
+      "grad_norm": 1.65657639503479,
+      "learning_rate": 0.0001910742508970622,
+      "loss": 0.343,
+      "step": 304
+    },
+    {
+      "epoch": 0.6714364336818932,
+      "grad_norm": 2.165513515472412,
+      "learning_rate": 0.00019096804145822655,
+      "loss": 0.3681,
+      "step": 305
+    },
+    {
+      "epoch": 0.6736378646119978,
+      "grad_norm": 1.0061649084091187,
+      "learning_rate": 0.00019086123372328746,
+      "loss": 0.1908,
+      "step": 306
+    },
+    {
+      "epoch": 0.6758392955421023,
+      "grad_norm": 0.8672468066215515,
+      "learning_rate": 0.00019075382839471857,
+      "loss": 0.167,
+      "step": 307
+    },
+    {
+      "epoch": 0.6780407264722069,
+      "grad_norm": 1.4202598333358765,
+      "learning_rate": 0.0001906458261789238,
+      "loss": 0.1973,
+      "step": 308
+    },
+    {
+      "epoch": 0.6802421574023115,
+      "grad_norm": 1.4729435443878174,
+      "learning_rate": 0.00019053722778623305,
+      "loss": 0.3278,
+      "step": 309
+    },
+    {
+      "epoch": 0.682443588332416,
+      "grad_norm": 1.2635149955749512,
+      "learning_rate": 0.000190428033930897,
+      "loss": 0.2677,
+      "step": 310
+    },
+    {
+      "epoch": 0.6846450192625206,
+      "grad_norm": 1.1371662616729736,
+      "learning_rate": 0.0001903182453310828,
+      "loss": 0.3166,
+      "step": 311
+    },
+    {
+      "epoch": 0.6868464501926252,
+      "grad_norm": 0.8570672869682312,
+      "learning_rate": 0.0001902078627088692,
+      "loss": 0.2779,
+      "step": 312
+    },
+    {
+      "epoch": 0.6890478811227297,
+      "grad_norm": 0.9639473557472229,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 0.1461,
+      "step": 313
+    },
+    {
+      "epoch": 0.6912493120528344,
+      "grad_norm": 1.0219823122024536,
+      "learning_rate": 0.00018998531830508869,
+      "loss": 0.259,
+      "step": 314
+    },
+    {
+      "epoch": 0.693450742982939,
+      "grad_norm": 0.8559497594833374,
+      "learning_rate": 0.00018987315798719452,
+      "loss": 0.2074,
+      "step": 315
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 1.8177236318588257,
+      "learning_rate": 0.000189760406574237,
+      "loss": 0.3885,
+      "step": 316
+    },
+    {
+      "epoch": 0.6978536048431481,
+      "grad_norm": 1.7845747470855713,
+      "learning_rate": 0.0001896470648077813,
+      "loss": 0.3579,
+      "step": 317
+    },
+    {
+      "epoch": 0.7000550357732526,
+      "grad_norm": 0.8975507020950317,
+      "learning_rate": 0.0001895331334332753,
+      "loss": 0.2452,
+      "step": 318
+    },
+    {
+      "epoch": 0.7022564667033572,
+      "grad_norm": 0.9428125023841858,
+      "learning_rate": 0.00018941861320004484,
+      "loss": 0.2423,
+      "step": 319
+    },
+    {
+      "epoch": 0.7044578976334618,
+      "grad_norm": 1.159013032913208,
+      "learning_rate": 0.00018930350486128856,
+      "loss": 0.4047,
+      "step": 320
+    },
+    {
+      "epoch": 0.7066593285635663,
+      "grad_norm": 0.8494213223457336,
+      "learning_rate": 0.00018918780917407307,
+      "loss": 0.2341,
+      "step": 321
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 1.068804144859314,
+      "learning_rate": 0.0001890715268993281,
+      "loss": 0.3007,
+      "step": 322
+    },
+    {
+      "epoch": 0.7110621904237755,
+      "grad_norm": 0.43914756178855896,
+      "learning_rate": 0.0001889546588018412,
+      "loss": 0.0936,
+      "step": 323
+    },
+    {
+      "epoch": 0.71326362135388,
+      "grad_norm": 0.7797726392745972,
+      "learning_rate": 0.00018883720565025295,
+      "loss": 0.2057,
+      "step": 324
+    },
+    {
+      "epoch": 0.7154650522839846,
+      "grad_norm": 0.7800988554954529,
+      "learning_rate": 0.00018871916821705184,
+      "loss": 0.2073,
+      "step": 325
+    },
+    {
+      "epoch": 0.7176664832140892,
+      "grad_norm": 0.5828583240509033,
+      "learning_rate": 0.00018860054727856915,
+      "loss": 0.1515,
+      "step": 326
+    },
+    {
+      "epoch": 0.7198679141441937,
+      "grad_norm": 1.471644401550293,
+      "learning_rate": 0.00018848134361497385,
+      "loss": 0.3854,
+      "step": 327
+    },
+    {
+      "epoch": 0.7220693450742983,
+      "grad_norm": 1.2705450057983398,
+      "learning_rate": 0.00018836155801026753,
+      "loss": 0.2625,
+      "step": 328
+    },
+    {
+      "epoch": 0.7242707760044028,
+      "grad_norm": 1.0784512758255005,
+      "learning_rate": 0.0001882411912522792,
+      "loss": 0.1387,
+      "step": 329
+    },
+    {
+      "epoch": 0.7264722069345074,
+      "grad_norm": 1.1987195014953613,
+      "learning_rate": 0.00018812024413266006,
+      "loss": 0.2613,
+      "step": 330
+    },
+    {
+      "epoch": 0.728673637864612,
+      "grad_norm": 0.9083921909332275,
+      "learning_rate": 0.0001879987174468784,
+      "loss": 0.1633,
+      "step": 331
+    },
+    {
+      "epoch": 0.7308750687947165,
+      "grad_norm": 1.1217319965362549,
+      "learning_rate": 0.00018787661199421427,
+      "loss": 0.158,
+      "step": 332
+    },
+    {
+      "epoch": 0.7330764997248211,
+      "grad_norm": 0.7454916834831238,
+      "learning_rate": 0.00018775392857775432,
+      "loss": 0.1587,
+      "step": 333
+    },
+    {
+      "epoch": 0.7352779306549257,
+      "grad_norm": 0.8220398426055908,
+      "learning_rate": 0.00018763066800438636,
+      "loss": 0.1583,
+      "step": 334
+    },
+    {
+      "epoch": 0.7374793615850302,
+      "grad_norm": 0.908816397190094,
+      "learning_rate": 0.00018750683108479432,
+      "loss": 0.2706,
+      "step": 335
+    },
+    {
+      "epoch": 0.7396807925151349,
+      "grad_norm": 1.1175532341003418,
+      "learning_rate": 0.0001873824186334526,
+      "loss": 0.241,
+      "step": 336
+    },
+    {
+      "epoch": 0.7418822234452394,
+      "grad_norm": 0.8782237768173218,
+      "learning_rate": 0.00018725743146862095,
+      "loss": 0.2187,
+      "step": 337
+    },
+    {
+      "epoch": 0.744083654375344,
+      "grad_norm": 1.36802339553833,
+      "learning_rate": 0.00018713187041233896,
+      "loss": 0.2422,
+      "step": 338
+    },
+    {
+      "epoch": 0.7462850853054486,
+      "grad_norm": 1.4567631483078003,
+      "learning_rate": 0.00018700573629042074,
+      "loss": 0.3323,
+      "step": 339
+    },
+    {
+      "epoch": 0.7484865162355531,
+      "grad_norm": 0.8530279994010925,
+      "learning_rate": 0.00018687902993244943,
+      "loss": 0.2511,
+      "step": 340
+    },
+    {
+      "epoch": 0.7506879471656577,
+      "grad_norm": 1.4750181436538696,
+      "learning_rate": 0.00018675175217177175,
+      "loss": 0.334,
+      "step": 341
+    },
+    {
+      "epoch": 0.7528893780957623,
+      "grad_norm": 0.5695929527282715,
+      "learning_rate": 0.0001866239038454926,
+      "loss": 0.0692,
+      "step": 342
+    },
+    {
+      "epoch": 0.7550908090258668,
+      "grad_norm": 1.2884031534194946,
+      "learning_rate": 0.00018649548579446936,
+      "loss": 0.3052,
+      "step": 343
+    },
+    {
+      "epoch": 0.7572922399559714,
+      "grad_norm": 1.2000869512557983,
+      "learning_rate": 0.0001863664988633066,
+      "loss": 0.1638,
+      "step": 344
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 1.3409312963485718,
+      "learning_rate": 0.00018623694390035035,
+      "loss": 0.2515,
+      "step": 345
+    },
+    {
+      "epoch": 0.7616951018161805,
+      "grad_norm": 0.6578596234321594,
+      "learning_rate": 0.00018610682175768258,
+      "loss": 0.1144,
+      "step": 346
+    },
+    {
+      "epoch": 0.7638965327462851,
+      "grad_norm": 1.0106149911880493,
+      "learning_rate": 0.00018597613329111566,
+      "loss": 0.1943,
+      "step": 347
+    },
+    {
+      "epoch": 0.7660979636763896,
+      "grad_norm": 1.586403489112854,
+      "learning_rate": 0.00018584487936018661,
+      "loss": 0.1804,
+      "step": 348
+    },
+    {
+      "epoch": 0.7682993946064942,
+      "grad_norm": 1.031691074371338,
+      "learning_rate": 0.0001857130608281515,
+      "loss": 0.419,
+      "step": 349
+    },
+    {
+      "epoch": 0.7705008255365988,
+      "grad_norm": 1.0916547775268555,
+      "learning_rate": 0.0001855806785619798,
+      "loss": 0.1373,
+      "step": 350
+    },
+    {
+      "epoch": 0.7727022564667033,
+      "grad_norm": 2.66514253616333,
+      "learning_rate": 0.00018544773343234862,
+      "loss": 0.2008,
+      "step": 351
+    },
+    {
+      "epoch": 0.7749036873968079,
+      "grad_norm": 1.0321251153945923,
+      "learning_rate": 0.00018531422631363706,
+      "loss": 0.168,
+      "step": 352
+    },
+    {
+      "epoch": 0.7771051183269125,
+      "grad_norm": 1.2961331605911255,
+      "learning_rate": 0.00018518015808392045,
+      "loss": 0.2361,
+      "step": 353
+    },
+    {
+      "epoch": 0.779306549257017,
+      "grad_norm": 0.719615638256073,
+      "learning_rate": 0.0001850455296249644,
+      "loss": 0.1776,
+      "step": 354
+    },
+    {
+      "epoch": 0.7815079801871216,
+      "grad_norm": 1.1901960372924805,
+      "learning_rate": 0.0001849103418222194,
+      "loss": 0.2842,
+      "step": 355
+    },
+    {
+      "epoch": 0.7837094111172263,
+      "grad_norm": 0.9724762439727783,
+      "learning_rate": 0.00018477459556481441,
+      "loss": 0.2397,
+      "step": 356
+    },
+    {
+      "epoch": 0.7859108420473307,
+      "grad_norm": 1.761529803276062,
+      "learning_rate": 0.0001846382917455516,
+      "loss": 0.313,
+      "step": 357
+    },
+    {
+      "epoch": 0.7881122729774354,
+      "grad_norm": 1.0816775560379028,
+      "learning_rate": 0.00018450143126090015,
+      "loss": 0.1472,
+      "step": 358
+    },
+    {
+      "epoch": 0.7903137039075399,
+      "grad_norm": 2.434422016143799,
+      "learning_rate": 0.00018436401501099035,
+      "loss": 0.177,
+      "step": 359
+    },
+    {
+      "epoch": 0.7925151348376445,
+      "grad_norm": 1.8141069412231445,
+      "learning_rate": 0.00018422604389960783,
+      "loss": 0.375,
+      "step": 360
+    },
+    {
+      "epoch": 0.7947165657677491,
+      "grad_norm": 1.1976251602172852,
+      "learning_rate": 0.00018408751883418754,
+      "loss": 0.2226,
+      "step": 361
+    },
+    {
+      "epoch": 0.7969179966978536,
+      "grad_norm": 1.4153374433517456,
+      "learning_rate": 0.00018394844072580773,
+      "loss": 0.3456,
+      "step": 362
+    },
+    {
+      "epoch": 0.7991194276279582,
+      "grad_norm": 1.5557993650436401,
+      "learning_rate": 0.00018380881048918405,
+      "loss": 0.2763,
+      "step": 363
+    },
+    {
+      "epoch": 0.8013208585580628,
+      "grad_norm": 1.2922636270523071,
+      "learning_rate": 0.00018366862904266356,
+      "loss": 0.3598,
+      "step": 364
+    },
+    {
+      "epoch": 0.8035222894881673,
+      "grad_norm": 1.2480388879776,
+      "learning_rate": 0.00018352789730821843,
+      "loss": 0.3873,
+      "step": 365
+    },
+    {
+      "epoch": 0.8057237204182719,
+      "grad_norm": 0.9546335339546204,
+      "learning_rate": 0.00018338661621144025,
+      "loss": 0.1378,
+      "step": 366
+    },
+    {
+      "epoch": 0.8079251513483764,
+      "grad_norm": 1.3006221055984497,
+      "learning_rate": 0.00018324478668153366,
+      "loss": 0.2566,
+      "step": 367
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 1.2144070863723755,
+      "learning_rate": 0.00018310240965131041,
+      "loss": 0.2278,
+      "step": 368
+    },
+    {
+      "epoch": 0.8123280132085856,
+      "grad_norm": 0.8332474827766418,
+      "learning_rate": 0.00018295948605718314,
+      "loss": 0.2634,
+      "step": 369
+    },
+    {
+      "epoch": 0.8145294441386901,
+      "grad_norm": 0.7528755068778992,
+      "learning_rate": 0.00018281601683915917,
+      "loss": 0.2693,
+      "step": 370
+    },
+    {
+      "epoch": 0.8167308750687947,
+      "grad_norm": 1.0342864990234375,
+      "learning_rate": 0.00018267200294083447,
+      "loss": 0.2044,
+      "step": 371
+    },
+    {
+      "epoch": 0.8189323059988993,
+      "grad_norm": 0.8297427296638489,
+      "learning_rate": 0.00018252744530938736,
+      "loss": 0.1956,
+      "step": 372
+    },
+    {
+      "epoch": 0.8211337369290038,
+      "grad_norm": 1.282098412513733,
+      "learning_rate": 0.00018238234489557215,
+      "loss": 0.2616,
+      "step": 373
+    },
+    {
+      "epoch": 0.8233351678591084,
+      "grad_norm": 0.6476250886917114,
+      "learning_rate": 0.0001822367026537133,
+      "loss": 0.1147,
+      "step": 374
+    },
+    {
+      "epoch": 0.825536598789213,
+      "grad_norm": 0.6506133079528809,
+      "learning_rate": 0.00018209051954169855,
+      "loss": 0.1762,
+      "step": 375
+    },
+    {
+      "epoch": 0.8277380297193175,
+      "grad_norm": 0.926846444606781,
+      "learning_rate": 0.0001819437965209732,
+      "loss": 0.1866,
+      "step": 376
+    },
+    {
+      "epoch": 0.8299394606494221,
+      "grad_norm": 0.9617591500282288,
+      "learning_rate": 0.0001817965345565334,
+      "loss": 0.2098,
+      "step": 377
+    },
+    {
+      "epoch": 0.8321408915795266,
+      "grad_norm": 1.071401834487915,
+      "learning_rate": 0.00018164873461691986,
+      "loss": 0.2428,
+      "step": 378
+    },
+    {
+      "epoch": 0.8343423225096313,
+      "grad_norm": 0.654083788394928,
+      "learning_rate": 0.00018150039767421178,
+      "loss": 0.0716,
+      "step": 379
+    },
+    {
+      "epoch": 0.8365437534397359,
+      "grad_norm": 1.6268547773361206,
+      "learning_rate": 0.00018135152470401996,
+      "loss": 0.2735,
+      "step": 380
+    },
+    {
+      "epoch": 0.8387451843698404,
+      "grad_norm": 0.4768295884132385,
+      "learning_rate": 0.00018120211668548088,
+      "loss": 0.0739,
+      "step": 381
+    },
+    {
+      "epoch": 0.840946615299945,
+      "grad_norm": 1.295459508895874,
+      "learning_rate": 0.00018105217460124981,
+      "loss": 0.2342,
+      "step": 382
+    },
+    {
+      "epoch": 0.8431480462300496,
+      "grad_norm": 0.7535791397094727,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.1549,
+      "step": 383
+    },
+    {
+      "epoch": 0.8453494771601541,
+      "grad_norm": 1.5335655212402344,
+      "learning_rate": 0.00018075069218388964,
+      "loss": 0.1621,
+      "step": 384
+    },
+    {
+      "epoch": 0.8475509080902587,
+      "grad_norm": 1.2511568069458008,
+      "learning_rate": 0.00018059915383360808,
+      "loss": 0.1517,
+      "step": 385
+    },
+    {
+      "epoch": 0.8497523390203633,
+      "grad_norm": 0.8873250484466553,
+      "learning_rate": 0.00018044708538331655,
+      "loss": 0.2602,
+      "step": 386
+    },
+    {
+      "epoch": 0.8519537699504678,
+      "grad_norm": 1.4157307147979736,
+      "learning_rate": 0.00018029448783316813,
+      "loss": 0.4575,
+      "step": 387
+    },
+    {
+      "epoch": 0.8541552008805724,
+      "grad_norm": 1.280274510383606,
+      "learning_rate": 0.00018014136218679567,
+      "loss": 0.1614,
+      "step": 388
+    },
+    {
+      "epoch": 0.8563566318106769,
+      "grad_norm": 0.8422673940658569,
+      "learning_rate": 0.0001799877094513054,
+      "loss": 0.1639,
+      "step": 389
+    },
+    {
+      "epoch": 0.8585580627407815,
+      "grad_norm": 0.9522169232368469,
+      "learning_rate": 0.00017983353063727016,
+      "loss": 0.1054,
+      "step": 390
+    },
+    {
+      "epoch": 0.8607594936708861,
+      "grad_norm": 1.752172589302063,
+      "learning_rate": 0.00017967882675872277,
+      "loss": 0.2285,
+      "step": 391
+    },
+    {
+      "epoch": 0.8629609246009906,
+      "grad_norm": 1.322982907295227,
+      "learning_rate": 0.00017952359883314955,
+      "loss": 0.1765,
+      "step": 392
+    },
+    {
+      "epoch": 0.8651623555310952,
+      "grad_norm": 0.9844683408737183,
+      "learning_rate": 0.00017936784788148328,
+      "loss": 0.1054,
+      "step": 393
+    },
+    {
+      "epoch": 0.8673637864611998,
+      "grad_norm": 1.8924535512924194,
+      "learning_rate": 0.00017921157492809672,
+      "loss": 0.2393,
+      "step": 394
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 1.1759653091430664,
+      "learning_rate": 0.000179054781000796,
+      "loss": 0.2491,
+      "step": 395
+    },
+    {
+      "epoch": 0.8717666483214089,
+      "grad_norm": 1.2846308946609497,
+      "learning_rate": 0.00017889746713081343,
+      "loss": 0.1594,
+      "step": 396
+    },
+    {
+      "epoch": 0.8739680792515134,
+      "grad_norm": 0.7170493602752686,
+      "learning_rate": 0.00017873963435280121,
+      "loss": 0.228,
+      "step": 397
+    },
+    {
+      "epoch": 0.876169510181618,
+      "grad_norm": 0.9288744926452637,
+      "learning_rate": 0.00017858128370482426,
+      "loss": 0.1143,
+      "step": 398
+    },
+    {
+      "epoch": 0.8783709411117226,
+      "grad_norm": 1.3167988061904907,
+      "learning_rate": 0.00017842241622835356,
+      "loss": 0.1729,
+      "step": 399
+    },
+    {
+      "epoch": 0.8805723720418271,
+      "grad_norm": 1.0657830238342285,
+      "learning_rate": 0.00017826303296825926,
+      "loss": 0.286,
+      "step": 400
+    },
+    {
+      "epoch": 0.8827738029719318,
+      "grad_norm": 0.8168554902076721,
+      "learning_rate": 0.0001781031349728039,
+      "loss": 0.2013,
+      "step": 401
+    },
+    {
+      "epoch": 0.8849752339020364,
+      "grad_norm": 0.882465660572052,
+      "learning_rate": 0.00017794272329363527,
+      "loss": 0.1545,
+      "step": 402
+    },
+    {
+      "epoch": 0.8871766648321409,
+      "grad_norm": 0.9866340160369873,
+      "learning_rate": 0.00017778179898577973,
+      "loss": 0.1833,
+      "step": 403
+    },
+    {
+      "epoch": 0.8893780957622455,
+      "grad_norm": 0.7578817009925842,
+      "learning_rate": 0.00017762036310763532,
+      "loss": 0.1541,
+      "step": 404
+    },
+    {
+      "epoch": 0.8915795266923501,
+      "grad_norm": 0.947085976600647,
+      "learning_rate": 0.00017745841672096442,
+      "loss": 0.2227,
+      "step": 405
+    },
+    {
+      "epoch": 0.8937809576224546,
+      "grad_norm": 1.0397213697433472,
+      "learning_rate": 0.0001772959608908873,
+      "loss": 0.1625,
+      "step": 406
+    },
+    {
+      "epoch": 0.8959823885525592,
+      "grad_norm": 1.140871286392212,
+      "learning_rate": 0.00017713299668587458,
+      "loss": 0.2139,
+      "step": 407
+    },
+    {
+      "epoch": 0.8981838194826637,
+      "grad_norm": 1.7579021453857422,
+      "learning_rate": 0.00017696952517774062,
+      "loss": 0.3054,
+      "step": 408
+    },
+    {
+      "epoch": 0.9003852504127683,
+      "grad_norm": 1.5178894996643066,
+      "learning_rate": 0.00017680554744163625,
+      "loss": 0.2656,
+      "step": 409
+    },
+    {
+      "epoch": 0.9025866813428729,
+      "grad_norm": 1.9655121564865112,
+      "learning_rate": 0.00017664106455604178,
+      "loss": 0.3215,
+      "step": 410
+    },
+    {
+      "epoch": 0.9047881122729774,
+      "grad_norm": 1.8185651302337646,
+      "learning_rate": 0.00017647607760275987,
+      "loss": 0.3021,
+      "step": 411
+    },
+    {
+      "epoch": 0.906989543203082,
+      "grad_norm": 0.8997313976287842,
+      "learning_rate": 0.0001763105876669084,
+      "loss": 0.2806,
+      "step": 412
+    },
+    {
+      "epoch": 0.9091909741331866,
+      "grad_norm": 1.726635456085205,
+      "learning_rate": 0.00017614459583691346,
+      "loss": 0.2853,
+      "step": 413
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 1.0895217657089233,
+      "learning_rate": 0.000175978103204502,
+      "loss": 0.28,
+      "step": 414
+    },
+    {
+      "epoch": 0.9135938359933957,
+      "grad_norm": 1.582118034362793,
+      "learning_rate": 0.00017581111086469476,
+      "loss": 0.2495,
+      "step": 415
+    },
+    {
+      "epoch": 0.9157952669235003,
+      "grad_norm": 1.6275440454483032,
+      "learning_rate": 0.00017564361991579904,
+      "loss": 0.2424,
+      "step": 416
+    },
+    {
+      "epoch": 0.9179966978536048,
+      "grad_norm": 1.0979431867599487,
+      "learning_rate": 0.00017547563145940158,
+      "loss": 0.1526,
+      "step": 417
+    },
+    {
+      "epoch": 0.9201981287837094,
+      "grad_norm": 0.8231645226478577,
+      "learning_rate": 0.00017530714660036112,
+      "loss": 0.1641,
+      "step": 418
+    },
+    {
+      "epoch": 0.9223995597138139,
+      "grad_norm": 0.8364385366439819,
+      "learning_rate": 0.00017513816644680125,
+      "loss": 0.2394,
+      "step": 419
+    },
+    {
+      "epoch": 0.9246009906439185,
+      "grad_norm": 0.9016224145889282,
+      "learning_rate": 0.00017496869211010315,
+      "loss": 0.2125,
+      "step": 420
+    },
+    {
+      "epoch": 0.9268024215740231,
+      "grad_norm": 1.0510677099227905,
+      "learning_rate": 0.00017479872470489824,
+      "loss": 0.1643,
+      "step": 421
+    },
+    {
+      "epoch": 0.9290038525041276,
+      "grad_norm": 0.7203924059867859,
+      "learning_rate": 0.0001746282653490608,
+      "loss": 0.1425,
+      "step": 422
+    },
+    {
+      "epoch": 0.9312052834342323,
+      "grad_norm": 1.231278657913208,
+      "learning_rate": 0.0001744573151637007,
+      "loss": 0.2488,
+      "step": 423
+    },
+    {
+      "epoch": 0.9334067143643369,
+      "grad_norm": 1.0261121988296509,
+      "learning_rate": 0.000174285875273156,
+      "loss": 0.2443,
+      "step": 424
+    },
+    {
+      "epoch": 0.9356081452944414,
+      "grad_norm": 1.9562700986862183,
+      "learning_rate": 0.0001741139468049855,
+      "loss": 0.1857,
+      "step": 425
+    },
+    {
+      "epoch": 0.937809576224546,
+      "grad_norm": 0.8829732537269592,
+      "learning_rate": 0.0001739415308899614,
+      "loss": 0.1966,
+      "step": 426
+    },
+    {
+      "epoch": 0.9400110071546505,
+      "grad_norm": 1.1366294622421265,
+      "learning_rate": 0.00017376862866206187,
+      "loss": 0.195,
+      "step": 427
+    },
+    {
+      "epoch": 0.9422124380847551,
+      "grad_norm": 1.005967617034912,
+      "learning_rate": 0.0001735952412584635,
+      "loss": 0.2883,
+      "step": 428
+    },
+    {
+      "epoch": 0.9444138690148597,
+      "grad_norm": 1.5798931121826172,
+      "learning_rate": 0.0001734213698195339,
+      "loss": 0.2963,
+      "step": 429
+    },
+    {
+      "epoch": 0.9466152999449642,
+      "grad_norm": 1.1144102811813354,
+      "learning_rate": 0.00017324701548882422,
+      "loss": 0.2546,
+      "step": 430
+    },
+    {
+      "epoch": 0.9488167308750688,
+      "grad_norm": 3.013399600982666,
+      "learning_rate": 0.00017307217941306145,
+      "loss": 0.3737,
+      "step": 431
+    },
+    {
+      "epoch": 0.9510181618051734,
+      "grad_norm": 1.1589728593826294,
+      "learning_rate": 0.00017289686274214118,
+      "loss": 0.124,
+      "step": 432
+    },
+    {
+      "epoch": 0.9532195927352779,
+      "grad_norm": 0.958777666091919,
+      "learning_rate": 0.00017272106662911973,
+      "loss": 0.1727,
+      "step": 433
+    },
+    {
+      "epoch": 0.9554210236653825,
+      "grad_norm": 1.4360767602920532,
+      "learning_rate": 0.00017254479223020684,
+      "loss": 0.2203,
+      "step": 434
+    },
+    {
+      "epoch": 0.9576224545954871,
+      "grad_norm": 1.6432348489761353,
+      "learning_rate": 0.00017236804070475787,
+      "loss": 0.16,
+      "step": 435
+    },
+    {
+      "epoch": 0.9598238855255916,
+      "grad_norm": 1.4297266006469727,
+      "learning_rate": 0.00017219081321526617,
+      "loss": 0.2383,
+      "step": 436
+    },
+    {
+      "epoch": 0.9620253164556962,
+      "grad_norm": 1.1295509338378906,
+      "learning_rate": 0.00017201311092735565,
+      "loss": 0.2701,
+      "step": 437
+    },
+    {
+      "epoch": 0.9642267473858007,
+      "grad_norm": 0.958570122718811,
+      "learning_rate": 0.00017183493500977278,
+      "loss": 0.1856,
+      "step": 438
+    },
+    {
+      "epoch": 0.9664281783159053,
+      "grad_norm": 0.8746889233589172,
+      "learning_rate": 0.00017165628663437922,
+      "loss": 0.2212,
+      "step": 439
+    },
+    {
+      "epoch": 0.9686296092460099,
+      "grad_norm": 1.3758814334869385,
+      "learning_rate": 0.000171477166976144,
+      "loss": 0.2748,
+      "step": 440
+    },
+    {
+      "epoch": 0.9708310401761144,
+      "grad_norm": 1.5859287977218628,
+      "learning_rate": 0.00017129757721313568,
+      "loss": 0.3003,
+      "step": 441
+    },
+    {
+      "epoch": 0.973032471106219,
+      "grad_norm": 0.7726479172706604,
+      "learning_rate": 0.00017111751852651473,
+      "loss": 0.2363,
+      "step": 442
+    },
+    {
+      "epoch": 0.9752339020363237,
+      "grad_norm": 1.1339430809020996,
+      "learning_rate": 0.0001709369921005258,
+      "loss": 0.2288,
+      "step": 443
+    },
+    {
+      "epoch": 0.9774353329664281,
+      "grad_norm": 1.4856394529342651,
+      "learning_rate": 0.0001707559991224897,
+      "loss": 0.4531,
+      "step": 444
+    },
+    {
+      "epoch": 0.9796367638965328,
+      "grad_norm": 1.0055327415466309,
+      "learning_rate": 0.0001705745407827959,
+      "loss": 0.2731,
+      "step": 445
+    },
+    {
+      "epoch": 0.9818381948266374,
+      "grad_norm": 0.8089772462844849,
+      "learning_rate": 0.0001703926182748945,
+      "loss": 0.2677,
+      "step": 446
+    },
+    {
+      "epoch": 0.9840396257567419,
+      "grad_norm": 0.907011866569519,
+      "learning_rate": 0.0001702102327952884,
+      "loss": 0.1726,
+      "step": 447
+    },
+    {
+      "epoch": 0.9862410566868465,
+      "grad_norm": 0.7658312916755676,
+      "learning_rate": 0.00017002738554352552,
+      "loss": 0.2949,
+      "step": 448
+    },
+    {
+      "epoch": 0.988442487616951,
+      "grad_norm": 1.8704255819320679,
+      "learning_rate": 0.00016984407772219077,
+      "loss": 0.1561,
+      "step": 449
+    },
+    {
+      "epoch": 0.9906439185470556,
+      "grad_norm": 0.9276111125946045,
+      "learning_rate": 0.00016966031053689828,
+      "loss": 0.2613,
+      "step": 450
+    },
+    {
+      "epoch": 0.9928453494771602,
+      "grad_norm": 1.162001132965088,
+      "learning_rate": 0.00016947608519628343,
+      "loss": 0.2439,
+      "step": 451
+    },
+    {
+      "epoch": 0.9950467804072647,
+      "grad_norm": 0.7540222406387329,
+      "learning_rate": 0.00016929140291199483,
+      "loss": 0.1343,
+      "step": 452
+    },
+    {
+      "epoch": 0.9972482113373693,
+      "grad_norm": 1.1655153036117554,
+      "learning_rate": 0.00016910626489868649,
+      "loss": 0.3202,
+      "step": 453
+    },
+    {
+      "epoch": 0.9994496422674739,
+      "grad_norm": 1.484333872795105,
+      "learning_rate": 0.00016892067237400958,
+      "loss": 0.2112,
+      "step": 454
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4983161985874176,
+      "learning_rate": 0.0001687346265586048,
+      "loss": 0.0172,
+      "step": 455
+    },
+    {
+      "epoch": 1.0022014309301046,
+      "grad_norm": 1.3438286781311035,
+      "learning_rate": 0.00016854812867609398,
+      "loss": 0.2078,
+      "step": 456
+    },
+    {
+      "epoch": 1.0044028618602092,
+      "grad_norm": 0.8171751499176025,
+      "learning_rate": 0.00016836117995307226,
+      "loss": 0.1252,
+      "step": 457
+    },
+    {
+      "epoch": 1.0066042927903136,
+      "grad_norm": 0.8060850501060486,
+      "learning_rate": 0.00016817378161909996,
+      "loss": 0.0609,
+      "step": 458
+    },
+    {
+      "epoch": 1.0088057237204182,
+      "grad_norm": 0.9792287945747375,
+      "learning_rate": 0.00016798593490669446,
+      "loss": 0.1923,
+      "step": 459
+    },
+    {
+      "epoch": 1.0110071546505228,
+      "grad_norm": 0.7354319095611572,
+      "learning_rate": 0.0001677976410513221,
+      "loss": 0.1241,
+      "step": 460
+    },
+    {
+      "epoch": 1.0132085855806274,
+      "grad_norm": 0.41206568479537964,
+      "learning_rate": 0.00016760890129139012,
+      "loss": 0.0991,
+      "step": 461
+    },
+    {
+      "epoch": 1.015410016510732,
+      "grad_norm": 1.4063104391098022,
+      "learning_rate": 0.0001674197168682385,
+      "loss": 0.2867,
+      "step": 462
+    },
+    {
+      "epoch": 1.0176114474408366,
+      "grad_norm": 0.6827117204666138,
+      "learning_rate": 0.0001672300890261317,
+      "loss": 0.1367,
+      "step": 463
+    },
+    {
+      "epoch": 1.019812878370941,
+      "grad_norm": 1.3953988552093506,
+      "learning_rate": 0.00016704001901225051,
+      "loss": 0.2906,
+      "step": 464
+    },
+    {
+      "epoch": 1.0220143093010456,
+      "grad_norm": 0.34882837533950806,
+      "learning_rate": 0.00016684950807668405,
+      "loss": 0.0403,
+      "step": 465
+    },
+    {
+      "epoch": 1.0242157402311503,
+      "grad_norm": 1.344286561012268,
+      "learning_rate": 0.00016665855747242117,
+      "loss": 0.1794,
+      "step": 466
+    },
+    {
+      "epoch": 1.0264171711612549,
+      "grad_norm": 1.2556302547454834,
+      "learning_rate": 0.0001664671684553426,
+      "loss": 0.1342,
+      "step": 467
+    },
+    {
+      "epoch": 1.0286186020913595,
+      "grad_norm": 0.5322070121765137,
+      "learning_rate": 0.0001662753422842123,
+      "loss": 0.0844,
+      "step": 468
+    },
+    {
+      "epoch": 1.0308200330214639,
+      "grad_norm": 1.2713350057601929,
+      "learning_rate": 0.0001660830802206696,
+      "loss": 0.1843,
+      "step": 469
+    },
+    {
+      "epoch": 1.0330214639515685,
+      "grad_norm": 1.4414249658584595,
+      "learning_rate": 0.00016589038352922062,
+      "loss": 0.3473,
+      "step": 470
+    },
+    {
+      "epoch": 1.035222894881673,
+      "grad_norm": 0.9836573600769043,
+      "learning_rate": 0.00016569725347722995,
+      "loss": 0.1188,
+      "step": 471
+    },
+    {
+      "epoch": 1.0374243258117777,
+      "grad_norm": 1.6469132900238037,
+      "learning_rate": 0.00016550369133491247,
+      "loss": 0.2775,
+      "step": 472
+    },
+    {
+      "epoch": 1.0396257567418823,
+      "grad_norm": 1.406348705291748,
+      "learning_rate": 0.00016530969837532487,
+      "loss": 0.2294,
+      "step": 473
+    },
+    {
+      "epoch": 1.041827187671987,
+      "grad_norm": 1.0122801065444946,
+      "learning_rate": 0.00016511527587435737,
+      "loss": 0.184,
+      "step": 474
+    },
+    {
+      "epoch": 1.0440286186020913,
+      "grad_norm": 0.5197072625160217,
+      "learning_rate": 0.0001649204251107252,
+      "loss": 0.0792,
+      "step": 475
+    },
+    {
+      "epoch": 1.046230049532196,
+      "grad_norm": 1.076872706413269,
+      "learning_rate": 0.0001647251473659604,
+      "loss": 0.1949,
+      "step": 476
+    },
+    {
+      "epoch": 1.0484314804623005,
+      "grad_norm": 0.6529151201248169,
+      "learning_rate": 0.00016452944392440313,
+      "loss": 0.1119,
+      "step": 477
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.9290107488632202,
+      "learning_rate": 0.00016433331607319343,
+      "loss": 0.139,
+      "step": 478
+    },
+    {
+      "epoch": 1.0528343423225097,
+      "grad_norm": 1.1986249685287476,
+      "learning_rate": 0.0001641367651022626,
+      "loss": 0.1887,
+      "step": 479
+    },
+    {
+      "epoch": 1.055035773252614,
+      "grad_norm": 0.7523142695426941,
+      "learning_rate": 0.00016393979230432495,
+      "loss": 0.1136,
+      "step": 480
+    },
+    {
+      "epoch": 1.0572372041827187,
+      "grad_norm": 0.8724271655082703,
+      "learning_rate": 0.000163742398974869,
+      "loss": 0.1314,
+      "step": 481
+    },
+    {
+      "epoch": 1.0594386351128233,
+      "grad_norm": 1.1953330039978027,
+      "learning_rate": 0.00016354458641214912,
+      "loss": 0.2461,
+      "step": 482
+    },
+    {
+      "epoch": 1.061640066042928,
+      "grad_norm": 0.635246217250824,
+      "learning_rate": 0.00016334635591717703,
+      "loss": 0.0915,
+      "step": 483
+    },
+    {
+      "epoch": 1.0638414969730325,
+      "grad_norm": 0.8510032892227173,
+      "learning_rate": 0.00016314770879371314,
+      "loss": 0.1231,
+      "step": 484
+    },
+    {
+      "epoch": 1.066042927903137,
+      "grad_norm": 0.772785484790802,
+      "learning_rate": 0.00016294864634825804,
+      "loss": 0.1074,
+      "step": 485
+    },
+    {
+      "epoch": 1.0682443588332415,
+      "grad_norm": 1.1119129657745361,
+      "learning_rate": 0.00016274916989004388,
+      "loss": 0.2063,
+      "step": 486
+    },
+    {
+      "epoch": 1.0704457897633461,
+      "grad_norm": 0.6170832514762878,
+      "learning_rate": 0.00016254928073102586,
+      "loss": 0.1263,
+      "step": 487
+    },
+    {
+      "epoch": 1.0726472206934508,
+      "grad_norm": 0.5298228859901428,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 0.0789,
+      "step": 488
+    },
+    {
+      "epoch": 1.0748486516235554,
+      "grad_norm": 1.2252048254013062,
+      "learning_rate": 0.00016214826957196152,
+      "loss": 0.168,
+      "step": 489
+    },
+    {
+      "epoch": 1.07705008255366,
+      "grad_norm": 1.7876496315002441,
+      "learning_rate": 0.00016194715020936249,
+      "loss": 0.2594,
+      "step": 490
+    },
+    {
+      "epoch": 1.0792515134837644,
+      "grad_norm": 1.0999908447265625,
+      "learning_rate": 0.00016174562342083677,
+      "loss": 0.2041,
+      "step": 491
+    },
+    {
+      "epoch": 1.081452944413869,
+      "grad_norm": 0.8964436650276184,
+      "learning_rate": 0.00016154369053182448,
+      "loss": 0.1144,
+      "step": 492
+    },
+    {
+      "epoch": 1.0836543753439736,
+      "grad_norm": 0.6843191981315613,
+      "learning_rate": 0.00016134135287043669,
+      "loss": 0.1088,
+      "step": 493
+    },
+    {
+      "epoch": 1.0858558062740782,
+      "grad_norm": 0.8405723571777344,
+      "learning_rate": 0.00016113861176744658,
+      "loss": 0.1813,
+      "step": 494
+    },
+    {
+      "epoch": 1.0880572372041828,
+      "grad_norm": 1.5375802516937256,
+      "learning_rate": 0.00016093546855628084,
+      "loss": 0.1916,
+      "step": 495
+    },
+    {
+      "epoch": 1.0902586681342874,
+      "grad_norm": 0.6426215171813965,
+      "learning_rate": 0.0001607319245730108,
+      "loss": 0.1028,
+      "step": 496
+    },
+    {
+      "epoch": 1.0924600990643918,
+      "grad_norm": 1.8465524911880493,
+      "learning_rate": 0.00016052798115634363,
+      "loss": 0.2657,
+      "step": 497
+    },
+    {
+      "epoch": 1.0946615299944964,
+      "grad_norm": 0.8891614079475403,
+      "learning_rate": 0.00016032363964761363,
+      "loss": 0.1123,
+      "step": 498
+    },
+    {
+      "epoch": 1.096862960924601,
+      "grad_norm": 1.3690117597579956,
+      "learning_rate": 0.00016011890139077336,
+      "loss": 0.1562,
+      "step": 499
+    },
+    {
+      "epoch": 1.0990643918547056,
+      "grad_norm": 0.9154755473136902,
+      "learning_rate": 0.00015991376773238466,
+      "loss": 0.1794,
+      "step": 500
+    },
+    {
+      "epoch": 1.1012658227848102,
+      "grad_norm": 0.8762046694755554,
+      "learning_rate": 0.00015970824002161006,
+      "loss": 0.2361,
+      "step": 501
+    },
+    {
+      "epoch": 1.1034672537149146,
+      "grad_norm": 0.7520323991775513,
+      "learning_rate": 0.00015950231961020373,
+      "loss": 0.0894,
+      "step": 502
+    },
+    {
+      "epoch": 1.1056686846450192,
+      "grad_norm": 1.0838078260421753,
+      "learning_rate": 0.00015929600785250257,
+      "loss": 0.1485,
+      "step": 503
+    },
+    {
+      "epoch": 1.1078701155751238,
+      "grad_norm": 1.2119700908660889,
+      "learning_rate": 0.00015908930610541747,
+      "loss": 0.2524,
+      "step": 504
+    },
+    {
+      "epoch": 1.1100715465052284,
+      "grad_norm": 1.8041168451309204,
+      "learning_rate": 0.00015888221572842422,
+      "loss": 0.2858,
+      "step": 505
+    },
+    {
+      "epoch": 1.112272977435333,
+      "grad_norm": 0.9385707378387451,
+      "learning_rate": 0.00015867473808355454,
+      "loss": 0.1342,
+      "step": 506
+    },
+    {
+      "epoch": 1.1144744083654374,
+      "grad_norm": 0.5351221561431885,
+      "learning_rate": 0.00015846687453538734,
+      "loss": 0.1107,
+      "step": 507
+    },
+    {
+      "epoch": 1.116675839295542,
+      "grad_norm": 0.820961594581604,
+      "learning_rate": 0.0001582586264510396,
+      "loss": 0.1046,
+      "step": 508
+    },
+    {
+      "epoch": 1.1188772702256466,
+      "grad_norm": 0.18447497487068176,
+      "learning_rate": 0.00015804999520015734,
+      "loss": 0.0299,
+      "step": 509
+    },
+    {
+      "epoch": 1.1210787011557513,
+      "grad_norm": 0.5453505516052246,
+      "learning_rate": 0.00015784098215490669,
+      "loss": 0.095,
+      "step": 510
+    },
+    {
+      "epoch": 1.1232801320858559,
+      "grad_norm": 1.1057016849517822,
+      "learning_rate": 0.00015763158868996476,
+      "loss": 0.2143,
+      "step": 511
+    },
+    {
+      "epoch": 1.1254815630159603,
+      "grad_norm": 0.9430788159370422,
+      "learning_rate": 0.00015742181618251083,
+      "loss": 0.2118,
+      "step": 512
+    },
+    {
+      "epoch": 1.1276829939460649,
+      "grad_norm": 0.694381833076477,
+      "learning_rate": 0.00015721166601221698,
+      "loss": 0.0908,
+      "step": 513
+    },
+    {
+      "epoch": 1.1298844248761695,
+      "grad_norm": 0.6499952077865601,
+      "learning_rate": 0.00015700113956123928,
+      "loss": 0.0757,
+      "step": 514
+    },
+    {
+      "epoch": 1.132085855806274,
+      "grad_norm": 0.2832721173763275,
+      "learning_rate": 0.0001567902382142086,
+      "loss": 0.0268,
+      "step": 515
+    },
+    {
+      "epoch": 1.1342872867363787,
+      "grad_norm": 1.6799596548080444,
+      "learning_rate": 0.00015657896335822147,
+      "loss": 0.4062,
+      "step": 516
+    },
+    {
+      "epoch": 1.1364887176664833,
+      "grad_norm": 0.8721389770507812,
+      "learning_rate": 0.0001563673163828309,
+      "loss": 0.1643,
+      "step": 517
+    },
+    {
+      "epoch": 1.138690148596588,
+      "grad_norm": 1.1544523239135742,
+      "learning_rate": 0.0001561552986800375,
+      "loss": 0.1405,
+      "step": 518
+    },
+    {
+      "epoch": 1.1408915795266923,
+      "grad_norm": 1.1611992120742798,
+      "learning_rate": 0.00015594291164427996,
+      "loss": 0.1905,
+      "step": 519
+    },
+    {
+      "epoch": 1.143093010456797,
+      "grad_norm": 0.7311229705810547,
+      "learning_rate": 0.00015573015667242625,
+      "loss": 0.1634,
+      "step": 520
+    },
+    {
+      "epoch": 1.1452944413869015,
+      "grad_norm": 1.469364881515503,
+      "learning_rate": 0.00015551703516376416,
+      "loss": 0.2812,
+      "step": 521
+    },
+    {
+      "epoch": 1.1474958723170061,
+      "grad_norm": 0.8396485447883606,
+      "learning_rate": 0.00015530354851999215,
+      "loss": 0.1058,
+      "step": 522
+    },
+    {
+      "epoch": 1.1496973032471107,
+      "grad_norm": 0.8667928576469421,
+      "learning_rate": 0.00015508969814521025,
+      "loss": 0.1181,
+      "step": 523
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.915149450302124,
+      "learning_rate": 0.00015487548544591075,
+      "loss": 0.1272,
+      "step": 524
+    },
+    {
+      "epoch": 1.1541001651073197,
+      "grad_norm": 1.0588988065719604,
+      "learning_rate": 0.00015466091183096884,
+      "loss": 0.1298,
+      "step": 525
+    },
+    {
+      "epoch": 1.1563015960374243,
+      "grad_norm": 1.0772982835769653,
+      "learning_rate": 0.0001544459787116336,
+      "loss": 0.1364,
+      "step": 526
+    },
+    {
+      "epoch": 1.158503026967529,
+      "grad_norm": 1.4910756349563599,
+      "learning_rate": 0.0001542306875015185,
+      "loss": 0.1675,
+      "step": 527
+    },
+    {
+      "epoch": 1.1607044578976335,
+      "grad_norm": 0.6280027031898499,
+      "learning_rate": 0.00015401503961659204,
+      "loss": 0.0791,
+      "step": 528
+    },
+    {
+      "epoch": 1.162905888827738,
+      "grad_norm": 0.6094567179679871,
+      "learning_rate": 0.0001537990364751688,
+      "loss": 0.096,
+      "step": 529
+    },
+    {
+      "epoch": 1.1651073197578425,
+      "grad_norm": 0.8040492534637451,
+      "learning_rate": 0.00015358267949789966,
+      "loss": 0.153,
+      "step": 530
+    },
+    {
+      "epoch": 1.1673087506879471,
+      "grad_norm": 0.397973895072937,
+      "learning_rate": 0.0001533659701077629,
+      "loss": 0.0614,
+      "step": 531
+    },
+    {
+      "epoch": 1.1695101816180518,
+      "grad_norm": 0.7243671417236328,
+      "learning_rate": 0.00015314890973005448,
+      "loss": 0.1102,
+      "step": 532
+    },
+    {
+      "epoch": 1.1717116125481564,
+      "grad_norm": 0.4522433876991272,
+      "learning_rate": 0.00015293149979237876,
+      "loss": 0.0712,
+      "step": 533
+    },
+    {
+      "epoch": 1.1739130434782608,
+      "grad_norm": 0.4651868939399719,
+      "learning_rate": 0.00015271374172463924,
+      "loss": 0.0917,
+      "step": 534
+    },
+    {
+      "epoch": 1.1761144744083654,
+      "grad_norm": 2.3681793212890625,
+      "learning_rate": 0.00015249563695902906,
+      "loss": 0.1205,
+      "step": 535
+    },
+    {
+      "epoch": 1.17831590533847,
+      "grad_norm": 1.3707005977630615,
+      "learning_rate": 0.00015227718693002154,
+      "loss": 0.1908,
+      "step": 536
+    },
+    {
+      "epoch": 1.1805173362685746,
+      "grad_norm": 1.3730310201644897,
+      "learning_rate": 0.00015205839307436088,
+      "loss": 0.3092,
+      "step": 537
+    },
+    {
+      "epoch": 1.1827187671986792,
+      "grad_norm": 0.6282625794410706,
+      "learning_rate": 0.00015183925683105254,
+      "loss": 0.1166,
+      "step": 538
+    },
+    {
+      "epoch": 1.1849201981287838,
+      "grad_norm": 0.6953176259994507,
+      "learning_rate": 0.00015161977964135386,
+      "loss": 0.0541,
+      "step": 539
+    },
+    {
+      "epoch": 1.1871216290588882,
+      "grad_norm": 1.0111442804336548,
+      "learning_rate": 0.00015139996294876465,
+      "loss": 0.1917,
+      "step": 540
+    },
+    {
+      "epoch": 1.1893230599889928,
+      "grad_norm": 2.3661305904388428,
+      "learning_rate": 0.00015117980819901762,
+      "loss": 0.2784,
+      "step": 541
+    },
+    {
+      "epoch": 1.1915244909190974,
+      "grad_norm": 0.5237095355987549,
+      "learning_rate": 0.00015095931684006882,
+      "loss": 0.0515,
+      "step": 542
+    },
+    {
+      "epoch": 1.193725921849202,
+      "grad_norm": 0.959276020526886,
+      "learning_rate": 0.00015073849032208822,
+      "loss": 0.1441,
+      "step": 543
+    },
+    {
+      "epoch": 1.1959273527793066,
+      "grad_norm": 1.3492348194122314,
+      "learning_rate": 0.00015051733009745013,
+      "loss": 0.2327,
+      "step": 544
+    },
+    {
+      "epoch": 1.1981287837094112,
+      "grad_norm": 0.9852789044380188,
+      "learning_rate": 0.00015029583762072357,
+      "loss": 0.1306,
+      "step": 545
+    },
+    {
+      "epoch": 1.2003302146395156,
+      "grad_norm": 0.60108482837677,
+      "learning_rate": 0.0001500740143486629,
+      "loss": 0.1524,
+      "step": 546
+    },
+    {
+      "epoch": 1.2025316455696202,
+      "grad_norm": 1.4141664505004883,
+      "learning_rate": 0.00014985186174019805,
+      "loss": 0.332,
+      "step": 547
+    },
+    {
+      "epoch": 1.2047330764997248,
+      "grad_norm": 0.6472118496894836,
+      "learning_rate": 0.00014962938125642503,
+      "loss": 0.0793,
+      "step": 548
+    },
+    {
+      "epoch": 1.2069345074298294,
+      "grad_norm": 0.9237731695175171,
+      "learning_rate": 0.00014940657436059618,
+      "loss": 0.1123,
+      "step": 549
+    },
+    {
+      "epoch": 1.209135938359934,
+      "grad_norm": 0.8630395531654358,
+      "learning_rate": 0.0001491834425181108,
+      "loss": 0.1285,
+      "step": 550
+    },
+    {
+      "epoch": 1.2113373692900384,
+      "grad_norm": 0.5758857727050781,
+      "learning_rate": 0.00014895998719650526,
+      "loss": 0.0914,
+      "step": 551
+    },
+    {
+      "epoch": 1.213538800220143,
+      "grad_norm": 0.7132592797279358,
+      "learning_rate": 0.00014873620986544347,
+      "loss": 0.0987,
+      "step": 552
+    },
+    {
+      "epoch": 1.2157402311502477,
+      "grad_norm": 1.0615679025650024,
+      "learning_rate": 0.00014851211199670721,
+      "loss": 0.1668,
+      "step": 553
+    },
+    {
+      "epoch": 1.2179416620803523,
+      "grad_norm": 0.6403416991233826,
+      "learning_rate": 0.00014828769506418645,
+      "loss": 0.0932,
+      "step": 554
+    },
+    {
+      "epoch": 1.2201430930104569,
+      "grad_norm": 0.8760681748390198,
+      "learning_rate": 0.0001480629605438696,
+      "loss": 0.1528,
+      "step": 555
+    },
+    {
+      "epoch": 1.2223445239405613,
+      "grad_norm": 0.9847756624221802,
+      "learning_rate": 0.0001478379099138338,
+      "loss": 0.1526,
+      "step": 556
+    },
+    {
+      "epoch": 1.2245459548706659,
+      "grad_norm": 0.3219001889228821,
+      "learning_rate": 0.00014761254465423538,
+      "loss": 0.0314,
+      "step": 557
+    },
+    {
+      "epoch": 1.2267473858007705,
+      "grad_norm": 0.6814752817153931,
+      "learning_rate": 0.00014738686624729986,
+      "loss": 0.0962,
+      "step": 558
+    },
+    {
+      "epoch": 1.228948816730875,
+      "grad_norm": 0.675477147102356,
+      "learning_rate": 0.00014716087617731242,
+      "loss": 0.0931,
+      "step": 559
+    },
+    {
+      "epoch": 1.2311502476609797,
+      "grad_norm": 1.599684476852417,
+      "learning_rate": 0.00014693457593060796,
+      "loss": 0.1536,
+      "step": 560
+    },
+    {
+      "epoch": 1.2333516785910843,
+      "grad_norm": 0.7459136247634888,
+      "learning_rate": 0.00014670796699556144,
+      "loss": 0.1696,
+      "step": 561
+    },
+    {
+      "epoch": 1.2355531095211887,
+      "grad_norm": 1.097995400428772,
+      "learning_rate": 0.00014648105086257807,
+      "loss": 0.1535,
+      "step": 562
+    },
+    {
+      "epoch": 1.2377545404512933,
+      "grad_norm": 0.36836791038513184,
+      "learning_rate": 0.00014625382902408356,
+      "loss": 0.0999,
+      "step": 563
+    },
+    {
+      "epoch": 1.239955971381398,
+      "grad_norm": 2.6130542755126953,
+      "learning_rate": 0.00014602630297451408,
+      "loss": 0.2079,
+      "step": 564
+    },
+    {
+      "epoch": 1.2421574023115025,
+      "grad_norm": 0.8240204453468323,
+      "learning_rate": 0.00014579847421030678,
+      "loss": 0.1552,
+      "step": 565
+    },
+    {
+      "epoch": 1.2443588332416071,
+      "grad_norm": 0.9083564877510071,
+      "learning_rate": 0.0001455703442298896,
+      "loss": 0.1677,
+      "step": 566
+    },
+    {
+      "epoch": 1.2465602641717117,
+      "grad_norm": 0.6768250465393066,
+      "learning_rate": 0.00014534191453367175,
+      "loss": 0.0819,
+      "step": 567
+    },
+    {
+      "epoch": 1.2487616951018161,
+      "grad_norm": 1.417345643043518,
+      "learning_rate": 0.00014511318662403347,
+      "loss": 0.2168,
+      "step": 568
+    },
+    {
+      "epoch": 1.2509631260319207,
+      "grad_norm": 1.0408700704574585,
+      "learning_rate": 0.00014488416200531651,
+      "loss": 0.1357,
+      "step": 569
+    },
+    {
+      "epoch": 1.2531645569620253,
+      "grad_norm": 0.49575138092041016,
+      "learning_rate": 0.000144654842183814,
+      "loss": 0.1213,
+      "step": 570
+    },
+    {
+      "epoch": 1.25536598789213,
+      "grad_norm": 0.47278714179992676,
+      "learning_rate": 0.00014442522866776062,
+      "loss": 0.0639,
+      "step": 571
+    },
+    {
+      "epoch": 1.2575674188222346,
+      "grad_norm": 0.6846803426742554,
+      "learning_rate": 0.0001441953229673227,
+      "loss": 0.0893,
+      "step": 572
+    },
+    {
+      "epoch": 1.259768849752339,
+      "grad_norm": 0.7213048934936523,
+      "learning_rate": 0.00014396512659458824,
+      "loss": 0.0702,
+      "step": 573
+    },
+    {
+      "epoch": 1.2619702806824435,
+      "grad_norm": 0.665086030960083,
+      "learning_rate": 0.00014373464106355696,
+      "loss": 0.0753,
+      "step": 574
+    },
+    {
+      "epoch": 1.2641717116125482,
+      "grad_norm": 0.5565153956413269,
+      "learning_rate": 0.00014350386789013046,
+      "loss": 0.0627,
+      "step": 575
+    },
+    {
+      "epoch": 1.2663731425426528,
+      "grad_norm": 0.5590353012084961,
+      "learning_rate": 0.00014327280859210214,
+      "loss": 0.0624,
+      "step": 576
+    },
+    {
+      "epoch": 1.2685745734727574,
+      "grad_norm": 0.8887538909912109,
+      "learning_rate": 0.00014304146468914713,
+      "loss": 0.1523,
+      "step": 577
+    },
+    {
+      "epoch": 1.2707760044028618,
+      "grad_norm": 0.37396806478500366,
+      "learning_rate": 0.0001428098377028126,
+      "loss": 0.0711,
+      "step": 578
+    },
+    {
+      "epoch": 1.2729774353329664,
+      "grad_norm": 0.6387276649475098,
+      "learning_rate": 0.00014257792915650728,
+      "loss": 0.1033,
+      "step": 579
+    },
+    {
+      "epoch": 1.275178866263071,
+      "grad_norm": 1.458855390548706,
+      "learning_rate": 0.000142345740575492,
+      "loss": 0.0562,
+      "step": 580
+    },
+    {
+      "epoch": 1.2773802971931756,
+      "grad_norm": 1.5706512928009033,
+      "learning_rate": 0.00014211327348686925,
+      "loss": 0.2097,
+      "step": 581
+    },
+    {
+      "epoch": 1.2795817281232802,
+      "grad_norm": 1.3540840148925781,
+      "learning_rate": 0.00014188052941957325,
+      "loss": 0.146,
+      "step": 582
+    },
+    {
+      "epoch": 1.2817831590533846,
+      "grad_norm": 1.8784266710281372,
+      "learning_rate": 0.0001416475099043599,
+      "loss": 0.2617,
+      "step": 583
+    },
+    {
+      "epoch": 1.2839845899834892,
+      "grad_norm": 0.8808488249778748,
+      "learning_rate": 0.0001414142164737968,
+      "loss": 0.1448,
+      "step": 584
+    },
+    {
+      "epoch": 1.2861860209135938,
+      "grad_norm": 0.9543077945709229,
+      "learning_rate": 0.00014118065066225303,
+      "loss": 0.1621,
+      "step": 585
+    },
+    {
+      "epoch": 1.2883874518436984,
+      "grad_norm": 0.8723201155662537,
+      "learning_rate": 0.00014094681400588906,
+      "loss": 0.1472,
+      "step": 586
+    },
+    {
+      "epoch": 1.290588882773803,
+      "grad_norm": 1.5932315587997437,
+      "learning_rate": 0.0001407127080426468,
+      "loss": 0.2377,
+      "step": 587
+    },
+    {
+      "epoch": 1.2927903137039076,
+      "grad_norm": 1.116560935974121,
+      "learning_rate": 0.00014047833431223938,
+      "loss": 0.2282,
+      "step": 588
+    },
+    {
+      "epoch": 1.2949917446340122,
+      "grad_norm": 0.62128746509552,
+      "learning_rate": 0.00014024369435614102,
+      "loss": 0.1086,
+      "step": 589
+    },
+    {
+      "epoch": 1.2971931755641166,
+      "grad_norm": 0.3911591172218323,
+      "learning_rate": 0.0001400087897175768,
+      "loss": 0.05,
+      "step": 590
+    },
+    {
+      "epoch": 1.2993946064942212,
+      "grad_norm": 0.8413006663322449,
+      "learning_rate": 0.0001397736219415128,
+      "loss": 0.1577,
+      "step": 591
+    },
+    {
+      "epoch": 1.3015960374243258,
+      "grad_norm": 1.405246615409851,
+      "learning_rate": 0.0001395381925746456,
+      "loss": 0.2257,
+      "step": 592
+    },
+    {
+      "epoch": 1.3037974683544304,
+      "grad_norm": 0.6087014079093933,
+      "learning_rate": 0.00013930250316539238,
+      "loss": 0.1669,
+      "step": 593
+    },
+    {
+      "epoch": 1.305998899284535,
+      "grad_norm": 0.5942371487617493,
+      "learning_rate": 0.0001390665552638805,
+      "loss": 0.1612,
+      "step": 594
+    },
+    {
+      "epoch": 1.3082003302146394,
+      "grad_norm": 1.0461444854736328,
+      "learning_rate": 0.00013883035042193753,
+      "loss": 0.1825,
+      "step": 595
+    },
+    {
+      "epoch": 1.310401761144744,
+      "grad_norm": 0.791422963142395,
+      "learning_rate": 0.00013859389019308085,
+      "loss": 0.0849,
+      "step": 596
+    },
+    {
+      "epoch": 1.3126031920748487,
+      "grad_norm": 0.43431755900382996,
+      "learning_rate": 0.00013835717613250754,
+      "loss": 0.063,
+      "step": 597
+    },
+    {
+      "epoch": 1.3148046230049533,
+      "grad_norm": 0.6389137506484985,
+      "learning_rate": 0.00013812020979708418,
+      "loss": 0.0962,
+      "step": 598
+    },
+    {
+      "epoch": 1.3170060539350579,
+      "grad_norm": 0.7983347773551941,
+      "learning_rate": 0.0001378829927453365,
+      "loss": 0.1981,
+      "step": 599
+    },
+    {
+      "epoch": 1.3192074848651623,
+      "grad_norm": 0.9866439700126648,
+      "learning_rate": 0.0001376455265374392,
+      "loss": 0.2923,
+      "step": 600
+    },
+    {
+      "epoch": 1.3214089157952669,
+      "grad_norm": 1.0503846406936646,
+      "learning_rate": 0.00013740781273520572,
+      "loss": 0.1727,
+      "step": 601
+    },
+    {
+      "epoch": 1.3236103467253715,
+      "grad_norm": 1.300238013267517,
+      "learning_rate": 0.00013716985290207788,
+      "loss": 0.1419,
+      "step": 602
+    },
+    {
+      "epoch": 1.325811777655476,
+      "grad_norm": 1.0286176204681396,
+      "learning_rate": 0.00013693164860311565,
+      "loss": 0.2083,
+      "step": 603
+    },
+    {
+      "epoch": 1.3280132085855807,
+      "grad_norm": 0.5792612433433533,
+      "learning_rate": 0.00013669320140498686,
+      "loss": 0.0672,
+      "step": 604
+    },
+    {
+      "epoch": 1.330214639515685,
+      "grad_norm": 0.5995973944664001,
+      "learning_rate": 0.00013645451287595686,
+      "loss": 0.1223,
+      "step": 605
+    },
+    {
+      "epoch": 1.3324160704457897,
+      "grad_norm": 1.065690517425537,
+      "learning_rate": 0.00013621558458587832,
+      "loss": 0.148,
+      "step": 606
+    },
+    {
+      "epoch": 1.3346175013758943,
+      "grad_norm": 1.202343463897705,
+      "learning_rate": 0.00013597641810618073,
+      "loss": 0.2709,
+      "step": 607
+    },
+    {
+      "epoch": 1.336818932305999,
+      "grad_norm": 0.6058682799339294,
+      "learning_rate": 0.0001357370150098601,
+      "loss": 0.0968,
+      "step": 608
+    },
+    {
+      "epoch": 1.3390203632361035,
+      "grad_norm": 0.9532506465911865,
+      "learning_rate": 0.0001354973768714688,
+      "loss": 0.0983,
+      "step": 609
+    },
+    {
+      "epoch": 1.341221794166208,
+      "grad_norm": 0.912371039390564,
+      "learning_rate": 0.00013525750526710502,
+      "loss": 0.2221,
+      "step": 610
+    },
+    {
+      "epoch": 1.3434232250963127,
+      "grad_norm": 0.5274757742881775,
+      "learning_rate": 0.00013501740177440243,
+      "loss": 0.0504,
+      "step": 611
+    },
+    {
+      "epoch": 1.3456246560264171,
+      "grad_norm": 1.182218074798584,
+      "learning_rate": 0.00013477706797251985,
+      "loss": 0.2155,
+      "step": 612
+    },
+    {
+      "epoch": 1.3478260869565217,
+      "grad_norm": 1.1047543287277222,
+      "learning_rate": 0.00013453650544213076,
+      "loss": 0.1104,
+      "step": 613
+    },
+    {
+      "epoch": 1.3500275178866263,
+      "grad_norm": 0.48692095279693604,
+      "learning_rate": 0.00013429571576541315,
+      "loss": 0.0808,
+      "step": 614
+    },
+    {
+      "epoch": 1.352228948816731,
+      "grad_norm": 1.084601879119873,
+      "learning_rate": 0.00013405470052603883,
+      "loss": 0.2284,
+      "step": 615
+    },
+    {
+      "epoch": 1.3544303797468356,
+      "grad_norm": 1.1326749324798584,
+      "learning_rate": 0.00013381346130916314,
+      "loss": 0.3293,
+      "step": 616
+    },
+    {
+      "epoch": 1.35663181067694,
+      "grad_norm": 2.1683349609375,
+      "learning_rate": 0.00013357199970141456,
+      "loss": 0.187,
+      "step": 617
+    },
+    {
+      "epoch": 1.3588332416070446,
+      "grad_norm": 1.2375034093856812,
+      "learning_rate": 0.00013333031729088419,
+      "loss": 0.1319,
+      "step": 618
+    },
+    {
+      "epoch": 1.3610346725371492,
+      "grad_norm": 1.8519262075424194,
+      "learning_rate": 0.00013308841566711538,
+      "loss": 0.2359,
+      "step": 619
+    },
+    {
+      "epoch": 1.3632361034672538,
+      "grad_norm": 1.2038832902908325,
+      "learning_rate": 0.00013284629642109326,
+      "loss": 0.1733,
+      "step": 620
+    },
+    {
+      "epoch": 1.3654375343973584,
+      "grad_norm": 0.7529458403587341,
+      "learning_rate": 0.0001326039611452342,
+      "loss": 0.1321,
+      "step": 621
+    },
+    {
+      "epoch": 1.3676389653274628,
+      "grad_norm": 0.4542110860347748,
+      "learning_rate": 0.0001323614114333754,
+      "loss": 0.0424,
+      "step": 622
+    },
+    {
+      "epoch": 1.3698403962575674,
+      "grad_norm": 1.1919138431549072,
+      "learning_rate": 0.00013211864888076457,
+      "loss": 0.1559,
+      "step": 623
+    },
+    {
+      "epoch": 1.372041827187672,
+      "grad_norm": 0.9733201265335083,
+      "learning_rate": 0.00013187567508404899,
+      "loss": 0.1407,
+      "step": 624
+    },
+    {
+      "epoch": 1.3742432581177766,
+      "grad_norm": 1.161691665649414,
+      "learning_rate": 0.00013163249164126548,
+      "loss": 0.2082,
+      "step": 625
+    },
+    {
+      "epoch": 1.3764446890478812,
+      "grad_norm": 0.7243776917457581,
+      "learning_rate": 0.0001313891001518297,
+      "loss": 0.1635,
+      "step": 626
+    },
+    {
+      "epoch": 1.3786461199779856,
+      "grad_norm": 1.8206489086151123,
+      "learning_rate": 0.00013114550221652552,
+      "loss": 0.1471,
+      "step": 627
+    },
+    {
+      "epoch": 1.3808475509080902,
+      "grad_norm": 0.9848284721374512,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.0941,
+      "step": 628
+    },
+    {
+      "epoch": 1.3830489818381948,
+      "grad_norm": 0.6347261667251587,
+      "learning_rate": 0.00013065769341822633,
+      "loss": 0.1172,
+      "step": 629
+    },
+    {
+      "epoch": 1.3852504127682994,
+      "grad_norm": 0.5745399594306946,
+      "learning_rate": 0.00013041348576354594,
+      "loss": 0.1323,
+      "step": 630
+    },
+    {
+      "epoch": 1.387451843698404,
+      "grad_norm": 1.2065634727478027,
+      "learning_rate": 0.0001301690780796055,
+      "loss": 0.1248,
+      "step": 631
+    },
+    {
+      "epoch": 1.3896532746285084,
+      "grad_norm": 0.7082602977752686,
+      "learning_rate": 0.00012992447197387238,
+      "loss": 0.1398,
+      "step": 632
+    },
+    {
+      "epoch": 1.3918547055586132,
+      "grad_norm": 0.7551670670509338,
+      "learning_rate": 0.00012967966905511906,
+      "loss": 0.0958,
+      "step": 633
+    },
+    {
+      "epoch": 1.3940561364887176,
+      "grad_norm": 0.9957584142684937,
+      "learning_rate": 0.00012943467093341247,
+      "loss": 0.114,
+      "step": 634
+    },
+    {
+      "epoch": 1.3962575674188222,
+      "grad_norm": 0.6230639815330505,
+      "learning_rate": 0.00012918947922010336,
+      "loss": 0.1101,
+      "step": 635
+    },
+    {
+      "epoch": 1.3984589983489268,
+      "grad_norm": 0.9289701581001282,
+      "learning_rate": 0.00012894409552781564,
+      "loss": 0.1617,
+      "step": 636
+    },
+    {
+      "epoch": 1.4006604292790314,
+      "grad_norm": 1.1827571392059326,
+      "learning_rate": 0.00012869852147043607,
+      "loss": 0.1768,
+      "step": 637
+    },
+    {
+      "epoch": 1.402861860209136,
+      "grad_norm": 0.6412714123725891,
+      "learning_rate": 0.00012845275866310324,
+      "loss": 0.082,
+      "step": 638
+    },
+    {
+      "epoch": 1.4050632911392404,
+      "grad_norm": 0.8040220141410828,
+      "learning_rate": 0.00012820680872219727,
+      "loss": 0.1265,
+      "step": 639
+    },
+    {
+      "epoch": 1.407264722069345,
+      "grad_norm": 0.885257363319397,
+      "learning_rate": 0.000127960673265329,
+      "loss": 0.141,
+      "step": 640
+    },
+    {
+      "epoch": 1.4094661529994497,
+      "grad_norm": 1.914450764656067,
+      "learning_rate": 0.00012771435391132943,
+      "loss": 0.4999,
+      "step": 641
+    },
+    {
+      "epoch": 1.4116675839295543,
+      "grad_norm": 1.6893082857131958,
+      "learning_rate": 0.00012746785228023904,
+      "loss": 0.2837,
+      "step": 642
+    },
+    {
+      "epoch": 1.4138690148596589,
+      "grad_norm": 0.7887259125709534,
+      "learning_rate": 0.00012722116999329712,
+      "loss": 0.1506,
+      "step": 643
+    },
+    {
+      "epoch": 1.4160704457897633,
+      "grad_norm": 0.9915180206298828,
+      "learning_rate": 0.00012697430867293118,
+      "loss": 0.1352,
+      "step": 644
+    },
+    {
+      "epoch": 1.4182718767198679,
+      "grad_norm": 0.6148739457130432,
+      "learning_rate": 0.00012672726994274622,
+      "loss": 0.1433,
+      "step": 645
+    },
+    {
+      "epoch": 1.4204733076499725,
+      "grad_norm": 0.6085465550422668,
+      "learning_rate": 0.00012648005542751405,
+      "loss": 0.128,
+      "step": 646
+    },
+    {
+      "epoch": 1.422674738580077,
+      "grad_norm": 1.2505244016647339,
+      "learning_rate": 0.00012623266675316263,
+      "loss": 0.1056,
+      "step": 647
+    },
+    {
+      "epoch": 1.4248761695101817,
+      "grad_norm": 1.1471660137176514,
+      "learning_rate": 0.0001259851055467653,
+      "loss": 0.181,
+      "step": 648
+    },
+    {
+      "epoch": 1.427077600440286,
+      "grad_norm": 0.570956289768219,
+      "learning_rate": 0.00012573737343653024,
+      "loss": 0.0658,
+      "step": 649
+    },
+    {
+      "epoch": 1.4292790313703907,
+      "grad_norm": 0.634922444820404,
+      "learning_rate": 0.00012548947205178962,
+      "loss": 0.1813,
+      "step": 650
+    },
+    {
+      "epoch": 1.4314804623004953,
+      "grad_norm": 0.567136287689209,
+      "learning_rate": 0.00012524140302298893,
+      "loss": 0.1212,
+      "step": 651
+    },
+    {
+      "epoch": 1.4336818932306,
+      "grad_norm": 0.8204668164253235,
+      "learning_rate": 0.00012499316798167623,
+      "loss": 0.134,
+      "step": 652
+    },
+    {
+      "epoch": 1.4358833241607045,
+      "grad_norm": 0.9615541696548462,
+      "learning_rate": 0.00012474476856049144,
+      "loss": 0.17,
+      "step": 653
+    },
+    {
+      "epoch": 1.438084755090809,
+      "grad_norm": 0.6434019207954407,
+      "learning_rate": 0.0001244962063931557,
+      "loss": 0.1325,
+      "step": 654
+    },
+    {
+      "epoch": 1.4402861860209135,
+      "grad_norm": 0.6178004145622253,
+      "learning_rate": 0.0001242474831144604,
+      "loss": 0.1104,
+      "step": 655
+    },
+    {
+      "epoch": 1.4424876169510181,
+      "grad_norm": 0.5323943495750427,
+      "learning_rate": 0.0001239986003602566,
+      "loss": 0.1131,
+      "step": 656
+    },
+    {
+      "epoch": 1.4446890478811227,
+      "grad_norm": 1.0922467708587646,
+      "learning_rate": 0.00012374955976744432,
+      "loss": 0.1887,
+      "step": 657
+    },
+    {
+      "epoch": 1.4468904788112273,
+      "grad_norm": 1.1991472244262695,
+      "learning_rate": 0.00012350036297396154,
+      "loss": 0.24,
+      "step": 658
+    },
+    {
+      "epoch": 1.4490919097413317,
+      "grad_norm": 1.2929867506027222,
+      "learning_rate": 0.00012325101161877363,
+      "loss": 0.1374,
+      "step": 659
+    },
+    {
+      "epoch": 1.4512933406714366,
+      "grad_norm": 0.5163611769676208,
+      "learning_rate": 0.00012300150734186258,
+      "loss": 0.0782,
+      "step": 660
+    },
+    {
+      "epoch": 1.453494771601541,
+      "grad_norm": 0.45565393567085266,
+      "learning_rate": 0.00012275185178421607,
+      "loss": 0.0621,
+      "step": 661
+    },
+    {
+      "epoch": 1.4556962025316456,
+      "grad_norm": 0.6713182926177979,
+      "learning_rate": 0.00012250204658781675,
+      "loss": 0.1608,
+      "step": 662
+    },
+    {
+      "epoch": 1.4578976334617502,
+      "grad_norm": 0.8330361247062683,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 0.1589,
+      "step": 663
+    },
+    {
+      "epoch": 1.4600990643918548,
+      "grad_norm": 0.5987771153450012,
+      "learning_rate": 0.00012200199385160041,
+      "loss": 0.0729,
+      "step": 664
+    },
+    {
+      "epoch": 1.4623004953219594,
+      "grad_norm": 0.5758820176124573,
+      "learning_rate": 0.00012175174960062637,
+      "loss": 0.0917,
+      "step": 665
+    },
+    {
+      "epoch": 1.4645019262520638,
+      "grad_norm": 1.1447787284851074,
+      "learning_rate": 0.00012150136228856381,
+      "loss": 0.1609,
+      "step": 666
+    },
+    {
+      "epoch": 1.4667033571821684,
+      "grad_norm": 1.4136334657669067,
+      "learning_rate": 0.00012125083356220817,
+      "loss": 0.1145,
+      "step": 667
+    },
+    {
+      "epoch": 1.468904788112273,
+      "grad_norm": 1.5868796110153198,
+      "learning_rate": 0.00012100016506928493,
+      "loss": 0.2602,
+      "step": 668
+    },
+    {
+      "epoch": 1.4711062190423776,
+      "grad_norm": 1.784375548362732,
+      "learning_rate": 0.00012074935845843887,
+      "loss": 0.1714,
+      "step": 669
+    },
+    {
+      "epoch": 1.4733076499724822,
+      "grad_norm": 1.1253080368041992,
+      "learning_rate": 0.00012049841537922307,
+      "loss": 0.207,
+      "step": 670
+    },
+    {
+      "epoch": 1.4755090809025866,
+      "grad_norm": 0.760846734046936,
+      "learning_rate": 0.00012024733748208819,
+      "loss": 0.1476,
+      "step": 671
+    },
+    {
+      "epoch": 1.4777105118326912,
+      "grad_norm": 0.581866443157196,
+      "learning_rate": 0.00011999612641837168,
+      "loss": 0.0876,
+      "step": 672
+    },
+    {
+      "epoch": 1.4799119427627958,
+      "grad_norm": 0.9836142659187317,
+      "learning_rate": 0.00011974478384028672,
+      "loss": 0.1357,
+      "step": 673
+    },
+    {
+      "epoch": 1.4821133736929004,
+      "grad_norm": 0.46399572491645813,
+      "learning_rate": 0.00011949331140091153,
+      "loss": 0.0771,
+      "step": 674
+    },
+    {
+      "epoch": 1.484314804623005,
+      "grad_norm": 0.7173429727554321,
+      "learning_rate": 0.00011924171075417836,
+      "loss": 0.1398,
+      "step": 675
+    },
+    {
+      "epoch": 1.4865162355531094,
+      "grad_norm": 1.0541839599609375,
+      "learning_rate": 0.00011898998355486273,
+      "loss": 0.2469,
+      "step": 676
+    },
+    {
+      "epoch": 1.488717666483214,
+      "grad_norm": 0.9371955394744873,
+      "learning_rate": 0.00011873813145857249,
+      "loss": 0.1503,
+      "step": 677
+    },
+    {
+      "epoch": 1.4909190974133186,
+      "grad_norm": 0.7344411611557007,
+      "learning_rate": 0.00011848615612173688,
+      "loss": 0.1076,
+      "step": 678
+    },
+    {
+      "epoch": 1.4931205283434232,
+      "grad_norm": 1.0650099515914917,
+      "learning_rate": 0.00011823405920159575,
+      "loss": 0.1796,
+      "step": 679
+    },
+    {
+      "epoch": 1.4953219592735278,
+      "grad_norm": 0.9302453994750977,
+      "learning_rate": 0.00011798184235618867,
+      "loss": 0.15,
+      "step": 680
+    },
+    {
+      "epoch": 1.4975233902036322,
+      "grad_norm": 0.5184688568115234,
+      "learning_rate": 0.0001177295072443438,
+      "loss": 0.1045,
+      "step": 681
+    },
+    {
+      "epoch": 1.499724821133737,
+      "grad_norm": 0.3599972426891327,
+      "learning_rate": 0.0001174770555256672,
+      "loss": 0.0485,
+      "step": 682
+    },
+    {
+      "epoch": 1.5019262520638414,
+      "grad_norm": 0.7699552774429321,
+      "learning_rate": 0.0001172244888605319,
+      "loss": 0.1161,
+      "step": 683
+    },
+    {
+      "epoch": 1.504127682993946,
+      "grad_norm": 1.2350573539733887,
+      "learning_rate": 0.00011697180891006689,
+      "loss": 0.1615,
+      "step": 684
+    },
+    {
+      "epoch": 1.5063291139240507,
+      "grad_norm": 1.0100566148757935,
+      "learning_rate": 0.00011671901733614628,
+      "loss": 0.1865,
+      "step": 685
+    },
+    {
+      "epoch": 1.508530544854155,
+      "grad_norm": 0.3905757963657379,
+      "learning_rate": 0.00011646611580137823,
+      "loss": 0.0464,
+      "step": 686
+    },
+    {
+      "epoch": 1.5107319757842599,
+      "grad_norm": 1.1495448350906372,
+      "learning_rate": 0.00011621310596909422,
+      "loss": 0.1326,
+      "step": 687
+    },
+    {
+      "epoch": 1.5129334067143643,
+      "grad_norm": 1.3454679250717163,
+      "learning_rate": 0.00011595998950333793,
+      "loss": 0.1748,
+      "step": 688
+    },
+    {
+      "epoch": 1.5151348376444689,
+      "grad_norm": 0.8291569352149963,
+      "learning_rate": 0.00011570676806885441,
+      "loss": 0.178,
+      "step": 689
+    },
+    {
+      "epoch": 1.5173362685745735,
+      "grad_norm": 1.183701515197754,
+      "learning_rate": 0.00011545344333107904,
+      "loss": 0.1655,
+      "step": 690
+    },
+    {
+      "epoch": 1.5195376995046779,
+      "grad_norm": 0.9870046377182007,
+      "learning_rate": 0.00011520001695612674,
+      "loss": 0.1865,
+      "step": 691
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 0.6644650101661682,
+      "learning_rate": 0.00011494649061078071,
+      "loss": 0.115,
+      "step": 692
+    },
+    {
+      "epoch": 1.523940561364887,
+      "grad_norm": 1.217482566833496,
+      "learning_rate": 0.00011469286596248181,
+      "loss": 0.2605,
+      "step": 693
+    },
+    {
+      "epoch": 1.5261419922949917,
+      "grad_norm": 0.9333559274673462,
+      "learning_rate": 0.00011443914467931736,
+      "loss": 0.1618,
+      "step": 694
+    },
+    {
+      "epoch": 1.5283434232250963,
+      "grad_norm": 0.839487612247467,
+      "learning_rate": 0.00011418532843001031,
+      "loss": 0.0942,
+      "step": 695
+    },
+    {
+      "epoch": 1.530544854155201,
+      "grad_norm": 1.1805278062820435,
+      "learning_rate": 0.00011393141888390814,
+      "loss": 0.1592,
+      "step": 696
+    },
+    {
+      "epoch": 1.5327462850853055,
+      "grad_norm": 0.6482624411582947,
+      "learning_rate": 0.00011367741771097197,
+      "loss": 0.0706,
+      "step": 697
+    },
+    {
+      "epoch": 1.53494771601541,
+      "grad_norm": 0.8422489762306213,
+      "learning_rate": 0.00011342332658176555,
+      "loss": 0.125,
+      "step": 698
+    },
+    {
+      "epoch": 1.5371491469455147,
+      "grad_norm": 0.38020074367523193,
+      "learning_rate": 0.00011316914716744426,
+      "loss": 0.0532,
+      "step": 699
+    },
+    {
+      "epoch": 1.5393505778756191,
+      "grad_norm": 0.7216457724571228,
+      "learning_rate": 0.00011291488113974416,
+      "loss": 0.1769,
+      "step": 700
+    },
+    {
+      "epoch": 1.5415520088057237,
+      "grad_norm": 0.5067474246025085,
+      "learning_rate": 0.00011266053017097091,
+      "loss": 0.0982,
+      "step": 701
+    },
+    {
+      "epoch": 1.5437534397358283,
+      "grad_norm": 0.7698938250541687,
+      "learning_rate": 0.00011240609593398886,
+      "loss": 0.097,
+      "step": 702
+    },
+    {
+      "epoch": 1.5459548706659327,
+      "grad_norm": 0.8412646055221558,
+      "learning_rate": 0.00011215158010221005,
+      "loss": 0.1075,
+      "step": 703
+    },
+    {
+      "epoch": 1.5481563015960376,
+      "grad_norm": 0.841958224773407,
+      "learning_rate": 0.00011189698434958309,
+      "loss": 0.1534,
+      "step": 704
+    },
+    {
+      "epoch": 1.550357732526142,
+      "grad_norm": 0.8684465289115906,
+      "learning_rate": 0.00011164231035058228,
+      "loss": 0.1269,
+      "step": 705
+    },
+    {
+      "epoch": 1.5525591634562466,
+      "grad_norm": 1.397005558013916,
+      "learning_rate": 0.00011138755978019657,
+      "loss": 0.2653,
+      "step": 706
+    },
+    {
+      "epoch": 1.5547605943863512,
+      "grad_norm": 0.6085904836654663,
+      "learning_rate": 0.00011113273431391849,
+      "loss": 0.0796,
+      "step": 707
+    },
+    {
+      "epoch": 1.5569620253164556,
+      "grad_norm": 0.6033327579498291,
+      "learning_rate": 0.00011087783562773311,
+      "loss": 0.1289,
+      "step": 708
+    },
+    {
+      "epoch": 1.5591634562465604,
+      "grad_norm": 0.9420616030693054,
+      "learning_rate": 0.00011062286539810711,
+      "loss": 0.2401,
+      "step": 709
+    },
+    {
+      "epoch": 1.5613648871766648,
+      "grad_norm": 0.825903058052063,
+      "learning_rate": 0.00011036782530197775,
+      "loss": 0.1729,
+      "step": 710
+    },
+    {
+      "epoch": 1.5635663181067694,
+      "grad_norm": 0.7171649932861328,
+      "learning_rate": 0.00011011271701674178,
+      "loss": 0.1321,
+      "step": 711
+    },
+    {
+      "epoch": 1.565767749036874,
+      "grad_norm": 1.4322612285614014,
+      "learning_rate": 0.00010985754222024436,
+      "loss": 0.1478,
+      "step": 712
+    },
+    {
+      "epoch": 1.5679691799669784,
+      "grad_norm": 0.6679661870002747,
+      "learning_rate": 0.00010960230259076818,
+      "loss": 0.1528,
+      "step": 713
+    },
+    {
+      "epoch": 1.5701706108970832,
+      "grad_norm": 1.174611210823059,
+      "learning_rate": 0.00010934699980702232,
+      "loss": 0.118,
+      "step": 714
+    },
+    {
+      "epoch": 1.5723720418271876,
+      "grad_norm": 0.5112281441688538,
+      "learning_rate": 0.00010909163554813117,
+      "loss": 0.1038,
+      "step": 715
+    },
+    {
+      "epoch": 1.5745734727572922,
+      "grad_norm": 1.4544810056686401,
+      "learning_rate": 0.0001088362114936235,
+      "loss": 0.2417,
+      "step": 716
+    },
+    {
+      "epoch": 1.5767749036873968,
+      "grad_norm": 1.2852834463119507,
+      "learning_rate": 0.00010858072932342134,
+      "loss": 0.2468,
+      "step": 717
+    },
+    {
+      "epoch": 1.5789763346175014,
+      "grad_norm": 0.4749760031700134,
+      "learning_rate": 0.00010832519071782894,
+      "loss": 0.0733,
+      "step": 718
+    },
+    {
+      "epoch": 1.581177765547606,
+      "grad_norm": 1.2053903341293335,
+      "learning_rate": 0.00010806959735752174,
+      "loss": 0.1455,
+      "step": 719
+    },
+    {
+      "epoch": 1.5833791964777104,
+      "grad_norm": 1.0660688877105713,
+      "learning_rate": 0.00010781395092353526,
+      "loss": 0.1825,
+      "step": 720
+    },
+    {
+      "epoch": 1.5855806274078152,
+      "grad_norm": 1.3690940141677856,
+      "learning_rate": 0.00010755825309725416,
+      "loss": 0.3656,
+      "step": 721
+    },
+    {
+      "epoch": 1.5877820583379196,
+      "grad_norm": 0.5797556042671204,
+      "learning_rate": 0.000107302505560401,
+      "loss": 0.0831,
+      "step": 722
+    },
+    {
+      "epoch": 1.5899834892680242,
+      "grad_norm": 0.5217254161834717,
+      "learning_rate": 0.0001070467099950254,
+      "loss": 0.0831,
+      "step": 723
+    },
+    {
+      "epoch": 1.5921849201981289,
+      "grad_norm": 0.5146464705467224,
+      "learning_rate": 0.00010679086808349278,
+      "loss": 0.0561,
+      "step": 724
+    },
+    {
+      "epoch": 1.5943863511282332,
+      "grad_norm": 0.48769646883010864,
+      "learning_rate": 0.00010653498150847343,
+      "loss": 0.0819,
+      "step": 725
+    },
+    {
+      "epoch": 1.596587782058338,
+      "grad_norm": 1.0709245204925537,
+      "learning_rate": 0.00010627905195293135,
+      "loss": 0.1851,
+      "step": 726
+    },
+    {
+      "epoch": 1.5987892129884425,
+      "grad_norm": 0.5213449001312256,
+      "learning_rate": 0.00010602308110011328,
+      "loss": 0.0837,
+      "step": 727
+    },
+    {
+      "epoch": 1.600990643918547,
+      "grad_norm": 0.7490752339363098,
+      "learning_rate": 0.00010576707063353746,
+      "loss": 0.1098,
+      "step": 728
+    },
+    {
+      "epoch": 1.6031920748486517,
+      "grad_norm": 0.7549910545349121,
+      "learning_rate": 0.00010551102223698282,
+      "loss": 0.1396,
+      "step": 729
+    },
+    {
+      "epoch": 1.605393505778756,
+      "grad_norm": 1.0106641054153442,
+      "learning_rate": 0.00010525493759447765,
+      "loss": 0.1466,
+      "step": 730
+    },
+    {
+      "epoch": 1.6075949367088609,
+      "grad_norm": 0.7263780832290649,
+      "learning_rate": 0.00010499881839028867,
+      "loss": 0.1372,
+      "step": 731
+    },
+    {
+      "epoch": 1.6097963676389653,
+      "grad_norm": 0.6660489439964294,
+      "learning_rate": 0.00010474266630890986,
+      "loss": 0.0806,
+      "step": 732
+    },
+    {
+      "epoch": 1.6119977985690699,
+      "grad_norm": 1.0522032976150513,
+      "learning_rate": 0.00010448648303505151,
+      "loss": 0.2161,
+      "step": 733
+    },
+    {
+      "epoch": 1.6141992294991745,
+      "grad_norm": 0.9830417633056641,
+      "learning_rate": 0.00010423027025362902,
+      "loss": 0.2289,
+      "step": 734
+    },
+    {
+      "epoch": 1.6164006604292789,
+      "grad_norm": 1.082297682762146,
+      "learning_rate": 0.00010397402964975187,
+      "loss": 0.1082,
+      "step": 735
+    },
+    {
+      "epoch": 1.6186020913593837,
+      "grad_norm": 1.1494207382202148,
+      "learning_rate": 0.00010371776290871252,
+      "loss": 0.2047,
+      "step": 736
+    },
+    {
+      "epoch": 1.620803522289488,
+      "grad_norm": 1.351162314414978,
+      "learning_rate": 0.00010346147171597535,
+      "loss": 0.2367,
+      "step": 737
+    },
+    {
+      "epoch": 1.6230049532195927,
+      "grad_norm": 0.8432272672653198,
+      "learning_rate": 0.00010320515775716555,
+      "loss": 0.2138,
+      "step": 738
+    },
+    {
+      "epoch": 1.6252063841496973,
+      "grad_norm": 0.9856284856796265,
+      "learning_rate": 0.00010294882271805799,
+      "loss": 0.2512,
+      "step": 739
+    },
+    {
+      "epoch": 1.627407815079802,
+      "grad_norm": 1.0501776933670044,
+      "learning_rate": 0.00010269246828456629,
+      "loss": 0.1582,
+      "step": 740
+    },
+    {
+      "epoch": 1.6296092460099065,
+      "grad_norm": 0.6857620477676392,
+      "learning_rate": 0.00010243609614273155,
+      "loss": 0.0702,
+      "step": 741
+    },
+    {
+      "epoch": 1.631810676940011,
+      "grad_norm": 0.8443143367767334,
+      "learning_rate": 0.0001021797079787114,
+      "loss": 0.1715,
+      "step": 742
+    },
+    {
+      "epoch": 1.6340121078701155,
+      "grad_norm": 0.7217234373092651,
+      "learning_rate": 0.00010192330547876871,
+      "loss": 0.1181,
+      "step": 743
+    },
+    {
+      "epoch": 1.6362135388002201,
+      "grad_norm": 1.8371859788894653,
+      "learning_rate": 0.00010166689032926085,
+      "loss": 0.3051,
+      "step": 744
+    },
+    {
+      "epoch": 1.6384149697303247,
+      "grad_norm": 0.14680980145931244,
+      "learning_rate": 0.00010141046421662821,
+      "loss": 0.0258,
+      "step": 745
+    },
+    {
+      "epoch": 1.6406164006604294,
+      "grad_norm": 0.5036325454711914,
+      "learning_rate": 0.00010115402882738334,
+      "loss": 0.0987,
+      "step": 746
+    },
+    {
+      "epoch": 1.6428178315905337,
+      "grad_norm": 0.7323853969573975,
+      "learning_rate": 0.00010089758584809979,
+      "loss": 0.11,
+      "step": 747
+    },
+    {
+      "epoch": 1.6450192625206386,
+      "grad_norm": 0.5266981720924377,
+      "learning_rate": 0.00010064113696540111,
+      "loss": 0.1494,
+      "step": 748
+    },
+    {
+      "epoch": 1.647220693450743,
+      "grad_norm": 0.30830544233322144,
+      "learning_rate": 0.0001003846838659496,
+      "loss": 0.0412,
+      "step": 749
+    },
+    {
+      "epoch": 1.6494221243808476,
+      "grad_norm": 0.9504794478416443,
+      "learning_rate": 0.00010012822823643525,
+      "loss": 0.1227,
+      "step": 750
+    },
+    {
+      "epoch": 1.6516235553109522,
+      "grad_norm": 1.2924413681030273,
+      "learning_rate": 9.987177176356476e-05,
+      "loss": 0.187,
+      "step": 751
+    },
+    {
+      "epoch": 1.6538249862410566,
+      "grad_norm": 0.5978471040725708,
+      "learning_rate": 9.961531613405043e-05,
+      "loss": 0.1778,
+      "step": 752
+    },
+    {
+      "epoch": 1.6560264171711614,
+      "grad_norm": 1.2347365617752075,
+      "learning_rate": 9.93588630345989e-05,
+      "loss": 0.1905,
+      "step": 753
+    },
+    {
+      "epoch": 1.6582278481012658,
+      "grad_norm": 0.7218102812767029,
+      "learning_rate": 9.910241415190021e-05,
+      "loss": 0.1779,
+      "step": 754
+    },
+    {
+      "epoch": 1.6604292790313704,
+      "grad_norm": 0.8096426725387573,
+      "learning_rate": 9.884597117261667e-05,
+      "loss": 0.099,
+      "step": 755
+    },
+    {
+      "epoch": 1.662630709961475,
+      "grad_norm": 1.0941493511199951,
+      "learning_rate": 9.85895357833718e-05,
+      "loss": 0.095,
+      "step": 756
+    },
+    {
+      "epoch": 1.6648321408915794,
+      "grad_norm": 0.661952018737793,
+      "learning_rate": 9.83331096707392e-05,
+      "loss": 0.1123,
+      "step": 757
+    },
+    {
+      "epoch": 1.6670335718216842,
+      "grad_norm": 1.0296493768692017,
+      "learning_rate": 9.80766945212313e-05,
+      "loss": 0.1239,
+      "step": 758
+    },
+    {
+      "epoch": 1.6692350027517886,
+      "grad_norm": 0.7456392049789429,
+      "learning_rate": 9.782029202128864e-05,
+      "loss": 0.1132,
+      "step": 759
+    },
+    {
+      "epoch": 1.6714364336818932,
+      "grad_norm": 0.9941757917404175,
+      "learning_rate": 9.756390385726846e-05,
+      "loss": 0.1752,
+      "step": 760
+    },
+    {
+      "epoch": 1.6736378646119978,
+      "grad_norm": 0.8386315107345581,
+      "learning_rate": 9.730753171543374e-05,
+      "loss": 0.127,
+      "step": 761
+    },
+    {
+      "epoch": 1.6758392955421022,
+      "grad_norm": 0.4235280752182007,
+      "learning_rate": 9.705117728194202e-05,
+      "loss": 0.0495,
+      "step": 762
+    },
+    {
+      "epoch": 1.678040726472207,
+      "grad_norm": 0.9623192548751831,
+      "learning_rate": 9.679484224283449e-05,
+      "loss": 0.1748,
+      "step": 763
+    },
+    {
+      "epoch": 1.6802421574023114,
+      "grad_norm": 0.5407153964042664,
+      "learning_rate": 9.653852828402467e-05,
+      "loss": 0.0788,
+      "step": 764
+    },
+    {
+      "epoch": 1.682443588332416,
+      "grad_norm": 0.9053894281387329,
+      "learning_rate": 9.628223709128749e-05,
+      "loss": 0.1913,
+      "step": 765
+    },
+    {
+      "epoch": 1.6846450192625206,
+      "grad_norm": 0.5698592066764832,
+      "learning_rate": 9.602597035024815e-05,
+      "loss": 0.0605,
+      "step": 766
+    },
+    {
+      "epoch": 1.6868464501926252,
+      "grad_norm": 0.9939035773277283,
+      "learning_rate": 9.576972974637097e-05,
+      "loss": 0.161,
+      "step": 767
+    },
+    {
+      "epoch": 1.6890478811227299,
+      "grad_norm": 0.7236847281455994,
+      "learning_rate": 9.551351696494854e-05,
+      "loss": 0.0859,
+      "step": 768
+    },
+    {
+      "epoch": 1.6912493120528342,
+      "grad_norm": 1.412492275238037,
+      "learning_rate": 9.525733369109018e-05,
+      "loss": 0.1902,
+      "step": 769
+    },
+    {
+      "epoch": 1.693450742982939,
+      "grad_norm": 0.956356406211853,
+      "learning_rate": 9.500118160971138e-05,
+      "loss": 0.1631,
+      "step": 770
+    },
+    {
+      "epoch": 1.6956521739130435,
+      "grad_norm": 1.869669795036316,
+      "learning_rate": 9.474506240552239e-05,
+      "loss": 0.219,
+      "step": 771
+    },
+    {
+      "epoch": 1.697853604843148,
+      "grad_norm": 1.6506068706512451,
+      "learning_rate": 9.448897776301722e-05,
+      "loss": 0.1503,
+      "step": 772
+    },
+    {
+      "epoch": 1.7000550357732527,
+      "grad_norm": 0.5310888290405273,
+      "learning_rate": 9.423292936646257e-05,
+      "loss": 0.093,
+      "step": 773
+    },
+    {
+      "epoch": 1.702256466703357,
+      "grad_norm": 1.497267723083496,
+      "learning_rate": 9.397691889988676e-05,
+      "loss": 0.2146,
+      "step": 774
+    },
+    {
+      "epoch": 1.704457897633462,
+      "grad_norm": 1.5522136688232422,
+      "learning_rate": 9.372094804706867e-05,
+      "loss": 0.2829,
+      "step": 775
+    },
+    {
+      "epoch": 1.7066593285635663,
+      "grad_norm": 0.5641590356826782,
+      "learning_rate": 9.346501849152659e-05,
+      "loss": 0.0788,
+      "step": 776
+    },
+    {
+      "epoch": 1.7088607594936709,
+      "grad_norm": 0.18695193529129028,
+      "learning_rate": 9.320913191650724e-05,
+      "loss": 0.0299,
+      "step": 777
+    },
+    {
+      "epoch": 1.7110621904237755,
+      "grad_norm": 0.8758842349052429,
+      "learning_rate": 9.29532900049746e-05,
+      "loss": 0.122,
+      "step": 778
+    },
+    {
+      "epoch": 1.7132636213538799,
+      "grad_norm": 0.8107666969299316,
+      "learning_rate": 9.269749443959904e-05,
+      "loss": 0.0945,
+      "step": 779
+    },
+    {
+      "epoch": 1.7154650522839847,
+      "grad_norm": 0.8614788055419922,
+      "learning_rate": 9.24417469027459e-05,
+      "loss": 0.1672,
+      "step": 780
+    },
+    {
+      "epoch": 1.717666483214089,
+      "grad_norm": 1.3785629272460938,
+      "learning_rate": 9.218604907646475e-05,
+      "loss": 0.3274,
+      "step": 781
+    },
+    {
+      "epoch": 1.7198679141441937,
+      "grad_norm": 2.5916779041290283,
+      "learning_rate": 9.193040264247829e-05,
+      "loss": 0.2438,
+      "step": 782
+    },
+    {
+      "epoch": 1.7220693450742983,
+      "grad_norm": 0.8327628374099731,
+      "learning_rate": 9.167480928217108e-05,
+      "loss": 0.1232,
+      "step": 783
+    },
+    {
+      "epoch": 1.7242707760044027,
+      "grad_norm": 0.536798357963562,
+      "learning_rate": 9.141927067657869e-05,
+      "loss": 0.0611,
+      "step": 784
+    },
+    {
+      "epoch": 1.7264722069345075,
+      "grad_norm": 1.238837480545044,
+      "learning_rate": 9.11637885063765e-05,
+      "loss": 0.1771,
+      "step": 785
+    },
+    {
+      "epoch": 1.728673637864612,
+      "grad_norm": 0.6637739539146423,
+      "learning_rate": 9.090836445186885e-05,
+      "loss": 0.1656,
+      "step": 786
+    },
+    {
+      "epoch": 1.7308750687947165,
+      "grad_norm": 0.6841593980789185,
+      "learning_rate": 9.06530001929777e-05,
+      "loss": 0.093,
+      "step": 787
+    },
+    {
+      "epoch": 1.7330764997248211,
+      "grad_norm": 1.7791855335235596,
+      "learning_rate": 9.039769740923183e-05,
+      "loss": 0.251,
+      "step": 788
+    },
+    {
+      "epoch": 1.7352779306549257,
+      "grad_norm": 1.0243498086929321,
+      "learning_rate": 9.014245777975565e-05,
+      "loss": 0.1087,
+      "step": 789
+    },
+    {
+      "epoch": 1.7374793615850304,
+      "grad_norm": 0.9615514278411865,
+      "learning_rate": 8.988728298325822e-05,
+      "loss": 0.1978,
+      "step": 790
+    },
+    {
+      "epoch": 1.7396807925151347,
+      "grad_norm": 0.4960983693599701,
+      "learning_rate": 8.963217469802228e-05,
+      "loss": 0.0553,
+      "step": 791
+    },
+    {
+      "epoch": 1.7418822234452394,
+      "grad_norm": 0.8816950917243958,
+      "learning_rate": 8.937713460189291e-05,
+      "loss": 0.1071,
+      "step": 792
+    },
+    {
+      "epoch": 1.744083654375344,
+      "grad_norm": 0.9826503396034241,
+      "learning_rate": 8.912216437226693e-05,
+      "loss": 0.1488,
+      "step": 793
+    },
+    {
+      "epoch": 1.7462850853054486,
+      "grad_norm": 0.5744142532348633,
+      "learning_rate": 8.886726568608155e-05,
+      "loss": 0.0908,
+      "step": 794
+    },
+    {
+      "epoch": 1.7484865162355532,
+      "grad_norm": 0.5662325620651245,
+      "learning_rate": 8.861244021980344e-05,
+      "loss": 0.0897,
+      "step": 795
+    },
+    {
+      "epoch": 1.7506879471656576,
+      "grad_norm": 0.39198780059814453,
+      "learning_rate": 8.835768964941773e-05,
+      "loss": 0.0491,
+      "step": 796
+    },
+    {
+      "epoch": 1.7528893780957624,
+      "grad_norm": 0.7171629667282104,
+      "learning_rate": 8.810301565041692e-05,
+      "loss": 0.1096,
+      "step": 797
+    },
+    {
+      "epoch": 1.7550908090258668,
+      "grad_norm": 0.8525797724723816,
+      "learning_rate": 8.784841989778996e-05,
+      "loss": 0.0978,
+      "step": 798
+    },
+    {
+      "epoch": 1.7572922399559714,
+      "grad_norm": 0.7842770218849182,
+      "learning_rate": 8.759390406601115e-05,
+      "loss": 0.143,
+      "step": 799
+    },
+    {
+      "epoch": 1.759493670886076,
+      "grad_norm": 1.2253752946853638,
+      "learning_rate": 8.733946982902911e-05,
+      "loss": 0.148,
+      "step": 800
+    },
+    {
+      "epoch": 1.7616951018161804,
+      "grad_norm": 0.7642062902450562,
+      "learning_rate": 8.708511886025585e-05,
+      "loss": 0.1092,
+      "step": 801
+    },
+    {
+      "epoch": 1.7638965327462852,
+      "grad_norm": 0.8164166212081909,
+      "learning_rate": 8.683085283255576e-05,
+      "loss": 0.1523,
+      "step": 802
+    },
+    {
+      "epoch": 1.7660979636763896,
+      "grad_norm": 0.9109005928039551,
+      "learning_rate": 8.657667341823448e-05,
+      "loss": 0.1462,
+      "step": 803
+    },
+    {
+      "epoch": 1.7682993946064942,
+      "grad_norm": 1.1961185932159424,
+      "learning_rate": 8.632258228902805e-05,
+      "loss": 0.1098,
+      "step": 804
+    },
+    {
+      "epoch": 1.7705008255365988,
+      "grad_norm": 0.6288356184959412,
+      "learning_rate": 8.606858111609188e-05,
+      "loss": 0.126,
+      "step": 805
+    },
+    {
+      "epoch": 1.7727022564667032,
+      "grad_norm": 0.6300954222679138,
+      "learning_rate": 8.581467156998971e-05,
+      "loss": 0.1364,
+      "step": 806
+    },
+    {
+      "epoch": 1.774903687396808,
+      "grad_norm": 0.8954706788063049,
+      "learning_rate": 8.556085532068266e-05,
+      "loss": 0.0989,
+      "step": 807
+    },
+    {
+      "epoch": 1.7771051183269124,
+      "grad_norm": 1.5506774187088013,
+      "learning_rate": 8.530713403751821e-05,
+      "loss": 0.1683,
+      "step": 808
+    },
+    {
+      "epoch": 1.779306549257017,
+      "grad_norm": 0.7719908952713013,
+      "learning_rate": 8.505350938921931e-05,
+      "loss": 0.0985,
+      "step": 809
+    },
+    {
+      "epoch": 1.7815079801871216,
+      "grad_norm": 0.22699938714504242,
+      "learning_rate": 8.479998304387329e-05,
+      "loss": 0.0205,
+      "step": 810
+    },
+    {
+      "epoch": 1.7837094111172263,
+      "grad_norm": 0.6535486578941345,
+      "learning_rate": 8.454655666892095e-05,
+      "loss": 0.0905,
+      "step": 811
+    },
+    {
+      "epoch": 1.7859108420473309,
+      "grad_norm": 0.47723376750946045,
+      "learning_rate": 8.42932319311456e-05,
+      "loss": 0.1109,
+      "step": 812
+    },
+    {
+      "epoch": 1.7881122729774352,
+      "grad_norm": 0.8404837846755981,
+      "learning_rate": 8.404001049666211e-05,
+      "loss": 0.1419,
+      "step": 813
+    },
+    {
+      "epoch": 1.7903137039075399,
+      "grad_norm": 0.5765104293823242,
+      "learning_rate": 8.378689403090583e-05,
+      "loss": 0.0634,
+      "step": 814
+    },
+    {
+      "epoch": 1.7925151348376445,
+      "grad_norm": 0.8536213636398315,
+      "learning_rate": 8.353388419862178e-05,
+      "loss": 0.1279,
+      "step": 815
+    },
+    {
+      "epoch": 1.794716565767749,
+      "grad_norm": 0.7238773107528687,
+      "learning_rate": 8.328098266385375e-05,
+      "loss": 0.093,
+      "step": 816
+    },
+    {
+      "epoch": 1.7969179966978537,
+      "grad_norm": 1.3495720624923706,
+      "learning_rate": 8.302819108993312e-05,
+      "loss": 0.2,
+      "step": 817
+    },
+    {
+      "epoch": 1.799119427627958,
+      "grad_norm": 0.6155779957771301,
+      "learning_rate": 8.277551113946812e-05,
+      "loss": 0.1371,
+      "step": 818
+    },
+    {
+      "epoch": 1.801320858558063,
+      "grad_norm": 0.8336213231086731,
+      "learning_rate": 8.252294447433283e-05,
+      "loss": 0.1992,
+      "step": 819
+    },
+    {
+      "epoch": 1.8035222894881673,
+      "grad_norm": 0.394396036863327,
+      "learning_rate": 8.227049275565622e-05,
+      "loss": 0.0589,
+      "step": 820
+    },
+    {
+      "epoch": 1.805723720418272,
+      "grad_norm": 0.5639936923980713,
+      "learning_rate": 8.201815764381133e-05,
+      "loss": 0.1162,
+      "step": 821
+    },
+    {
+      "epoch": 1.8079251513483765,
+      "grad_norm": 1.2527427673339844,
+      "learning_rate": 8.176594079840423e-05,
+      "loss": 0.1426,
+      "step": 822
+    },
+    {
+      "epoch": 1.810126582278481,
+      "grad_norm": 0.8214813470840454,
+      "learning_rate": 8.151384387826313e-05,
+      "loss": 0.132,
+      "step": 823
+    },
+    {
+      "epoch": 1.8123280132085857,
+      "grad_norm": 0.6703236699104309,
+      "learning_rate": 8.126186854142752e-05,
+      "loss": 0.1678,
+      "step": 824
+    },
+    {
+      "epoch": 1.81452944413869,
+      "grad_norm": 0.5412653088569641,
+      "learning_rate": 8.101001644513731e-05,
+      "loss": 0.113,
+      "step": 825
+    },
+    {
+      "epoch": 1.8167308750687947,
+      "grad_norm": 0.8707739114761353,
+      "learning_rate": 8.075828924582169e-05,
+      "loss": 0.1577,
+      "step": 826
+    },
+    {
+      "epoch": 1.8189323059988993,
+      "grad_norm": 1.103877067565918,
+      "learning_rate": 8.05066885990885e-05,
+      "loss": 0.1617,
+      "step": 827
+    },
+    {
+      "epoch": 1.8211337369290037,
+      "grad_norm": 1.0585663318634033,
+      "learning_rate": 8.02552161597133e-05,
+      "loss": 0.2364,
+      "step": 828
+    },
+    {
+      "epoch": 1.8233351678591085,
+      "grad_norm": 0.46425899863243103,
+      "learning_rate": 8.000387358162835e-05,
+      "loss": 0.1124,
+      "step": 829
+    },
+    {
+      "epoch": 1.825536598789213,
+      "grad_norm": 0.5401412844657898,
+      "learning_rate": 7.975266251791185e-05,
+      "loss": 0.1355,
+      "step": 830
+    },
+    {
+      "epoch": 1.8277380297193175,
+      "grad_norm": 0.3869989812374115,
+      "learning_rate": 7.950158462077697e-05,
+      "loss": 0.0837,
+      "step": 831
+    },
+    {
+      "epoch": 1.8299394606494221,
+      "grad_norm": 0.4511391818523407,
+      "learning_rate": 7.925064154156116e-05,
+      "loss": 0.1075,
+      "step": 832
+    },
+    {
+      "epoch": 1.8321408915795265,
+      "grad_norm": 0.960554301738739,
+      "learning_rate": 7.899983493071507e-05,
+      "loss": 0.1864,
+      "step": 833
+    },
+    {
+      "epoch": 1.8343423225096314,
+      "grad_norm": 1.1453462839126587,
+      "learning_rate": 7.874916643779184e-05,
+      "loss": 0.1911,
+      "step": 834
+    },
+    {
+      "epoch": 1.8365437534397357,
+      "grad_norm": 1.1819422245025635,
+      "learning_rate": 7.849863771143621e-05,
+      "loss": 0.2681,
+      "step": 835
+    },
+    {
+      "epoch": 1.8387451843698404,
+      "grad_norm": 0.7269442677497864,
+      "learning_rate": 7.82482503993737e-05,
+      "loss": 0.0988,
+      "step": 836
+    },
+    {
+      "epoch": 1.840946615299945,
+      "grad_norm": 0.9441477060317993,
+      "learning_rate": 7.799800614839964e-05,
+      "loss": 0.2001,
+      "step": 837
+    },
+    {
+      "epoch": 1.8431480462300496,
+      "grad_norm": 0.8774495124816895,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 0.1978,
+      "step": 838
+    },
+    {
+      "epoch": 1.8453494771601542,
+      "grad_norm": 0.4876162111759186,
+      "learning_rate": 7.749795341218327e-05,
+      "loss": 0.0747,
+      "step": 839
+    },
+    {
+      "epoch": 1.8475509080902586,
+      "grad_norm": 0.48862627148628235,
+      "learning_rate": 7.724814821578396e-05,
+      "loss": 0.0701,
+      "step": 840
+    },
+    {
+      "epoch": 1.8497523390203634,
+      "grad_norm": 1.3835673332214355,
+      "learning_rate": 7.699849265813743e-05,
+      "loss": 0.162,
+      "step": 841
+    },
+    {
+      "epoch": 1.8519537699504678,
+      "grad_norm": 0.6900964975357056,
+      "learning_rate": 7.67489883812264e-05,
+      "loss": 0.0586,
+      "step": 842
+    },
+    {
+      "epoch": 1.8541552008805724,
+      "grad_norm": 0.7376418113708496,
+      "learning_rate": 7.649963702603849e-05,
+      "loss": 0.079,
+      "step": 843
+    },
+    {
+      "epoch": 1.856356631810677,
+      "grad_norm": 0.6786530613899231,
+      "learning_rate": 7.625044023255571e-05,
+      "loss": 0.1063,
+      "step": 844
+    },
+    {
+      "epoch": 1.8585580627407814,
+      "grad_norm": 0.7012770771980286,
+      "learning_rate": 7.600139963974341e-05,
+      "loss": 0.1238,
+      "step": 845
+    },
+    {
+      "epoch": 1.8607594936708862,
+      "grad_norm": 0.7398976683616638,
+      "learning_rate": 7.575251688553963e-05,
+      "loss": 0.1393,
+      "step": 846
+    },
+    {
+      "epoch": 1.8629609246009906,
+      "grad_norm": 0.5676134824752808,
+      "learning_rate": 7.550379360684434e-05,
+      "loss": 0.1269,
+      "step": 847
+    },
+    {
+      "epoch": 1.8651623555310952,
+      "grad_norm": 0.70937180519104,
+      "learning_rate": 7.525523143950859e-05,
+      "loss": 0.0519,
+      "step": 848
+    },
+    {
+      "epoch": 1.8673637864611998,
+      "grad_norm": 0.563579261302948,
+      "learning_rate": 7.500683201832382e-05,
+      "loss": 0.105,
+      "step": 849
+    },
+    {
+      "epoch": 1.8695652173913042,
+      "grad_norm": 0.6759403347969055,
+      "learning_rate": 7.47585969770111e-05,
+      "loss": 0.1216,
+      "step": 850
+    },
+    {
+      "epoch": 1.871766648321409,
+      "grad_norm": 0.7409111857414246,
+      "learning_rate": 7.45105279482104e-05,
+      "loss": 0.1031,
+      "step": 851
+    },
+    {
+      "epoch": 1.8739680792515134,
+      "grad_norm": 0.7576317191123962,
+      "learning_rate": 7.426262656346978e-05,
+      "loss": 0.0719,
+      "step": 852
+    },
+    {
+      "epoch": 1.876169510181618,
+      "grad_norm": 0.8437058329582214,
+      "learning_rate": 7.401489445323473e-05,
+      "loss": 0.0878,
+      "step": 853
+    },
+    {
+      "epoch": 1.8783709411117226,
+      "grad_norm": 0.7089110016822815,
+      "learning_rate": 7.37673332468374e-05,
+      "loss": 0.1523,
+      "step": 854
+    },
+    {
+      "epoch": 1.880572372041827,
+      "grad_norm": 0.9702976942062378,
+      "learning_rate": 7.351994457248595e-05,
+      "loss": 0.0713,
+      "step": 855
+    },
+    {
+      "epoch": 1.8827738029719319,
+      "grad_norm": 1.018698811531067,
+      "learning_rate": 7.327273005725379e-05,
+      "loss": 0.1516,
+      "step": 856
+    },
+    {
+      "epoch": 1.8849752339020363,
+      "grad_norm": 0.47892457246780396,
+      "learning_rate": 7.302569132706882e-05,
+      "loss": 0.109,
+      "step": 857
+    },
+    {
+      "epoch": 1.8871766648321409,
+      "grad_norm": 0.732261598110199,
+      "learning_rate": 7.27788300067029e-05,
+      "loss": 0.092,
+      "step": 858
+    },
+    {
+      "epoch": 1.8893780957622455,
+      "grad_norm": 1.109673261642456,
+      "learning_rate": 7.2532147719761e-05,
+      "loss": 0.1103,
+      "step": 859
+    },
+    {
+      "epoch": 1.89157952669235,
+      "grad_norm": 0.5722037553787231,
+      "learning_rate": 7.22856460886706e-05,
+      "loss": 0.1094,
+      "step": 860
+    },
+    {
+      "epoch": 1.8937809576224547,
+      "grad_norm": 0.7374326586723328,
+      "learning_rate": 7.203932673467101e-05,
+      "loss": 0.109,
+      "step": 861
+    },
+    {
+      "epoch": 1.895982388552559,
+      "grad_norm": 1.0216845273971558,
+      "learning_rate": 7.179319127780274e-05,
+      "loss": 0.2088,
+      "step": 862
+    },
+    {
+      "epoch": 1.8981838194826637,
+      "grad_norm": 0.8911333084106445,
+      "learning_rate": 7.154724133689677e-05,
+      "loss": 0.126,
+      "step": 863
+    },
+    {
+      "epoch": 1.9003852504127683,
+      "grad_norm": 0.6132821440696716,
+      "learning_rate": 7.130147852956394e-05,
+      "loss": 0.0664,
+      "step": 864
+    },
+    {
+      "epoch": 1.902586681342873,
+      "grad_norm": 0.6062414646148682,
+      "learning_rate": 7.105590447218438e-05,
+      "loss": 0.1096,
+      "step": 865
+    },
+    {
+      "epoch": 1.9047881122729775,
+      "grad_norm": 0.7949546575546265,
+      "learning_rate": 7.081052077989667e-05,
+      "loss": 0.0898,
+      "step": 866
+    },
+    {
+      "epoch": 1.906989543203082,
+      "grad_norm": 0.546043872833252,
+      "learning_rate": 7.056532906658752e-05,
+      "loss": 0.0606,
+      "step": 867
+    },
+    {
+      "epoch": 1.9091909741331867,
+      "grad_norm": 0.5800933837890625,
+      "learning_rate": 7.032033094488095e-05,
+      "loss": 0.1102,
+      "step": 868
+    },
+    {
+      "epoch": 1.9113924050632911,
+      "grad_norm": 0.6095998883247375,
+      "learning_rate": 7.007552802612765e-05,
+      "loss": 0.0896,
+      "step": 869
+    },
+    {
+      "epoch": 1.9135938359933957,
+      "grad_norm": 1.112326741218567,
+      "learning_rate": 6.983092192039456e-05,
+      "loss": 0.1556,
+      "step": 870
+    },
+    {
+      "epoch": 1.9157952669235003,
+      "grad_norm": 0.6165559887886047,
+      "learning_rate": 6.958651423645409e-05,
+      "loss": 0.0835,
+      "step": 871
+    },
+    {
+      "epoch": 1.9179966978536047,
+      "grad_norm": 0.49726906418800354,
+      "learning_rate": 6.934230658177372e-05,
+      "loss": 0.0748,
+      "step": 872
+    },
+    {
+      "epoch": 1.9201981287837095,
+      "grad_norm": 2.08432936668396,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.335,
+      "step": 873
+    },
+    {
+      "epoch": 1.922399559713814,
+      "grad_norm": 0.37368956208229065,
+      "learning_rate": 6.885449778347448e-05,
+      "loss": 0.0427,
+      "step": 874
+    },
+    {
+      "epoch": 1.9246009906439185,
+      "grad_norm": 0.616974949836731,
+      "learning_rate": 6.861089984817032e-05,
+      "loss": 0.1036,
+      "step": 875
+    },
+    {
+      "epoch": 1.9268024215740231,
+      "grad_norm": 0.4907055199146271,
+      "learning_rate": 6.836750835873453e-05,
+      "loss": 0.0526,
+      "step": 876
+    },
+    {
+      "epoch": 1.9290038525041275,
+      "grad_norm": 0.8442423939704895,
+      "learning_rate": 6.812432491595102e-05,
+      "loss": 0.1206,
+      "step": 877
+    },
+    {
+      "epoch": 1.9312052834342324,
+      "grad_norm": 2.4751176834106445,
+      "learning_rate": 6.788135111923545e-05,
+      "loss": 0.3944,
+      "step": 878
+    },
+    {
+      "epoch": 1.9334067143643368,
+      "grad_norm": 0.6310155987739563,
+      "learning_rate": 6.763858856662457e-05,
+      "loss": 0.1503,
+      "step": 879
+    },
+    {
+      "epoch": 1.9356081452944414,
+      "grad_norm": 1.2114028930664062,
+      "learning_rate": 6.739603885476582e-05,
+      "loss": 0.1939,
+      "step": 880
+    },
+    {
+      "epoch": 1.937809576224546,
+      "grad_norm": 1.4767590761184692,
+      "learning_rate": 6.715370357890678e-05,
+      "loss": 0.2673,
+      "step": 881
+    },
+    {
+      "epoch": 1.9400110071546504,
+      "grad_norm": 1.4075496196746826,
+      "learning_rate": 6.691158433288465e-05,
+      "loss": 0.2565,
+      "step": 882
+    },
+    {
+      "epoch": 1.9422124380847552,
+      "grad_norm": 0.49719229340553284,
+      "learning_rate": 6.666968270911584e-05,
+      "loss": 0.0479,
+      "step": 883
+    },
+    {
+      "epoch": 1.9444138690148596,
+      "grad_norm": 0.7342202067375183,
+      "learning_rate": 6.642800029858546e-05,
+      "loss": 0.0923,
+      "step": 884
+    },
+    {
+      "epoch": 1.9466152999449642,
+      "grad_norm": 0.8439943790435791,
+      "learning_rate": 6.618653869083688e-05,
+      "loss": 0.1153,
+      "step": 885
+    },
+    {
+      "epoch": 1.9488167308750688,
+      "grad_norm": 0.944554328918457,
+      "learning_rate": 6.594529947396119e-05,
+      "loss": 0.1236,
+      "step": 886
+    },
+    {
+      "epoch": 1.9510181618051734,
+      "grad_norm": 0.7831180691719055,
+      "learning_rate": 6.570428423458687e-05,
+      "loss": 0.1326,
+      "step": 887
+    },
+    {
+      "epoch": 1.953219592735278,
+      "grad_norm": 0.8560040593147278,
+      "learning_rate": 6.546349455786926e-05,
+      "loss": 0.1364,
+      "step": 888
+    },
+    {
+      "epoch": 1.9554210236653824,
+      "grad_norm": 0.7042924761772156,
+      "learning_rate": 6.522293202748017e-05,
+      "loss": 0.1286,
+      "step": 889
+    },
+    {
+      "epoch": 1.9576224545954872,
+      "grad_norm": 0.7423930764198303,
+      "learning_rate": 6.498259822559758e-05,
+      "loss": 0.0448,
+      "step": 890
+    },
+    {
+      "epoch": 1.9598238855255916,
+      "grad_norm": 0.8293138742446899,
+      "learning_rate": 6.474249473289498e-05,
+      "loss": 0.1723,
+      "step": 891
+    },
+    {
+      "epoch": 1.9620253164556962,
+      "grad_norm": 0.8757152557373047,
+      "learning_rate": 6.450262312853121e-05,
+      "loss": 0.1538,
+      "step": 892
+    },
+    {
+      "epoch": 1.9642267473858008,
+      "grad_norm": 0.902170717716217,
+      "learning_rate": 6.426298499013994e-05,
+      "loss": 0.1142,
+      "step": 893
+    },
+    {
+      "epoch": 1.9664281783159052,
+      "grad_norm": 0.8055010437965393,
+      "learning_rate": 6.402358189381934e-05,
+      "loss": 0.1195,
+      "step": 894
+    },
+    {
+      "epoch": 1.96862960924601,
+      "grad_norm": 0.7822755575180054,
+      "learning_rate": 6.37844154141217e-05,
+      "loss": 0.0841,
+      "step": 895
+    },
+    {
+      "epoch": 1.9708310401761144,
+      "grad_norm": 0.6275139451026917,
+      "learning_rate": 6.354548712404313e-05,
+      "loss": 0.0836,
+      "step": 896
+    },
+    {
+      "epoch": 1.973032471106219,
+      "grad_norm": 0.8963627219200134,
+      "learning_rate": 6.330679859501316e-05,
+      "loss": 0.1696,
+      "step": 897
+    },
+    {
+      "epoch": 1.9752339020363237,
+      "grad_norm": 0.6084915399551392,
+      "learning_rate": 6.306835139688438e-05,
+      "loss": 0.0936,
+      "step": 898
+    },
+    {
+      "epoch": 1.977435332966428,
+      "grad_norm": 0.5180437564849854,
+      "learning_rate": 6.283014709792215e-05,
+      "loss": 0.0877,
+      "step": 899
+    },
+    {
+      "epoch": 1.9796367638965329,
+      "grad_norm": 0.5324386954307556,
+      "learning_rate": 6.259218726479427e-05,
+      "loss": 0.0562,
+      "step": 900
+    },
+    {
+      "epoch": 1.9818381948266373,
+      "grad_norm": 0.6329354047775269,
+      "learning_rate": 6.23544734625608e-05,
+      "loss": 0.1695,
+      "step": 901
+    },
+    {
+      "epoch": 1.9840396257567419,
+      "grad_norm": 0.6291633248329163,
+      "learning_rate": 6.211700725466352e-05,
+      "loss": 0.0946,
+      "step": 902
+    },
+    {
+      "epoch": 1.9862410566868465,
+      "grad_norm": 0.9961429834365845,
+      "learning_rate": 6.187979020291583e-05,
+      "loss": 0.1711,
+      "step": 903
+    },
+    {
+      "epoch": 1.9884424876169509,
+      "grad_norm": 0.46691733598709106,
+      "learning_rate": 6.164282386749249e-05,
+      "loss": 0.0967,
+      "step": 904
+    },
+    {
+      "epoch": 1.9906439185470557,
+      "grad_norm": 0.9631969928741455,
+      "learning_rate": 6.140610980691921e-05,
+      "loss": 0.1244,
+      "step": 905
+    },
+    {
+      "epoch": 1.99284534947716,
+      "grad_norm": 0.5582860112190247,
+      "learning_rate": 6.116964957806253e-05,
+      "loss": 0.0603,
+      "step": 906
+    },
+    {
+      "epoch": 1.9950467804072647,
+      "grad_norm": 0.7027082443237305,
+      "learning_rate": 6.0933444736119514e-05,
+      "loss": 0.1172,
+      "step": 907
+    },
+    {
+      "epoch": 1.9972482113373693,
+      "grad_norm": 0.5455462336540222,
+      "learning_rate": 6.069749683460765e-05,
+      "loss": 0.1221,
+      "step": 908
+    },
+    {
+      "epoch": 1.999449642267474,
+      "grad_norm": 0.6726623177528381,
+      "learning_rate": 6.046180742535441e-05,
+      "loss": 0.0661,
+      "step": 909
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9510299563407898,
+      "learning_rate": 6.022637805848723e-05,
+      "loss": 0.0573,
+      "step": 910
+    },
+    {
+      "epoch": 2.0022014309301044,
+      "grad_norm": 0.18712180852890015,
+      "learning_rate": 5.9991210282423224e-05,
+      "loss": 0.0184,
+      "step": 911
+    },
+    {
+      "epoch": 2.004402861860209,
+      "grad_norm": 0.5940394997596741,
+      "learning_rate": 5.975630564385901e-05,
+      "loss": 0.1299,
+      "step": 912
+    },
+    {
+      "epoch": 2.0066042927903136,
+      "grad_norm": 0.4946439564228058,
+      "learning_rate": 5.952166568776062e-05,
+      "loss": 0.0594,
+      "step": 913
+    },
+    {
+      "epoch": 2.0088057237204184,
+      "grad_norm": 0.18642058968544006,
+      "learning_rate": 5.928729195735319e-05,
+      "loss": 0.0206,
+      "step": 914
+    },
+    {
+      "epoch": 2.011007154650523,
+      "grad_norm": 0.5645319223403931,
+      "learning_rate": 5.9053185994110974e-05,
+      "loss": 0.1114,
+      "step": 915
+    },
+    {
+      "epoch": 2.013208585580627,
+      "grad_norm": 0.6106786131858826,
+      "learning_rate": 5.8819349337747024e-05,
+      "loss": 0.0853,
+      "step": 916
+    },
+    {
+      "epoch": 2.015410016510732,
+      "grad_norm": 0.4686431586742401,
+      "learning_rate": 5.858578352620322e-05,
+      "loss": 0.0312,
+      "step": 917
+    },
+    {
+      "epoch": 2.0176114474408364,
+      "grad_norm": 0.4860677123069763,
+      "learning_rate": 5.835249009564012e-05,
+      "loss": 0.059,
+      "step": 918
+    },
+    {
+      "epoch": 2.0198128783709413,
+      "grad_norm": 0.3909473121166229,
+      "learning_rate": 5.811947058042677e-05,
+      "loss": 0.0291,
+      "step": 919
+    },
+    {
+      "epoch": 2.0220143093010456,
+      "grad_norm": 0.7224319577217102,
+      "learning_rate": 5.7886726513130784e-05,
+      "loss": 0.0636,
+      "step": 920
+    },
+    {
+      "epoch": 2.02421574023115,
+      "grad_norm": 0.5846138596534729,
+      "learning_rate": 5.765425942450802e-05,
+      "loss": 0.0464,
+      "step": 921
+    },
+    {
+      "epoch": 2.026417171161255,
+      "grad_norm": 0.6430914402008057,
+      "learning_rate": 5.7422070843492734e-05,
+      "loss": 0.0387,
+      "step": 922
+    },
+    {
+      "epoch": 2.0286186020913592,
+      "grad_norm": 0.9888489246368408,
+      "learning_rate": 5.7190162297187475e-05,
+      "loss": 0.0953,
+      "step": 923
+    },
+    {
+      "epoch": 2.030820033021464,
+      "grad_norm": 0.6899676322937012,
+      "learning_rate": 5.695853531085287e-05,
+      "loss": 0.0829,
+      "step": 924
+    },
+    {
+      "epoch": 2.0330214639515685,
+      "grad_norm": 0.44638827443122864,
+      "learning_rate": 5.672719140789786e-05,
+      "loss": 0.0613,
+      "step": 925
+    },
+    {
+      "epoch": 2.0352228948816733,
+      "grad_norm": 0.27008816599845886,
+      "learning_rate": 5.6496132109869535e-05,
+      "loss": 0.0314,
+      "step": 926
+    },
+    {
+      "epoch": 2.0374243258117777,
+      "grad_norm": 0.11629890650510788,
+      "learning_rate": 5.626535893644307e-05,
+      "loss": 0.0054,
+      "step": 927
+    },
+    {
+      "epoch": 2.039625756741882,
+      "grad_norm": 0.47062039375305176,
+      "learning_rate": 5.60348734054118e-05,
+      "loss": 0.0458,
+      "step": 928
+    },
+    {
+      "epoch": 2.041827187671987,
+      "grad_norm": 0.6004695892333984,
+      "learning_rate": 5.5804677032677354e-05,
+      "loss": 0.0696,
+      "step": 929
+    },
+    {
+      "epoch": 2.0440286186020913,
+      "grad_norm": 0.42884814739227295,
+      "learning_rate": 5.5574771332239406e-05,
+      "loss": 0.0371,
+      "step": 930
+    },
+    {
+      "epoch": 2.046230049532196,
+      "grad_norm": 0.2548038065433502,
+      "learning_rate": 5.534515781618604e-05,
+      "loss": 0.0154,
+      "step": 931
+    },
+    {
+      "epoch": 2.0484314804623005,
+      "grad_norm": 0.8348471522331238,
+      "learning_rate": 5.5115837994683515e-05,
+      "loss": 0.0524,
+      "step": 932
+    },
+    {
+      "epoch": 2.050632911392405,
+      "grad_norm": 1.414502739906311,
+      "learning_rate": 5.488681337596653e-05,
+      "loss": 0.0924,
+      "step": 933
+    },
+    {
+      "epoch": 2.0528343423225097,
+      "grad_norm": 1.2586158514022827,
+      "learning_rate": 5.4658085466328294e-05,
+      "loss": 0.0615,
+      "step": 934
+    },
+    {
+      "epoch": 2.055035773252614,
+      "grad_norm": 0.6914826035499573,
+      "learning_rate": 5.442965577011039e-05,
+      "loss": 0.0396,
+      "step": 935
+    },
+    {
+      "epoch": 2.057237204182719,
+      "grad_norm": 0.28960347175598145,
+      "learning_rate": 5.420152578969326e-05,
+      "loss": 0.0082,
+      "step": 936
+    },
+    {
+      "epoch": 2.0594386351128233,
+      "grad_norm": 0.8626177310943604,
+      "learning_rate": 5.3973697025485934e-05,
+      "loss": 0.1057,
+      "step": 937
+    },
+    {
+      "epoch": 2.0616400660429277,
+      "grad_norm": 0.3073163330554962,
+      "learning_rate": 5.37461709759165e-05,
+      "loss": 0.0143,
+      "step": 938
+    },
+    {
+      "epoch": 2.0638414969730325,
+      "grad_norm": 0.7219668626785278,
+      "learning_rate": 5.351894913742193e-05,
+      "loss": 0.1273,
+      "step": 939
+    },
+    {
+      "epoch": 2.066042927903137,
+      "grad_norm": 1.7083491086959839,
+      "learning_rate": 5.32920330044386e-05,
+      "loss": 0.1951,
+      "step": 940
+    },
+    {
+      "epoch": 2.0682443588332418,
+      "grad_norm": 0.7081314325332642,
+      "learning_rate": 5.3065424069392064e-05,
+      "loss": 0.0876,
+      "step": 941
+    },
+    {
+      "epoch": 2.070445789763346,
+      "grad_norm": 0.358746737241745,
+      "learning_rate": 5.283912382268762e-05,
+      "loss": 0.0139,
+      "step": 942
+    },
+    {
+      "epoch": 2.0726472206934505,
+      "grad_norm": 0.539964497089386,
+      "learning_rate": 5.261313375270014e-05,
+      "loss": 0.0389,
+      "step": 943
+    },
+    {
+      "epoch": 2.0748486516235554,
+      "grad_norm": 0.6323754191398621,
+      "learning_rate": 5.238745534576461e-05,
+      "loss": 0.0589,
+      "step": 944
+    },
+    {
+      "epoch": 2.0770500825536597,
+      "grad_norm": 1.231791615486145,
+      "learning_rate": 5.2162090086166215e-05,
+      "loss": 0.312,
+      "step": 945
+    },
+    {
+      "epoch": 2.0792515134837646,
+      "grad_norm": 0.4685794711112976,
+      "learning_rate": 5.1937039456130435e-05,
+      "loss": 0.0482,
+      "step": 946
+    },
+    {
+      "epoch": 2.081452944413869,
+      "grad_norm": 0.15832555294036865,
+      "learning_rate": 5.171230493581358e-05,
+      "loss": 0.0182,
+      "step": 947
+    },
+    {
+      "epoch": 2.083654375343974,
+      "grad_norm": 0.6063134074211121,
+      "learning_rate": 5.148788800329278e-05,
+      "loss": 0.0802,
+      "step": 948
+    },
+    {
+      "epoch": 2.085855806274078,
+      "grad_norm": 0.4775516092777252,
+      "learning_rate": 5.126379013455656e-05,
+      "loss": 0.0462,
+      "step": 949
+    },
+    {
+      "epoch": 2.0880572372041826,
+      "grad_norm": 0.6208484768867493,
+      "learning_rate": 5.1040012803494795e-05,
+      "loss": 0.039,
+      "step": 950
+    },
+    {
+      "epoch": 2.0902586681342874,
+      "grad_norm": 0.6610376238822937,
+      "learning_rate": 5.081655748188924e-05,
+      "loss": 0.0688,
+      "step": 951
+    },
+    {
+      "epoch": 2.092460099064392,
+      "grad_norm": 0.20292621850967407,
+      "learning_rate": 5.059342563940384e-05,
+      "loss": 0.0249,
+      "step": 952
+    },
+    {
+      "epoch": 2.0946615299944966,
+      "grad_norm": 0.579462468624115,
+      "learning_rate": 5.0370618743575026e-05,
+      "loss": 0.0871,
+      "step": 953
+    },
+    {
+      "epoch": 2.096862960924601,
+      "grad_norm": 0.178482323884964,
+      "learning_rate": 5.014813825980196e-05,
+      "loss": 0.0197,
+      "step": 954
+    },
+    {
+      "epoch": 2.0990643918547054,
+      "grad_norm": 0.6209052801132202,
+      "learning_rate": 4.992598565133709e-05,
+      "loss": 0.1135,
+      "step": 955
+    },
+    {
+      "epoch": 2.1012658227848102,
+      "grad_norm": 0.3643285036087036,
+      "learning_rate": 4.970416237927645e-05,
+      "loss": 0.0426,
+      "step": 956
+    },
+    {
+      "epoch": 2.1034672537149146,
+      "grad_norm": 0.9399219155311584,
+      "learning_rate": 4.9482669902549894e-05,
+      "loss": 0.0895,
+      "step": 957
+    },
+    {
+      "epoch": 2.1056686846450194,
+      "grad_norm": 1.401642084121704,
+      "learning_rate": 4.92615096779118e-05,
+      "loss": 0.0861,
+      "step": 958
+    },
+    {
+      "epoch": 2.107870115575124,
+      "grad_norm": 0.856060266494751,
+      "learning_rate": 4.904068315993118e-05,
+      "loss": 0.1042,
+      "step": 959
+    },
+    {
+      "epoch": 2.110071546505228,
+      "grad_norm": 0.646442174911499,
+      "learning_rate": 4.882019180098236e-05,
+      "loss": 0.0765,
+      "step": 960
+    },
+    {
+      "epoch": 2.112272977435333,
+      "grad_norm": 0.6977699995040894,
+      "learning_rate": 4.860003705123538e-05,
+      "loss": 0.094,
+      "step": 961
+    },
+    {
+      "epoch": 2.1144744083654374,
+      "grad_norm": 0.41784411668777466,
+      "learning_rate": 4.838022035864618e-05,
+      "loss": 0.0435,
+      "step": 962
+    },
+    {
+      "epoch": 2.1166758392955423,
+      "grad_norm": 0.8824518918991089,
+      "learning_rate": 4.8160743168947496e-05,
+      "loss": 0.1005,
+      "step": 963
+    },
+    {
+      "epoch": 2.1188772702256466,
+      "grad_norm": 0.3907751739025116,
+      "learning_rate": 4.794160692563917e-05,
+      "loss": 0.0564,
+      "step": 964
+    },
+    {
+      "epoch": 2.121078701155751,
+      "grad_norm": 1.0237921476364136,
+      "learning_rate": 4.772281306997848e-05,
+      "loss": 0.1325,
+      "step": 965
+    },
+    {
+      "epoch": 2.123280132085856,
+      "grad_norm": 0.45119601488113403,
+      "learning_rate": 4.750436304097099e-05,
+      "loss": 0.042,
+      "step": 966
+    },
+    {
+      "epoch": 2.1254815630159603,
+      "grad_norm": 0.4329589903354645,
+      "learning_rate": 4.7286258275360796e-05,
+      "loss": 0.0516,
+      "step": 967
+    },
+    {
+      "epoch": 2.127682993946065,
+      "grad_norm": 0.34292638301849365,
+      "learning_rate": 4.706850020762126e-05,
+      "loss": 0.025,
+      "step": 968
+    },
+    {
+      "epoch": 2.1298844248761695,
+      "grad_norm": 0.6047242879867554,
+      "learning_rate": 4.685109026994556e-05,
+      "loss": 0.0519,
+      "step": 969
+    },
+    {
+      "epoch": 2.132085855806274,
+      "grad_norm": 0.31319746375083923,
+      "learning_rate": 4.66340298922371e-05,
+      "loss": 0.0652,
+      "step": 970
+    },
+    {
+      "epoch": 2.1342872867363787,
+      "grad_norm": 0.45534616708755493,
+      "learning_rate": 4.6417320502100316e-05,
+      "loss": 0.0546,
+      "step": 971
+    },
+    {
+      "epoch": 2.136488717666483,
+      "grad_norm": 0.28852415084838867,
+      "learning_rate": 4.620096352483129e-05,
+      "loss": 0.0362,
+      "step": 972
+    },
+    {
+      "epoch": 2.138690148596588,
+      "grad_norm": 0.6879104971885681,
+      "learning_rate": 4.5984960383408005e-05,
+      "loss": 0.0801,
+      "step": 973
+    },
+    {
+      "epoch": 2.1408915795266923,
+      "grad_norm": 0.3644227087497711,
+      "learning_rate": 4.5769312498481554e-05,
+      "loss": 0.0376,
+      "step": 974
+    },
+    {
+      "epoch": 2.143093010456797,
+      "grad_norm": 0.2540081739425659,
+      "learning_rate": 4.555402128836642e-05,
+      "loss": 0.0182,
+      "step": 975
+    },
+    {
+      "epoch": 2.1452944413869015,
+      "grad_norm": 0.40438270568847656,
+      "learning_rate": 4.533908816903115e-05,
+      "loss": 0.0278,
+      "step": 976
+    },
+    {
+      "epoch": 2.147495872317006,
+      "grad_norm": 0.40767142176628113,
+      "learning_rate": 4.512451455408929e-05,
+      "loss": 0.0435,
+      "step": 977
+    },
+    {
+      "epoch": 2.1496973032471107,
+      "grad_norm": 0.4094831049442291,
+      "learning_rate": 4.491030185478976e-05,
+      "loss": 0.0443,
+      "step": 978
+    },
+    {
+      "epoch": 2.151898734177215,
+      "grad_norm": 0.8176153302192688,
+      "learning_rate": 4.469645148000785e-05,
+      "loss": 0.0795,
+      "step": 979
+    },
+    {
+      "epoch": 2.15410016510732,
+      "grad_norm": 1.6155697107315063,
+      "learning_rate": 4.4482964836235875e-05,
+      "loss": 0.2632,
+      "step": 980
+    },
+    {
+      "epoch": 2.1563015960374243,
+      "grad_norm": 0.2684282660484314,
+      "learning_rate": 4.426984332757375e-05,
+      "loss": 0.0324,
+      "step": 981
+    },
+    {
+      "epoch": 2.1585030269675287,
+      "grad_norm": 0.312021404504776,
+      "learning_rate": 4.4057088355720054e-05,
+      "loss": 0.0185,
+      "step": 982
+    },
+    {
+      "epoch": 2.1607044578976335,
+      "grad_norm": 0.4544532895088196,
+      "learning_rate": 4.384470131996252e-05,
+      "loss": 0.0332,
+      "step": 983
+    },
+    {
+      "epoch": 2.162905888827738,
+      "grad_norm": 0.348593145608902,
+      "learning_rate": 4.3632683617169124e-05,
+      "loss": 0.0505,
+      "step": 984
+    },
+    {
+      "epoch": 2.1651073197578428,
+      "grad_norm": 0.5245181918144226,
+      "learning_rate": 4.3421036641778556e-05,
+      "loss": 0.0464,
+      "step": 985
+    },
+    {
+      "epoch": 2.167308750687947,
+      "grad_norm": 0.6938517689704895,
+      "learning_rate": 4.3209761785791416e-05,
+      "loss": 0.1021,
+      "step": 986
+    },
+    {
+      "epoch": 2.1695101816180515,
+      "grad_norm": 1.0149823427200317,
+      "learning_rate": 4.299886043876071e-05,
+      "loss": 0.0921,
+      "step": 987
+    },
+    {
+      "epoch": 2.1717116125481564,
+      "grad_norm": 0.5596351027488708,
+      "learning_rate": 4.278833398778306e-05,
+      "loss": 0.031,
+      "step": 988
+    },
+    {
+      "epoch": 2.1739130434782608,
+      "grad_norm": 0.8479326963424683,
+      "learning_rate": 4.2578183817489216e-05,
+      "loss": 0.0533,
+      "step": 989
+    },
+    {
+      "epoch": 2.1761144744083656,
+      "grad_norm": 0.9328631162643433,
+      "learning_rate": 4.236841131003524e-05,
+      "loss": 0.1186,
+      "step": 990
+    },
+    {
+      "epoch": 2.17831590533847,
+      "grad_norm": 0.9092453122138977,
+      "learning_rate": 4.215901784509335e-05,
+      "loss": 0.0417,
+      "step": 991
+    },
+    {
+      "epoch": 2.180517336268575,
+      "grad_norm": 0.30856090784072876,
+      "learning_rate": 4.195000479984265e-05,
+      "loss": 0.0182,
+      "step": 992
+    },
+    {
+      "epoch": 2.182718767198679,
+      "grad_norm": 1.2621026039123535,
+      "learning_rate": 4.174137354896039e-05,
+      "loss": 0.1173,
+      "step": 993
+    },
+    {
+      "epoch": 2.1849201981287836,
+      "grad_norm": 0.4759381413459778,
+      "learning_rate": 4.153312546461264e-05,
+      "loss": 0.063,
+      "step": 994
+    },
+    {
+      "epoch": 2.1871216290588884,
+      "grad_norm": 0.458916574716568,
+      "learning_rate": 4.132526191644549e-05,
+      "loss": 0.0752,
+      "step": 995
+    },
+    {
+      "epoch": 2.189323059988993,
+      "grad_norm": 1.4117220640182495,
+      "learning_rate": 4.111778427157586e-05,
+      "loss": 0.2388,
+      "step": 996
+    },
+    {
+      "epoch": 2.1915244909190976,
+      "grad_norm": 0.2887744903564453,
+      "learning_rate": 4.091069389458255e-05,
+      "loss": 0.0243,
+      "step": 997
+    },
+    {
+      "epoch": 2.193725921849202,
+      "grad_norm": 0.3760842978954315,
+      "learning_rate": 4.0703992147497425e-05,
+      "loss": 0.0427,
+      "step": 998
+    },
+    {
+      "epoch": 2.1959273527793064,
+      "grad_norm": 0.3974992632865906,
+      "learning_rate": 4.049768038979631e-05,
+      "loss": 0.0353,
+      "step": 999
+    },
+    {
+      "epoch": 2.1981287837094112,
+      "grad_norm": 0.5743415951728821,
+      "learning_rate": 4.029175997838996e-05,
+      "loss": 0.0487,
+      "step": 1000
+    },
+    {
+      "epoch": 2.2003302146395156,
+      "grad_norm": 0.5577808022499084,
+      "learning_rate": 4.008623226761534e-05,
+      "loss": 0.0599,
+      "step": 1001
+    },
+    {
+      "epoch": 2.2025316455696204,
+      "grad_norm": 0.2979278862476349,
+      "learning_rate": 3.9881098609226666e-05,
+      "loss": 0.04,
+      "step": 1002
+    },
+    {
+      "epoch": 2.204733076499725,
+      "grad_norm": 0.439411461353302,
+      "learning_rate": 3.9676360352386356e-05,
+      "loss": 0.0251,
+      "step": 1003
+    },
+    {
+      "epoch": 2.206934507429829,
+      "grad_norm": 0.5079408288002014,
+      "learning_rate": 3.947201884365639e-05,
+      "loss": 0.054,
+      "step": 1004
+    },
+    {
+      "epoch": 2.209135938359934,
+      "grad_norm": 0.6212779879570007,
+      "learning_rate": 3.926807542698922e-05,
+      "loss": 0.0379,
+      "step": 1005
+    },
+    {
+      "epoch": 2.2113373692900384,
+      "grad_norm": 1.1278338432312012,
+      "learning_rate": 3.90645314437192e-05,
+      "loss": 0.0723,
+      "step": 1006
+    },
+    {
+      "epoch": 2.2135388002201433,
+      "grad_norm": 0.842240571975708,
+      "learning_rate": 3.8861388232553486e-05,
+      "loss": 0.0702,
+      "step": 1007
+    },
+    {
+      "epoch": 2.2157402311502477,
+      "grad_norm": 0.3266470730304718,
+      "learning_rate": 3.8658647129563364e-05,
+      "loss": 0.0208,
+      "step": 1008
+    },
+    {
+      "epoch": 2.217941662080352,
+      "grad_norm": 2.0979557037353516,
+      "learning_rate": 3.845630946817553e-05,
+      "loss": 0.3657,
+      "step": 1009
+    },
+    {
+      "epoch": 2.220143093010457,
+      "grad_norm": 0.33687448501586914,
+      "learning_rate": 3.8254376579163255e-05,
+      "loss": 0.0321,
+      "step": 1010
+    },
+    {
+      "epoch": 2.2223445239405613,
+      "grad_norm": 0.9908032417297363,
+      "learning_rate": 3.8052849790637524e-05,
+      "loss": 0.0541,
+      "step": 1011
+    },
+    {
+      "epoch": 2.224545954870666,
+      "grad_norm": 1.3826929330825806,
+      "learning_rate": 3.785173042803848e-05,
+      "loss": 0.1712,
+      "step": 1012
+    },
+    {
+      "epoch": 2.2267473858007705,
+      "grad_norm": 0.5907435417175293,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 0.0812,
+      "step": 1013
+    },
+    {
+      "epoch": 2.228948816730875,
+      "grad_norm": 0.4497785270214081,
+      "learning_rate": 3.7450719268974136e-05,
+      "loss": 0.0515,
+      "step": 1014
+    },
+    {
+      "epoch": 2.2311502476609797,
+      "grad_norm": 0.6755375266075134,
+      "learning_rate": 3.725083010995611e-05,
+      "loss": 0.0513,
+      "step": 1015
+    },
+    {
+      "epoch": 2.233351678591084,
+      "grad_norm": 0.5040784478187561,
+      "learning_rate": 3.705135365174197e-05,
+      "loss": 0.0256,
+      "step": 1016
+    },
+    {
+      "epoch": 2.235553109521189,
+      "grad_norm": 0.6088061332702637,
+      "learning_rate": 3.685229120628688e-05,
+      "loss": 0.068,
+      "step": 1017
+    },
+    {
+      "epoch": 2.2377545404512933,
+      "grad_norm": 1.4253185987472534,
+      "learning_rate": 3.665364408282305e-05,
+      "loss": 0.0648,
+      "step": 1018
+    },
+    {
+      "epoch": 2.2399559713813977,
+      "grad_norm": 0.7487550377845764,
+      "learning_rate": 3.645541358785093e-05,
+      "loss": 0.0932,
+      "step": 1019
+    },
+    {
+      "epoch": 2.2421574023115025,
+      "grad_norm": 1.2077827453613281,
+      "learning_rate": 3.6257601025131026e-05,
+      "loss": 0.1227,
+      "step": 1020
+    },
+    {
+      "epoch": 2.244358833241607,
+      "grad_norm": 0.5581986308097839,
+      "learning_rate": 3.6060207695675075e-05,
+      "loss": 0.0445,
+      "step": 1021
+    },
+    {
+      "epoch": 2.2465602641717117,
+      "grad_norm": 0.48802509903907776,
+      "learning_rate": 3.5863234897737395e-05,
+      "loss": 0.0582,
+      "step": 1022
+    },
+    {
+      "epoch": 2.248761695101816,
+      "grad_norm": 0.674498975276947,
+      "learning_rate": 3.566668392680662e-05,
+      "loss": 0.0877,
+      "step": 1023
+    },
+    {
+      "epoch": 2.2509631260319205,
+      "grad_norm": 0.49444684386253357,
+      "learning_rate": 3.547055607559688e-05,
+      "loss": 0.0512,
+      "step": 1024
+    },
+    {
+      "epoch": 2.2531645569620253,
+      "grad_norm": 0.914577841758728,
+      "learning_rate": 3.52748526340396e-05,
+      "loss": 0.124,
+      "step": 1025
+    },
+    {
+      "epoch": 2.2553659878921297,
+      "grad_norm": 0.6520012617111206,
+      "learning_rate": 3.50795748892748e-05,
+      "loss": 0.0856,
+      "step": 1026
+    },
+    {
+      "epoch": 2.2575674188222346,
+      "grad_norm": 0.522365391254425,
+      "learning_rate": 3.488472412564264e-05,
+      "loss": 0.0431,
+      "step": 1027
+    },
+    {
+      "epoch": 2.259768849752339,
+      "grad_norm": 0.5674323439598083,
+      "learning_rate": 3.469030162467513e-05,
+      "loss": 0.0738,
+      "step": 1028
+    },
+    {
+      "epoch": 2.2619702806824438,
+      "grad_norm": 0.2617637515068054,
+      "learning_rate": 3.449630866508757e-05,
+      "loss": 0.0154,
+      "step": 1029
+    },
+    {
+      "epoch": 2.264171711612548,
+      "grad_norm": 0.5464626550674438,
+      "learning_rate": 3.430274652277008e-05,
+      "loss": 0.0509,
+      "step": 1030
+    },
+    {
+      "epoch": 2.2663731425426525,
+      "grad_norm": 0.2028905153274536,
+      "learning_rate": 3.4109616470779394e-05,
+      "loss": 0.0213,
+      "step": 1031
+    },
+    {
+      "epoch": 2.2685745734727574,
+      "grad_norm": 0.1778368353843689,
+      "learning_rate": 3.39169197793304e-05,
+      "loss": 0.0191,
+      "step": 1032
+    },
+    {
+      "epoch": 2.2707760044028618,
+      "grad_norm": 0.12138936668634415,
+      "learning_rate": 3.372465771578771e-05,
+      "loss": 0.0121,
+      "step": 1033
+    },
+    {
+      "epoch": 2.2729774353329666,
+      "grad_norm": 0.18100698292255402,
+      "learning_rate": 3.353283154465746e-05,
+      "loss": 0.0115,
+      "step": 1034
+    },
+    {
+      "epoch": 2.275178866263071,
+      "grad_norm": 0.301810622215271,
+      "learning_rate": 3.3341442527578836e-05,
+      "loss": 0.0232,
+      "step": 1035
+    },
+    {
+      "epoch": 2.277380297193176,
+      "grad_norm": 0.7786280512809753,
+      "learning_rate": 3.315049192331595e-05,
+      "loss": 0.0828,
+      "step": 1036
+    },
+    {
+      "epoch": 2.27958172812328,
+      "grad_norm": 1.0666403770446777,
+      "learning_rate": 3.295998098774948e-05,
+      "loss": 0.0557,
+      "step": 1037
+    },
+    {
+      "epoch": 2.2817831590533846,
+      "grad_norm": 0.1729314625263214,
+      "learning_rate": 3.276991097386831e-05,
+      "loss": 0.0144,
+      "step": 1038
+    },
+    {
+      "epoch": 2.2839845899834894,
+      "grad_norm": 0.45126134157180786,
+      "learning_rate": 3.2580283131761515e-05,
+      "loss": 0.0442,
+      "step": 1039
+    },
+    {
+      "epoch": 2.286186020913594,
+      "grad_norm": 0.8280728459358215,
+      "learning_rate": 3.2391098708609905e-05,
+      "loss": 0.1192,
+      "step": 1040
+    },
+    {
+      "epoch": 2.2883874518436986,
+      "grad_norm": 0.3810870945453644,
+      "learning_rate": 3.220235894867794e-05,
+      "loss": 0.0628,
+      "step": 1041
+    },
+    {
+      "epoch": 2.290588882773803,
+      "grad_norm": 0.3795784115791321,
+      "learning_rate": 3.2014065093305565e-05,
+      "loss": 0.0403,
+      "step": 1042
+    },
+    {
+      "epoch": 2.2927903137039074,
+      "grad_norm": 0.2101019322872162,
+      "learning_rate": 3.1826218380900064e-05,
+      "loss": 0.0125,
+      "step": 1043
+    },
+    {
+      "epoch": 2.2949917446340122,
+      "grad_norm": 0.8287165760993958,
+      "learning_rate": 3.163882004692774e-05,
+      "loss": 0.0879,
+      "step": 1044
+    },
+    {
+      "epoch": 2.2971931755641166,
+      "grad_norm": 0.6925108432769775,
+      "learning_rate": 3.1451871323906046e-05,
+      "loss": 0.0599,
+      "step": 1045
+    },
+    {
+      "epoch": 2.2993946064942214,
+      "grad_norm": 0.18960566818714142,
+      "learning_rate": 3.1265373441395205e-05,
+      "loss": 0.0171,
+      "step": 1046
+    },
+    {
+      "epoch": 2.301596037424326,
+      "grad_norm": 0.4648805558681488,
+      "learning_rate": 3.107932762599041e-05,
+      "loss": 0.0564,
+      "step": 1047
+    },
+    {
+      "epoch": 2.3037974683544302,
+      "grad_norm": 0.8541012406349182,
+      "learning_rate": 3.089373510131354e-05,
+      "loss": 0.0716,
+      "step": 1048
+    },
+    {
+      "epoch": 2.305998899284535,
+      "grad_norm": 0.2366633266210556,
+      "learning_rate": 3.0708597088005156e-05,
+      "loss": 0.0247,
+      "step": 1049
+    },
+    {
+      "epoch": 2.3082003302146394,
+      "grad_norm": 0.286531537771225,
+      "learning_rate": 3.052391480371658e-05,
+      "loss": 0.0429,
+      "step": 1050
+    },
+    {
+      "epoch": 2.3104017611447443,
+      "grad_norm": 0.5673994421958923,
+      "learning_rate": 3.0339689463101718e-05,
+      "loss": 0.0594,
+      "step": 1051
+    },
+    {
+      "epoch": 2.3126031920748487,
+      "grad_norm": 1.424087405204773,
+      "learning_rate": 3.0155922277809257e-05,
+      "loss": 0.1185,
+      "step": 1052
+    },
+    {
+      "epoch": 2.314804623004953,
+      "grad_norm": 0.4548904001712799,
+      "learning_rate": 2.9972614456474536e-05,
+      "loss": 0.0281,
+      "step": 1053
+    },
+    {
+      "epoch": 2.317006053935058,
+      "grad_norm": 0.27611851692199707,
+      "learning_rate": 2.9789767204711618e-05,
+      "loss": 0.0377,
+      "step": 1054
+    },
+    {
+      "epoch": 2.3192074848651623,
+      "grad_norm": 0.4677520990371704,
+      "learning_rate": 2.960738172510551e-05,
+      "loss": 0.0692,
+      "step": 1055
+    },
+    {
+      "epoch": 2.321408915795267,
+      "grad_norm": 0.6894844770431519,
+      "learning_rate": 2.9425459217204123e-05,
+      "loss": 0.0921,
+      "step": 1056
+    },
+    {
+      "epoch": 2.3236103467253715,
+      "grad_norm": 0.8794598579406738,
+      "learning_rate": 2.9244000877510314e-05,
+      "loss": 0.0692,
+      "step": 1057
+    },
+    {
+      "epoch": 2.325811777655476,
+      "grad_norm": 0.19780100882053375,
+      "learning_rate": 2.9063007899474216e-05,
+      "loss": 0.0174,
+      "step": 1058
+    },
+    {
+      "epoch": 2.3280132085855807,
+      "grad_norm": 0.4535125195980072,
+      "learning_rate": 2.888248147348528e-05,
+      "loss": 0.0309,
+      "step": 1059
+    },
+    {
+      "epoch": 2.330214639515685,
+      "grad_norm": 1.4072175025939941,
+      "learning_rate": 2.8702422786864326e-05,
+      "loss": 0.1489,
+      "step": 1060
+    },
+    {
+      "epoch": 2.33241607044579,
+      "grad_norm": 0.38870853185653687,
+      "learning_rate": 2.852283302385602e-05,
+      "loss": 0.0372,
+      "step": 1061
+    },
+    {
+      "epoch": 2.3346175013758943,
+      "grad_norm": 0.3546408414840698,
+      "learning_rate": 2.8343713365620772e-05,
+      "loss": 0.0222,
+      "step": 1062
+    },
+    {
+      "epoch": 2.3368189323059987,
+      "grad_norm": 0.39257052540779114,
+      "learning_rate": 2.8165064990227252e-05,
+      "loss": 0.0375,
+      "step": 1063
+    },
+    {
+      "epoch": 2.3390203632361035,
+      "grad_norm": 0.7221199870109558,
+      "learning_rate": 2.7986889072644406e-05,
+      "loss": 0.0546,
+      "step": 1064
+    },
+    {
+      "epoch": 2.341221794166208,
+      "grad_norm": 0.9033507704734802,
+      "learning_rate": 2.7809186784733833e-05,
+      "loss": 0.1005,
+      "step": 1065
+    },
+    {
+      "epoch": 2.3434232250963127,
+      "grad_norm": 0.23366965353488922,
+      "learning_rate": 2.7631959295242126e-05,
+      "loss": 0.0202,
+      "step": 1066
+    },
+    {
+      "epoch": 2.345624656026417,
+      "grad_norm": 0.5312491059303284,
+      "learning_rate": 2.7455207769793158e-05,
+      "loss": 0.0485,
+      "step": 1067
+    },
+    {
+      "epoch": 2.3478260869565215,
+      "grad_norm": 1.1412675380706787,
+      "learning_rate": 2.7278933370880265e-05,
+      "loss": 0.0476,
+      "step": 1068
+    },
+    {
+      "epoch": 2.3500275178866263,
+      "grad_norm": 0.46408095955848694,
+      "learning_rate": 2.7103137257858868e-05,
+      "loss": 0.0335,
+      "step": 1069
+    },
+    {
+      "epoch": 2.3522289488167307,
+      "grad_norm": 0.4340619742870331,
+      "learning_rate": 2.692782058693858e-05,
+      "loss": 0.0681,
+      "step": 1070
+    },
+    {
+      "epoch": 2.3544303797468356,
+      "grad_norm": 0.2784384489059448,
+      "learning_rate": 2.6752984511175818e-05,
+      "loss": 0.0265,
+      "step": 1071
+    },
+    {
+      "epoch": 2.35663181067694,
+      "grad_norm": 0.1808812916278839,
+      "learning_rate": 2.6578630180466114e-05,
+      "loss": 0.0088,
+      "step": 1072
+    },
+    {
+      "epoch": 2.3588332416070448,
+      "grad_norm": 0.5902241468429565,
+      "learning_rate": 2.6404758741536505e-05,
+      "loss": 0.0487,
+      "step": 1073
+    },
+    {
+      "epoch": 2.361034672537149,
+      "grad_norm": 1.0578711032867432,
+      "learning_rate": 2.6231371337938148e-05,
+      "loss": 0.2177,
+      "step": 1074
+    },
+    {
+      "epoch": 2.3632361034672535,
+      "grad_norm": 0.7263875007629395,
+      "learning_rate": 2.605846911003863e-05,
+      "loss": 0.1247,
+      "step": 1075
+    },
+    {
+      "epoch": 2.3654375343973584,
+      "grad_norm": 0.34880584478378296,
+      "learning_rate": 2.5886053195014538e-05,
+      "loss": 0.0326,
+      "step": 1076
+    },
+    {
+      "epoch": 2.3676389653274628,
+      "grad_norm": 0.28905266523361206,
+      "learning_rate": 2.5714124726844014e-05,
+      "loss": 0.0282,
+      "step": 1077
+    },
+    {
+      "epoch": 2.3698403962575676,
+      "grad_norm": 0.5269353985786438,
+      "learning_rate": 2.5542684836299313e-05,
+      "loss": 0.0965,
+      "step": 1078
+    },
+    {
+      "epoch": 2.372041827187672,
+      "grad_norm": 0.41657519340515137,
+      "learning_rate": 2.5371734650939205e-05,
+      "loss": 0.0348,
+      "step": 1079
+    },
+    {
+      "epoch": 2.3742432581177764,
+      "grad_norm": 0.3339771032333374,
+      "learning_rate": 2.520127529510178e-05,
+      "loss": 0.0282,
+      "step": 1080
+    },
+    {
+      "epoch": 2.376444689047881,
+      "grad_norm": 0.56503826379776,
+      "learning_rate": 2.5031307889896847e-05,
+      "loss": 0.051,
+      "step": 1081
+    },
+    {
+      "epoch": 2.3786461199779856,
+      "grad_norm": 0.1981649547815323,
+      "learning_rate": 2.4861833553198756e-05,
+      "loss": 0.0194,
+      "step": 1082
+    },
+    {
+      "epoch": 2.3808475509080904,
+      "grad_norm": 0.3235817551612854,
+      "learning_rate": 2.4692853399638917e-05,
+      "loss": 0.0451,
+      "step": 1083
+    },
+    {
+      "epoch": 2.383048981838195,
+      "grad_norm": 0.4440438449382782,
+      "learning_rate": 2.452436854059843e-05,
+      "loss": 0.0249,
+      "step": 1084
+    },
+    {
+      "epoch": 2.3852504127682996,
+      "grad_norm": 0.2182845175266266,
+      "learning_rate": 2.4356380084200982e-05,
+      "loss": 0.0288,
+      "step": 1085
+    },
+    {
+      "epoch": 2.387451843698404,
+      "grad_norm": 0.9195117354393005,
+      "learning_rate": 2.4188889135305305e-05,
+      "loss": 0.1122,
+      "step": 1086
+    },
+    {
+      "epoch": 2.3896532746285084,
+      "grad_norm": 0.4600276052951813,
+      "learning_rate": 2.4021896795498043e-05,
+      "loss": 0.0539,
+      "step": 1087
+    },
+    {
+      "epoch": 2.3918547055586132,
+      "grad_norm": 0.5593721270561218,
+      "learning_rate": 2.3855404163086558e-05,
+      "loss": 0.0525,
+      "step": 1088
+    },
+    {
+      "epoch": 2.3940561364887176,
+      "grad_norm": 0.7966195940971375,
+      "learning_rate": 2.368941233309162e-05,
+      "loss": 0.0916,
+      "step": 1089
+    },
+    {
+      "epoch": 2.3962575674188225,
+      "grad_norm": 0.9492884278297424,
+      "learning_rate": 2.352392239724016e-05,
+      "loss": 0.0875,
+      "step": 1090
+    },
+    {
+      "epoch": 2.398458998348927,
+      "grad_norm": 1.1402668952941895,
+      "learning_rate": 2.3358935443958262e-05,
+      "loss": 0.1569,
+      "step": 1091
+    },
+    {
+      "epoch": 2.4006604292790312,
+      "grad_norm": 0.17830142378807068,
+      "learning_rate": 2.3194452558363778e-05,
+      "loss": 0.0115,
+      "step": 1092
+    },
+    {
+      "epoch": 2.402861860209136,
+      "grad_norm": 0.270605206489563,
+      "learning_rate": 2.3030474822259397e-05,
+      "loss": 0.0375,
+      "step": 1093
+    },
+    {
+      "epoch": 2.4050632911392404,
+      "grad_norm": 0.5833712220191956,
+      "learning_rate": 2.2867003314125446e-05,
+      "loss": 0.0638,
+      "step": 1094
+    },
+    {
+      "epoch": 2.4072647220693453,
+      "grad_norm": 0.6032495498657227,
+      "learning_rate": 2.270403910911272e-05,
+      "loss": 0.0368,
+      "step": 1095
+    },
+    {
+      "epoch": 2.4094661529994497,
+      "grad_norm": 0.8065091371536255,
+      "learning_rate": 2.254158327903557e-05,
+      "loss": 0.0925,
+      "step": 1096
+    },
+    {
+      "epoch": 2.411667583929554,
+      "grad_norm": 1.636093020439148,
+      "learning_rate": 2.237963689236472e-05,
+      "loss": 0.1379,
+      "step": 1097
+    },
+    {
+      "epoch": 2.413869014859659,
+      "grad_norm": 0.6910865902900696,
+      "learning_rate": 2.2218201014220263e-05,
+      "loss": 0.0626,
+      "step": 1098
+    },
+    {
+      "epoch": 2.4160704457897633,
+      "grad_norm": 0.1868569552898407,
+      "learning_rate": 2.2057276706364784e-05,
+      "loss": 0.0211,
+      "step": 1099
+    },
+    {
+      "epoch": 2.418271876719868,
+      "grad_norm": 0.20337679982185364,
+      "learning_rate": 2.1896865027196146e-05,
+      "loss": 0.0198,
+      "step": 1100
+    },
+    {
+      "epoch": 2.4204733076499725,
+      "grad_norm": 0.3975900113582611,
+      "learning_rate": 2.173696703174074e-05,
+      "loss": 0.0277,
+      "step": 1101
+    },
+    {
+      "epoch": 2.422674738580077,
+      "grad_norm": 0.4950021803379059,
+      "learning_rate": 2.157758377164647e-05,
+      "loss": 0.0524,
+      "step": 1102
+    },
+    {
+      "epoch": 2.4248761695101817,
+      "grad_norm": 1.2918648719787598,
+      "learning_rate": 2.141871629517577e-05,
+      "loss": 0.0979,
+      "step": 1103
+    },
+    {
+      "epoch": 2.427077600440286,
+      "grad_norm": 0.3743513524532318,
+      "learning_rate": 2.1260365647198798e-05,
+      "loss": 0.0486,
+      "step": 1104
+    },
+    {
+      "epoch": 2.429279031370391,
+      "grad_norm": 0.2542741894721985,
+      "learning_rate": 2.1102532869186587e-05,
+      "loss": 0.0221,
+      "step": 1105
+    },
+    {
+      "epoch": 2.4314804623004953,
+      "grad_norm": 0.10288311541080475,
+      "learning_rate": 2.094521899920403e-05,
+      "loss": 0.0076,
+      "step": 1106
+    },
+    {
+      "epoch": 2.4336818932305997,
+      "grad_norm": 0.4356939494609833,
+      "learning_rate": 2.0788425071903283e-05,
+      "loss": 0.0567,
+      "step": 1107
+    },
+    {
+      "epoch": 2.4358833241607045,
+      "grad_norm": 0.2828386127948761,
+      "learning_rate": 2.063215211851678e-05,
+      "loss": 0.043,
+      "step": 1108
+    },
+    {
+      "epoch": 2.438084755090809,
+      "grad_norm": 0.6171638369560242,
+      "learning_rate": 2.047640116685048e-05,
+      "loss": 0.1012,
+      "step": 1109
+    },
+    {
+      "epoch": 2.4402861860209137,
+      "grad_norm": 0.3744174540042877,
+      "learning_rate": 2.0321173241277237e-05,
+      "loss": 0.0317,
+      "step": 1110
+    },
+    {
+      "epoch": 2.442487616951018,
+      "grad_norm": 0.4005078971385956,
+      "learning_rate": 2.016646936272987e-05,
+      "loss": 0.0377,
+      "step": 1111
+    },
+    {
+      "epoch": 2.4446890478811225,
+      "grad_norm": 0.7585984468460083,
+      "learning_rate": 2.0012290548694612e-05,
+      "loss": 0.1142,
+      "step": 1112
+    },
+    {
+      "epoch": 2.4468904788112273,
+      "grad_norm": 0.5197720527648926,
+      "learning_rate": 1.985863781320435e-05,
+      "loss": 0.0358,
+      "step": 1113
+    },
+    {
+      "epoch": 2.4490919097413317,
+      "grad_norm": 0.8446899652481079,
+      "learning_rate": 1.9705512166831886e-05,
+      "loss": 0.0389,
+      "step": 1114
+    },
+    {
+      "epoch": 2.4512933406714366,
+      "grad_norm": 0.6728304624557495,
+      "learning_rate": 1.955291461668347e-05,
+      "loss": 0.045,
+      "step": 1115
+    },
+    {
+      "epoch": 2.453494771601541,
+      "grad_norm": 0.6178024411201477,
+      "learning_rate": 1.940084616639194e-05,
+      "loss": 0.0649,
+      "step": 1116
+    },
+    {
+      "epoch": 2.4556962025316453,
+      "grad_norm": 0.9444505572319031,
+      "learning_rate": 1.924930781611034e-05,
+      "loss": 0.0514,
+      "step": 1117
+    },
+    {
+      "epoch": 2.45789763346175,
+      "grad_norm": 0.6526023745536804,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.0732,
+      "step": 1118
+    },
+    {
+      "epoch": 2.4600990643918546,
+      "grad_norm": 0.6759251952171326,
+      "learning_rate": 1.8947825398750186e-05,
+      "loss": 0.0415,
+      "step": 1119
+    },
+    {
+      "epoch": 2.4623004953219594,
+      "grad_norm": 0.44757458567619324,
+      "learning_rate": 1.8797883314519148e-05,
+      "loss": 0.1711,
+      "step": 1120
+    },
+    {
+      "epoch": 2.4645019262520638,
+      "grad_norm": 0.49497660994529724,
+      "learning_rate": 1.864847529598005e-05,
+      "loss": 0.0549,
+      "step": 1121
+    },
+    {
+      "epoch": 2.4667033571821686,
+      "grad_norm": 0.504433810710907,
+      "learning_rate": 1.8499602325788245e-05,
+      "loss": 0.0751,
+      "step": 1122
+    },
+    {
+      "epoch": 2.468904788112273,
+      "grad_norm": 0.19270025193691254,
+      "learning_rate": 1.835126538308013e-05,
+      "loss": 0.0247,
+      "step": 1123
+    },
+    {
+      "epoch": 2.4711062190423774,
+      "grad_norm": 0.6644667387008667,
+      "learning_rate": 1.820346544346665e-05,
+      "loss": 0.0925,
+      "step": 1124
+    },
+    {
+      "epoch": 2.473307649972482,
+      "grad_norm": 0.42249488830566406,
+      "learning_rate": 1.805620347902681e-05,
+      "loss": 0.0551,
+      "step": 1125
+    },
+    {
+      "epoch": 2.4755090809025866,
+      "grad_norm": 0.42521026730537415,
+      "learning_rate": 1.7909480458301464e-05,
+      "loss": 0.0317,
+      "step": 1126
+    },
+    {
+      "epoch": 2.4777105118326914,
+      "grad_norm": 0.45745185017585754,
+      "learning_rate": 1.7763297346286735e-05,
+      "loss": 0.0419,
+      "step": 1127
+    },
+    {
+      "epoch": 2.479911942762796,
+      "grad_norm": 0.47970589995384216,
+      "learning_rate": 1.7617655104427832e-05,
+      "loss": 0.034,
+      "step": 1128
+    },
+    {
+      "epoch": 2.4821133736929,
+      "grad_norm": 0.399545282125473,
+      "learning_rate": 1.7472554690612684e-05,
+      "loss": 0.0346,
+      "step": 1129
+    },
+    {
+      "epoch": 2.484314804623005,
+      "grad_norm": 0.3248477280139923,
+      "learning_rate": 1.7327997059165535e-05,
+      "loss": 0.0257,
+      "step": 1130
+    },
+    {
+      "epoch": 2.4865162355531094,
+      "grad_norm": 0.536158561706543,
+      "learning_rate": 1.7183983160840844e-05,
+      "loss": 0.0384,
+      "step": 1131
+    },
+    {
+      "epoch": 2.4887176664832142,
+      "grad_norm": 0.522911548614502,
+      "learning_rate": 1.7040513942816906e-05,
+      "loss": 0.082,
+      "step": 1132
+    },
+    {
+      "epoch": 2.4909190974133186,
+      "grad_norm": 0.3248532712459564,
+      "learning_rate": 1.689759034868961e-05,
+      "loss": 0.0748,
+      "step": 1133
+    },
+    {
+      "epoch": 2.4931205283434235,
+      "grad_norm": 0.7561299800872803,
+      "learning_rate": 1.6755213318466357e-05,
+      "loss": 0.0796,
+      "step": 1134
+    },
+    {
+      "epoch": 2.495321959273528,
+      "grad_norm": 0.4897083044052124,
+      "learning_rate": 1.661338378855979e-05,
+      "loss": 0.0427,
+      "step": 1135
+    },
+    {
+      "epoch": 2.4975233902036322,
+      "grad_norm": 0.7056190967559814,
+      "learning_rate": 1.64721026917816e-05,
+      "loss": 0.07,
+      "step": 1136
+    },
+    {
+      "epoch": 2.499724821133737,
+      "grad_norm": 0.42163705825805664,
+      "learning_rate": 1.6331370957336477e-05,
+      "loss": 0.0522,
+      "step": 1137
+    },
+    {
+      "epoch": 2.5019262520638414,
+      "grad_norm": 0.7106038928031921,
+      "learning_rate": 1.619118951081594e-05,
+      "loss": 0.0894,
+      "step": 1138
+    },
+    {
+      "epoch": 2.5041276829939463,
+      "grad_norm": 0.4210283160209656,
+      "learning_rate": 1.6051559274192275e-05,
+      "loss": 0.0407,
+      "step": 1139
+    },
+    {
+      "epoch": 2.5063291139240507,
+      "grad_norm": 0.5599700808525085,
+      "learning_rate": 1.5912481165812488e-05,
+      "loss": 0.0974,
+      "step": 1140
+    },
+    {
+      "epoch": 2.508530544854155,
+      "grad_norm": 0.5361760258674622,
+      "learning_rate": 1.5773956100392183e-05,
+      "loss": 0.0447,
+      "step": 1141
+    },
+    {
+      "epoch": 2.51073197578426,
+      "grad_norm": 0.7134577631950378,
+      "learning_rate": 1.563598498900969e-05,
+      "loss": 0.0813,
+      "step": 1142
+    },
+    {
+      "epoch": 2.5129334067143643,
+      "grad_norm": 0.5170993804931641,
+      "learning_rate": 1.5498568739099906e-05,
+      "loss": 0.0384,
+      "step": 1143
+    },
+    {
+      "epoch": 2.515134837644469,
+      "grad_norm": 0.7268331050872803,
+      "learning_rate": 1.5361708254448427e-05,
+      "loss": 0.0234,
+      "step": 1144
+    },
+    {
+      "epoch": 2.5173362685745735,
+      "grad_norm": 0.316818505525589,
+      "learning_rate": 1.522540443518563e-05,
+      "loss": 0.0266,
+      "step": 1145
+    },
+    {
+      "epoch": 2.519537699504678,
+      "grad_norm": 0.3618791997432709,
+      "learning_rate": 1.5089658177780653e-05,
+      "loss": 0.0386,
+      "step": 1146
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 0.3421090543270111,
+      "learning_rate": 1.495447037503559e-05,
+      "loss": 0.0299,
+      "step": 1147
+    },
+    {
+      "epoch": 2.523940561364887,
+      "grad_norm": 0.29327401518821716,
+      "learning_rate": 1.481984191607959e-05,
+      "loss": 0.0358,
+      "step": 1148
+    },
+    {
+      "epoch": 2.526141992294992,
+      "grad_norm": 0.21743616461753845,
+      "learning_rate": 1.4685773686362958e-05,
+      "loss": 0.031,
+      "step": 1149
+    },
+    {
+      "epoch": 2.5283434232250963,
+      "grad_norm": 0.5527085661888123,
+      "learning_rate": 1.4552266567651396e-05,
+      "loss": 0.0421,
+      "step": 1150
+    },
+    {
+      "epoch": 2.5305448541552007,
+      "grad_norm": 0.5072147250175476,
+      "learning_rate": 1.441932143802024e-05,
+      "loss": 0.0436,
+      "step": 1151
+    },
+    {
+      "epoch": 2.5327462850853055,
+      "grad_norm": 0.32063746452331543,
+      "learning_rate": 1.4286939171848501e-05,
+      "loss": 0.0378,
+      "step": 1152
+    },
+    {
+      "epoch": 2.53494771601541,
+      "grad_norm": 0.9271298050880432,
+      "learning_rate": 1.415512063981339e-05,
+      "loss": 0.1437,
+      "step": 1153
+    },
+    {
+      "epoch": 2.5371491469455147,
+      "grad_norm": 0.24707064032554626,
+      "learning_rate": 1.4023866708884337e-05,
+      "loss": 0.0291,
+      "step": 1154
+    },
+    {
+      "epoch": 2.539350577875619,
+      "grad_norm": 0.24650706350803375,
+      "learning_rate": 1.3893178242317406e-05,
+      "loss": 0.0392,
+      "step": 1155
+    },
+    {
+      "epoch": 2.5415520088057235,
+      "grad_norm": 0.32058873772621155,
+      "learning_rate": 1.3763056099649674e-05,
+      "loss": 0.0277,
+      "step": 1156
+    },
+    {
+      "epoch": 2.5437534397358283,
+      "grad_norm": 0.36130160093307495,
+      "learning_rate": 1.3633501136693417e-05,
+      "loss": 0.0174,
+      "step": 1157
+    },
+    {
+      "epoch": 2.5459548706659327,
+      "grad_norm": 0.5230188369750977,
+      "learning_rate": 1.350451420553065e-05,
+      "loss": 0.047,
+      "step": 1158
+    },
+    {
+      "epoch": 2.5481563015960376,
+      "grad_norm": 0.2182776927947998,
+      "learning_rate": 1.3376096154507422e-05,
+      "loss": 0.0148,
+      "step": 1159
+    },
+    {
+      "epoch": 2.550357732526142,
+      "grad_norm": 0.21885547041893005,
+      "learning_rate": 1.3248247828228245e-05,
+      "loss": 0.0179,
+      "step": 1160
+    },
+    {
+      "epoch": 2.5525591634562463,
+      "grad_norm": 0.7712263464927673,
+      "learning_rate": 1.3120970067550598e-05,
+      "loss": 0.0687,
+      "step": 1161
+    },
+    {
+      "epoch": 2.554760594386351,
+      "grad_norm": 0.8403673768043518,
+      "learning_rate": 1.2994263709579279e-05,
+      "loss": 0.1235,
+      "step": 1162
+    },
+    {
+      "epoch": 2.5569620253164556,
+      "grad_norm": 0.6125810742378235,
+      "learning_rate": 1.286812958766106e-05,
+      "loss": 0.0381,
+      "step": 1163
+    },
+    {
+      "epoch": 2.5591634562465604,
+      "grad_norm": 0.40800410509109497,
+      "learning_rate": 1.2742568531379084e-05,
+      "loss": 0.0352,
+      "step": 1164
+    },
+    {
+      "epoch": 2.5613648871766648,
+      "grad_norm": 0.5998936891555786,
+      "learning_rate": 1.2617581366547437e-05,
+      "loss": 0.0862,
+      "step": 1165
+    },
+    {
+      "epoch": 2.563566318106769,
+      "grad_norm": 0.3383854925632477,
+      "learning_rate": 1.24931689152057e-05,
+      "loss": 0.0362,
+      "step": 1166
+    },
+    {
+      "epoch": 2.565767749036874,
+      "grad_norm": 0.310473769903183,
+      "learning_rate": 1.2369331995613665e-05,
+      "loss": 0.0348,
+      "step": 1167
+    },
+    {
+      "epoch": 2.5679691799669784,
+      "grad_norm": 0.9457272887229919,
+      "learning_rate": 1.224607142224572e-05,
+      "loss": 0.0623,
+      "step": 1168
+    },
+    {
+      "epoch": 2.570170610897083,
+      "grad_norm": 0.35449472069740295,
+      "learning_rate": 1.212338800578574e-05,
+      "loss": 0.0408,
+      "step": 1169
+    },
+    {
+      "epoch": 2.5723720418271876,
+      "grad_norm": 0.3177931606769562,
+      "learning_rate": 1.200128255312163e-05,
+      "loss": 0.0338,
+      "step": 1170
+    },
+    {
+      "epoch": 2.574573472757292,
+      "grad_norm": 0.47864291071891785,
+      "learning_rate": 1.1879755867339959e-05,
+      "loss": 0.0258,
+      "step": 1171
+    },
+    {
+      "epoch": 2.576774903687397,
+      "grad_norm": 0.48185205459594727,
+      "learning_rate": 1.1758808747720828e-05,
+      "loss": 0.0397,
+      "step": 1172
+    },
+    {
+      "epoch": 2.5789763346175016,
+      "grad_norm": 0.5277111530303955,
+      "learning_rate": 1.1638441989732473e-05,
+      "loss": 0.1052,
+      "step": 1173
+    },
+    {
+      "epoch": 2.581177765547606,
+      "grad_norm": 0.4601213335990906,
+      "learning_rate": 1.151865638502615e-05,
+      "loss": 0.046,
+      "step": 1174
+    },
+    {
+      "epoch": 2.5833791964777104,
+      "grad_norm": 0.7493718266487122,
+      "learning_rate": 1.139945272143087e-05,
+      "loss": 0.0407,
+      "step": 1175
+    },
+    {
+      "epoch": 2.5855806274078152,
+      "grad_norm": 0.12364010512828827,
+      "learning_rate": 1.1280831782948186e-05,
+      "loss": 0.0163,
+      "step": 1176
+    },
+    {
+      "epoch": 2.5877820583379196,
+      "grad_norm": 1.085183024406433,
+      "learning_rate": 1.1162794349747063e-05,
+      "loss": 0.0759,
+      "step": 1177
+    },
+    {
+      "epoch": 2.5899834892680245,
+      "grad_norm": 1.4777817726135254,
+      "learning_rate": 1.1045341198158831e-05,
+      "loss": 0.1314,
+      "step": 1178
+    },
+    {
+      "epoch": 2.592184920198129,
+      "grad_norm": 0.5860021710395813,
+      "learning_rate": 1.092847310067192e-05,
+      "loss": 0.0383,
+      "step": 1179
+    },
+    {
+      "epoch": 2.5943863511282332,
+      "grad_norm": 0.80066978931427,
+      "learning_rate": 1.0812190825926926e-05,
+      "loss": 0.0566,
+      "step": 1180
+    },
+    {
+      "epoch": 2.596587782058338,
+      "grad_norm": 0.5353748202323914,
+      "learning_rate": 1.069649513871147e-05,
+      "loss": 0.0829,
+      "step": 1181
+    },
+    {
+      "epoch": 2.5987892129884425,
+      "grad_norm": 0.4419805407524109,
+      "learning_rate": 1.0581386799955163e-05,
+      "loss": 0.0655,
+      "step": 1182
+    },
+    {
+      "epoch": 2.6009906439185473,
+      "grad_norm": 0.7315040826797485,
+      "learning_rate": 1.0466866566724698e-05,
+      "loss": 0.0421,
+      "step": 1183
+    },
+    {
+      "epoch": 2.6031920748486517,
+      "grad_norm": 0.44108641147613525,
+      "learning_rate": 1.0352935192218716e-05,
+      "loss": 0.0508,
+      "step": 1184
+    },
+    {
+      "epoch": 2.605393505778756,
+      "grad_norm": 0.8455576300621033,
+      "learning_rate": 1.0239593425762995e-05,
+      "loss": 0.1038,
+      "step": 1185
+    },
+    {
+      "epoch": 2.607594936708861,
+      "grad_norm": 0.23156088590621948,
+      "learning_rate": 1.0126842012805492e-05,
+      "loss": 0.0166,
+      "step": 1186
+    },
+    {
+      "epoch": 2.6097963676389653,
+      "grad_norm": 0.6034132242202759,
+      "learning_rate": 1.0014681694911333e-05,
+      "loss": 0.0961,
+      "step": 1187
+    },
+    {
+      "epoch": 2.61199779856907,
+      "grad_norm": 0.5041428804397583,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 0.046,
+      "step": 1188
+    },
+    {
+      "epoch": 2.6141992294991745,
+      "grad_norm": 0.3568980395793915,
+      "learning_rate": 9.79213729113081e-06,
+      "loss": 0.0368,
+      "step": 1189
+    },
+    {
+      "epoch": 2.616400660429279,
+      "grad_norm": 0.13806018233299255,
+      "learning_rate": 9.681754668917231e-06,
+      "loss": 0.0073,
+      "step": 1190
+    },
+    {
+      "epoch": 2.6186020913593837,
+      "grad_norm": 0.4549630880355835,
+      "learning_rate": 9.571966069103045e-06,
+      "loss": 0.0436,
+      "step": 1191
+    },
+    {
+      "epoch": 2.620803522289488,
+      "grad_norm": 0.4328806400299072,
+      "learning_rate": 9.462772213766969e-06,
+      "loss": 0.0267,
+      "step": 1192
+    },
+    {
+      "epoch": 2.623004953219593,
+      "grad_norm": 0.2977164685726166,
+      "learning_rate": 9.354173821076184e-06,
+      "loss": 0.0379,
+      "step": 1193
+    },
+    {
+      "epoch": 2.6252063841496973,
+      "grad_norm": 0.5366312861442566,
+      "learning_rate": 9.246171605281472e-06,
+      "loss": 0.0889,
+      "step": 1194
+    },
+    {
+      "epoch": 2.6274078150798017,
+      "grad_norm": 0.3855116665363312,
+      "learning_rate": 9.138766276712552e-06,
+      "loss": 0.0254,
+      "step": 1195
+    },
+    {
+      "epoch": 2.6296092460099065,
+      "grad_norm": 0.6682915091514587,
+      "learning_rate": 9.031958541773455e-06,
+      "loss": 0.0728,
+      "step": 1196
+    },
+    {
+      "epoch": 2.631810676940011,
+      "grad_norm": 0.4184034466743469,
+      "learning_rate": 8.925749102937831e-06,
+      "loss": 0.0372,
+      "step": 1197
+    },
+    {
+      "epoch": 2.6340121078701157,
+      "grad_norm": 0.5024555921554565,
+      "learning_rate": 8.820138658744304e-06,
+      "loss": 0.065,
+      "step": 1198
+    },
+    {
+      "epoch": 2.63621353880022,
+      "grad_norm": 0.3233795464038849,
+      "learning_rate": 8.715127903791976e-06,
+      "loss": 0.0185,
+      "step": 1199
+    },
+    {
+      "epoch": 2.6384149697303245,
+      "grad_norm": 0.3128807246685028,
+      "learning_rate": 8.610717528735767e-06,
+      "loss": 0.0216,
+      "step": 1200
+    },
+    {
+      "epoch": 2.6406164006604294,
+      "grad_norm": 0.6256663799285889,
+      "learning_rate": 8.506908220281884e-06,
+      "loss": 0.0401,
+      "step": 1201
+    },
+    {
+      "epoch": 2.6428178315905337,
+      "grad_norm": 0.9225179553031921,
+      "learning_rate": 8.403700661183355e-06,
+      "loss": 0.0783,
+      "step": 1202
+    },
+    {
+      "epoch": 2.6450192625206386,
+      "grad_norm": 0.5188078284263611,
+      "learning_rate": 8.301095530235492e-06,
+      "loss": 0.0689,
+      "step": 1203
+    },
+    {
+      "epoch": 2.647220693450743,
+      "grad_norm": 0.49392715096473694,
+      "learning_rate": 8.199093502271449e-06,
+      "loss": 0.031,
+      "step": 1204
+    },
+    {
+      "epoch": 2.6494221243808473,
+      "grad_norm": 0.6362511515617371,
+      "learning_rate": 8.097695248157799e-06,
+      "loss": 0.0629,
+      "step": 1205
+    },
+    {
+      "epoch": 2.651623555310952,
+      "grad_norm": 0.10085786879062653,
+      "learning_rate": 7.996901434790027e-06,
+      "loss": 0.0047,
+      "step": 1206
+    },
+    {
+      "epoch": 2.6538249862410566,
+      "grad_norm": 0.792805552482605,
+      "learning_rate": 7.896712725088285e-06,
+      "loss": 0.0788,
+      "step": 1207
+    },
+    {
+      "epoch": 2.6560264171711614,
+      "grad_norm": 0.32481032609939575,
+      "learning_rate": 7.797129777992952e-06,
+      "loss": 0.0209,
+      "step": 1208
+    },
+    {
+      "epoch": 2.6582278481012658,
+      "grad_norm": 0.586054801940918,
+      "learning_rate": 7.698153248460271e-06,
+      "loss": 0.0507,
+      "step": 1209
+    },
+    {
+      "epoch": 2.66042927903137,
+      "grad_norm": 0.5913929343223572,
+      "learning_rate": 7.599783787458137e-06,
+      "loss": 0.0456,
+      "step": 1210
+    },
+    {
+      "epoch": 2.662630709961475,
+      "grad_norm": 0.605243980884552,
+      "learning_rate": 7.5020220419617315e-06,
+      "loss": 0.0544,
+      "step": 1211
+    },
+    {
+      "epoch": 2.6648321408915794,
+      "grad_norm": 0.5508314967155457,
+      "learning_rate": 7.40486865494926e-06,
+      "loss": 0.0515,
+      "step": 1212
+    },
+    {
+      "epoch": 2.667033571821684,
+      "grad_norm": 0.6130932569503784,
+      "learning_rate": 7.308324265397836e-06,
+      "loss": 0.0582,
+      "step": 1213
+    },
+    {
+      "epoch": 2.6692350027517886,
+      "grad_norm": 0.49070799350738525,
+      "learning_rate": 7.212389508279138e-06,
+      "loss": 0.0478,
+      "step": 1214
+    },
+    {
+      "epoch": 2.671436433681893,
+      "grad_norm": 0.4508554935455322,
+      "learning_rate": 7.1170650145553245e-06,
+      "loss": 0.0343,
+      "step": 1215
+    },
+    {
+      "epoch": 2.673637864611998,
+      "grad_norm": 0.6292645931243896,
+      "learning_rate": 7.022351411174866e-06,
+      "loss": 0.1009,
+      "step": 1216
+    },
+    {
+      "epoch": 2.675839295542102,
+      "grad_norm": 0.9365841150283813,
+      "learning_rate": 6.928249321068381e-06,
+      "loss": 0.0665,
+      "step": 1217
+    },
+    {
+      "epoch": 2.678040726472207,
+      "grad_norm": 1.076918601989746,
+      "learning_rate": 6.834759363144594e-06,
+      "loss": 0.1623,
+      "step": 1218
+    },
+    {
+      "epoch": 2.6802421574023114,
+      "grad_norm": 0.5346421599388123,
+      "learning_rate": 6.741882152286227e-06,
+      "loss": 0.0347,
+      "step": 1219
+    },
+    {
+      "epoch": 2.682443588332416,
+      "grad_norm": 0.3021618723869324,
+      "learning_rate": 6.6496182993459836e-06,
+      "loss": 0.0486,
+      "step": 1220
+    },
+    {
+      "epoch": 2.6846450192625206,
+      "grad_norm": 0.08443254232406616,
+      "learning_rate": 6.557968411142523e-06,
+      "loss": 0.0106,
+      "step": 1221
+    },
+    {
+      "epoch": 2.6868464501926255,
+      "grad_norm": 0.3483668267726898,
+      "learning_rate": 6.466933090456428e-06,
+      "loss": 0.0177,
+      "step": 1222
+    },
+    {
+      "epoch": 2.68904788112273,
+      "grad_norm": 0.4496798515319824,
+      "learning_rate": 6.37651293602628e-06,
+      "loss": 0.0386,
+      "step": 1223
+    },
+    {
+      "epoch": 2.6912493120528342,
+      "grad_norm": 0.4968661665916443,
+      "learning_rate": 6.286708542544784e-06,
+      "loss": 0.0328,
+      "step": 1224
+    },
+    {
+      "epoch": 2.693450742982939,
+      "grad_norm": 0.6091892719268799,
+      "learning_rate": 6.197520500654685e-06,
+      "loss": 0.0691,
+      "step": 1225
+    },
+    {
+      "epoch": 2.6956521739130435,
+      "grad_norm": 0.3410601317882538,
+      "learning_rate": 6.108949396945074e-06,
+      "loss": 0.0303,
+      "step": 1226
+    },
+    {
+      "epoch": 2.6978536048431483,
+      "grad_norm": 0.8630918264389038,
+      "learning_rate": 6.020995813947428e-06,
+      "loss": 0.1078,
+      "step": 1227
+    },
+    {
+      "epoch": 2.7000550357732527,
+      "grad_norm": 0.7620694637298584,
+      "learning_rate": 5.933660330131752e-06,
+      "loss": 0.1604,
+      "step": 1228
+    },
+    {
+      "epoch": 2.702256466703357,
+      "grad_norm": 0.43725255131721497,
+      "learning_rate": 5.846943519902859e-06,
+      "loss": 0.05,
+      "step": 1229
+    },
+    {
+      "epoch": 2.704457897633462,
+      "grad_norm": 0.48732689023017883,
+      "learning_rate": 5.760845953596527e-06,
+      "loss": 0.0572,
+      "step": 1230
+    },
+    {
+      "epoch": 2.7066593285635663,
+      "grad_norm": 0.5541000366210938,
+      "learning_rate": 5.675368197475783e-06,
+      "loss": 0.0676,
+      "step": 1231
+    },
+    {
+      "epoch": 2.708860759493671,
+      "grad_norm": 0.32182785868644714,
+      "learning_rate": 5.59051081372719e-06,
+      "loss": 0.0192,
+      "step": 1232
+    },
+    {
+      "epoch": 2.7110621904237755,
+      "grad_norm": 0.3678714334964752,
+      "learning_rate": 5.506274360457086e-06,
+      "loss": 0.0476,
+      "step": 1233
+    },
+    {
+      "epoch": 2.71326362135388,
+      "grad_norm": 0.3759974539279938,
+      "learning_rate": 5.422659391687956e-06,
+      "loss": 0.0325,
+      "step": 1234
+    },
+    {
+      "epoch": 2.7154650522839847,
+      "grad_norm": 0.5801433324813843,
+      "learning_rate": 5.33966645735483e-06,
+      "loss": 0.0653,
+      "step": 1235
+    },
+    {
+      "epoch": 2.717666483214089,
+      "grad_norm": 0.2991897165775299,
+      "learning_rate": 5.257296103301568e-06,
+      "loss": 0.0419,
+      "step": 1236
+    },
+    {
+      "epoch": 2.719867914144194,
+      "grad_norm": 0.8437144756317139,
+      "learning_rate": 5.175548871277358e-06,
+      "loss": 0.0509,
+      "step": 1237
+    },
+    {
+      "epoch": 2.7220693450742983,
+      "grad_norm": 0.8704203367233276,
+      "learning_rate": 5.094425298933136e-06,
+      "loss": 0.1519,
+      "step": 1238
+    },
+    {
+      "epoch": 2.7242707760044027,
+      "grad_norm": 0.6470730304718018,
+      "learning_rate": 5.013925919817975e-06,
+      "loss": 0.0686,
+      "step": 1239
+    },
+    {
+      "epoch": 2.7264722069345075,
+      "grad_norm": 0.6081740856170654,
+      "learning_rate": 4.9340512633756945e-06,
+      "loss": 0.0731,
+      "step": 1240
+    },
+    {
+      "epoch": 2.728673637864612,
+      "grad_norm": 0.26082944869995117,
+      "learning_rate": 4.854801854941249e-06,
+      "loss": 0.0333,
+      "step": 1241
+    },
+    {
+      "epoch": 2.7308750687947168,
+      "grad_norm": 1.0025891065597534,
+      "learning_rate": 4.776178215737426e-06,
+      "loss": 0.0796,
+      "step": 1242
+    },
+    {
+      "epoch": 2.733076499724821,
+      "grad_norm": 1.0358490943908691,
+      "learning_rate": 4.698180862871282e-06,
+      "loss": 0.0575,
+      "step": 1243
+    },
+    {
+      "epoch": 2.7352779306549255,
+      "grad_norm": 0.6997416019439697,
+      "learning_rate": 4.620810309330803e-06,
+      "loss": 0.0613,
+      "step": 1244
+    },
+    {
+      "epoch": 2.7374793615850304,
+      "grad_norm": 0.7522681355476379,
+      "learning_rate": 4.544067063981494e-06,
+      "loss": 0.0959,
+      "step": 1245
+    },
+    {
+      "epoch": 2.7396807925151347,
+      "grad_norm": 0.7872377634048462,
+      "learning_rate": 4.467951631563117e-06,
+      "loss": 0.0907,
+      "step": 1246
+    },
+    {
+      "epoch": 2.7418822234452396,
+      "grad_norm": 0.77159184217453,
+      "learning_rate": 4.392464512686257e-06,
+      "loss": 0.0453,
+      "step": 1247
+    },
+    {
+      "epoch": 2.744083654375344,
+      "grad_norm": 0.8469691276550293,
+      "learning_rate": 4.317606203829127e-06,
+      "loss": 0.0953,
+      "step": 1248
+    },
+    {
+      "epoch": 2.7462850853054483,
+      "grad_norm": 0.45564761757850647,
+      "learning_rate": 4.243377197334231e-06,
+      "loss": 0.0382,
+      "step": 1249
+    },
+    {
+      "epoch": 2.748486516235553,
+      "grad_norm": 0.3689277470111847,
+      "learning_rate": 4.1697779814051765e-06,
+      "loss": 0.0273,
+      "step": 1250
+    },
+    {
+      "epoch": 2.7506879471656576,
+      "grad_norm": 0.664315402507782,
+      "learning_rate": 4.096809040103444e-06,
+      "loss": 0.106,
+      "step": 1251
+    },
+    {
+      "epoch": 2.7528893780957624,
+      "grad_norm": 0.34314075112342834,
+      "learning_rate": 4.024470853345175e-06,
+      "loss": 0.0332,
+      "step": 1252
+    },
+    {
+      "epoch": 2.755090809025867,
+      "grad_norm": 0.1801203191280365,
+      "learning_rate": 3.952763896898071e-06,
+      "loss": 0.0202,
+      "step": 1253
+    },
+    {
+      "epoch": 2.757292239955971,
+      "grad_norm": 0.9782657027244568,
+      "learning_rate": 3.881688642378245e-06,
+      "loss": 0.0512,
+      "step": 1254
+    },
+    {
+      "epoch": 2.759493670886076,
+      "grad_norm": 0.8883532285690308,
+      "learning_rate": 3.8112455572470672e-06,
+      "loss": 0.0971,
+      "step": 1255
+    },
+    {
+      "epoch": 2.7616951018161804,
+      "grad_norm": 0.2342640459537506,
+      "learning_rate": 3.7414351048081996e-06,
+      "loss": 0.0169,
+      "step": 1256
+    },
+    {
+      "epoch": 2.763896532746285,
+      "grad_norm": 0.5581437349319458,
+      "learning_rate": 3.6722577442044102e-06,
+      "loss": 0.0539,
+      "step": 1257
+    },
+    {
+      "epoch": 2.7660979636763896,
+      "grad_norm": 0.6687867641448975,
+      "learning_rate": 3.6037139304146762e-06,
+      "loss": 0.1083,
+      "step": 1258
+    },
+    {
+      "epoch": 2.768299394606494,
+      "grad_norm": 0.6550372838973999,
+      "learning_rate": 3.53580411425114e-06,
+      "loss": 0.0693,
+      "step": 1259
+    },
+    {
+      "epoch": 2.770500825536599,
+      "grad_norm": 0.2852613031864166,
+      "learning_rate": 3.468528742356103e-06,
+      "loss": 0.0277,
+      "step": 1260
+    },
+    {
+      "epoch": 2.772702256466703,
+      "grad_norm": 0.6610994338989258,
+      "learning_rate": 3.401888257199204e-06,
+      "loss": 0.0704,
+      "step": 1261
+    },
+    {
+      "epoch": 2.774903687396808,
+      "grad_norm": 0.5986510515213013,
+      "learning_rate": 3.3358830970743882e-06,
+      "loss": 0.0683,
+      "step": 1262
+    },
+    {
+      "epoch": 2.7771051183269124,
+      "grad_norm": 0.3271762430667877,
+      "learning_rate": 3.270513696097055e-06,
+      "loss": 0.0474,
+      "step": 1263
+    },
+    {
+      "epoch": 2.779306549257017,
+      "grad_norm": 0.6686267256736755,
+      "learning_rate": 3.205780484201293e-06,
+      "loss": 0.0584,
+      "step": 1264
+    },
+    {
+      "epoch": 2.7815079801871216,
+      "grad_norm": 0.7449634075164795,
+      "learning_rate": 3.1416838871368924e-06,
+      "loss": 0.1396,
+      "step": 1265
+    },
+    {
+      "epoch": 2.7837094111172265,
+      "grad_norm": 0.6699150204658508,
+      "learning_rate": 3.0782243264666944e-06,
+      "loss": 0.027,
+      "step": 1266
+    },
+    {
+      "epoch": 2.785910842047331,
+      "grad_norm": 0.22182409465312958,
+      "learning_rate": 3.0154022195637454e-06,
+      "loss": 0.0306,
+      "step": 1267
+    },
+    {
+      "epoch": 2.7881122729774352,
+      "grad_norm": 0.630264401435852,
+      "learning_rate": 2.9532179796085356e-06,
+      "loss": 0.1014,
+      "step": 1268
+    },
+    {
+      "epoch": 2.7903137039075396,
+      "grad_norm": 0.2437404841184616,
+      "learning_rate": 2.891672015586344e-06,
+      "loss": 0.0143,
+      "step": 1269
+    },
+    {
+      "epoch": 2.7925151348376445,
+      "grad_norm": 0.10994002968072891,
+      "learning_rate": 2.8307647322844964e-06,
+      "loss": 0.0131,
+      "step": 1270
+    },
+    {
+      "epoch": 2.7947165657677493,
+      "grad_norm": 0.22370854020118713,
+      "learning_rate": 2.7704965302897233e-06,
+      "loss": 0.0122,
+      "step": 1271
+    },
+    {
+      "epoch": 2.7969179966978537,
+      "grad_norm": 0.6404604315757751,
+      "learning_rate": 2.7108678059855065e-06,
+      "loss": 0.0553,
+      "step": 1272
+    },
+    {
+      "epoch": 2.799119427627958,
+      "grad_norm": 0.7777275443077087,
+      "learning_rate": 2.651878951549536e-06,
+      "loss": 0.0796,
+      "step": 1273
+    },
+    {
+      "epoch": 2.801320858558063,
+      "grad_norm": 0.4582979679107666,
+      "learning_rate": 2.5935303549510126e-06,
+      "loss": 0.0499,
+      "step": 1274
+    },
+    {
+      "epoch": 2.8035222894881673,
+      "grad_norm": 0.45491763949394226,
+      "learning_rate": 2.53582239994824e-06,
+      "loss": 0.0432,
+      "step": 1275
+    },
+    {
+      "epoch": 2.805723720418272,
+      "grad_norm": 0.4901306629180908,
+      "learning_rate": 2.478755466085958e-06,
+      "loss": 0.0416,
+      "step": 1276
+    },
+    {
+      "epoch": 2.8079251513483765,
+      "grad_norm": 0.43012434244155884,
+      "learning_rate": 2.422329928692957e-06,
+      "loss": 0.0354,
+      "step": 1277
+    },
+    {
+      "epoch": 2.810126582278481,
+      "grad_norm": 0.9710955619812012,
+      "learning_rate": 2.36654615887959e-06,
+      "loss": 0.0901,
+      "step": 1278
+    },
+    {
+      "epoch": 2.8123280132085857,
+      "grad_norm": 1.6936571598052979,
+      "learning_rate": 2.311404523535243e-06,
+      "loss": 0.1524,
+      "step": 1279
+    },
+    {
+      "epoch": 2.81452944413869,
+      "grad_norm": 0.965338945388794,
+      "learning_rate": 2.2569053853260246e-06,
+      "loss": 0.1646,
+      "step": 1280
+    },
+    {
+      "epoch": 2.816730875068795,
+      "grad_norm": 0.7708274722099304,
+      "learning_rate": 2.2030491026923672e-06,
+      "loss": 0.0516,
+      "step": 1281
+    },
+    {
+      "epoch": 2.8189323059988993,
+      "grad_norm": 0.43852585554122925,
+      "learning_rate": 2.149836029846608e-06,
+      "loss": 0.042,
+      "step": 1282
+    },
+    {
+      "epoch": 2.8211337369290037,
+      "grad_norm": 0.11031460762023926,
+      "learning_rate": 2.0972665167707126e-06,
+      "loss": 0.0115,
+      "step": 1283
+    },
+    {
+      "epoch": 2.8233351678591085,
+      "grad_norm": 0.5105849504470825,
+      "learning_rate": 2.0453409092139662e-06,
+      "loss": 0.0348,
+      "step": 1284
+    },
+    {
+      "epoch": 2.825536598789213,
+      "grad_norm": 0.44641202688217163,
+      "learning_rate": 1.9940595486906964e-06,
+      "loss": 0.055,
+      "step": 1285
+    },
+    {
+      "epoch": 2.8277380297193178,
+      "grad_norm": 0.7521995902061462,
+      "learning_rate": 1.9434227724779984e-06,
+      "loss": 0.1221,
+      "step": 1286
+    },
+    {
+      "epoch": 2.829939460649422,
+      "grad_norm": 0.6151113510131836,
+      "learning_rate": 1.8934309136135697e-06,
+      "loss": 0.0944,
+      "step": 1287
+    },
+    {
+      "epoch": 2.8321408915795265,
+      "grad_norm": 0.6635796427726746,
+      "learning_rate": 1.8440843008934561e-06,
+      "loss": 0.1008,
+      "step": 1288
+    },
+    {
+      "epoch": 2.8343423225096314,
+      "grad_norm": 0.1558506190776825,
+      "learning_rate": 1.795383258869987e-06,
+      "loss": 0.0114,
+      "step": 1289
+    },
+    {
+      "epoch": 2.8365437534397357,
+      "grad_norm": 0.5053389072418213,
+      "learning_rate": 1.7473281078494997e-06,
+      "loss": 0.07,
+      "step": 1290
+    },
+    {
+      "epoch": 2.8387451843698406,
+      "grad_norm": 0.594647228717804,
+      "learning_rate": 1.6999191638903623e-06,
+      "loss": 0.0662,
+      "step": 1291
+    },
+    {
+      "epoch": 2.840946615299945,
+      "grad_norm": 0.3465157747268677,
+      "learning_rate": 1.6531567388008429e-06,
+      "loss": 0.0322,
+      "step": 1292
+    },
+    {
+      "epoch": 2.8431480462300494,
+      "grad_norm": 0.38056671619415283,
+      "learning_rate": 1.6070411401370334e-06,
+      "loss": 0.0373,
+      "step": 1293
+    },
+    {
+      "epoch": 2.845349477160154,
+      "grad_norm": 0.88359135389328,
+      "learning_rate": 1.561572671200884e-06,
+      "loss": 0.0647,
+      "step": 1294
+    },
+    {
+      "epoch": 2.8475509080902586,
+      "grad_norm": 1.0830795764923096,
+      "learning_rate": 1.516751631038138e-06,
+      "loss": 0.0619,
+      "step": 1295
+    },
+    {
+      "epoch": 2.8497523390203634,
+      "grad_norm": 0.4293959438800812,
+      "learning_rate": 1.4725783144364568e-06,
+      "loss": 0.0624,
+      "step": 1296
+    },
+    {
+      "epoch": 2.851953769950468,
+      "grad_norm": 0.2833307683467865,
+      "learning_rate": 1.429053011923387e-06,
+      "loss": 0.0271,
+      "step": 1297
+    },
+    {
+      "epoch": 2.854155200880572,
+      "grad_norm": 0.427614688873291,
+      "learning_rate": 1.386176009764506e-06,
+      "loss": 0.0384,
+      "step": 1298
+    },
+    {
+      "epoch": 2.856356631810677,
+      "grad_norm": 1.211296796798706,
+      "learning_rate": 1.343947589961514e-06,
+      "loss": 0.1,
+      "step": 1299
+    },
+    {
+      "epoch": 2.8585580627407814,
+      "grad_norm": 0.33716437220573425,
+      "learning_rate": 1.3023680302504338e-06,
+      "loss": 0.0477,
+      "step": 1300
+    },
+    {
+      "epoch": 2.8607594936708862,
+      "grad_norm": 0.825700044631958,
+      "learning_rate": 1.26143760409968e-06,
+      "loss": 0.04,
+      "step": 1301
+    },
+    {
+      "epoch": 2.8629609246009906,
+      "grad_norm": 0.6166131496429443,
+      "learning_rate": 1.2211565807083492e-06,
+      "loss": 0.0391,
+      "step": 1302
+    },
+    {
+      "epoch": 2.865162355531095,
+      "grad_norm": 0.308035671710968,
+      "learning_rate": 1.1815252250044318e-06,
+      "loss": 0.0381,
+      "step": 1303
+    },
+    {
+      "epoch": 2.8673637864612,
+      "grad_norm": 0.49501779675483704,
+      "learning_rate": 1.1425437976430365e-06,
+      "loss": 0.0519,
+      "step": 1304
+    },
+    {
+      "epoch": 2.869565217391304,
+      "grad_norm": 0.3320211172103882,
+      "learning_rate": 1.1042125550047134e-06,
+      "loss": 0.0385,
+      "step": 1305
+    },
+    {
+      "epoch": 2.871766648321409,
+      "grad_norm": 0.26195937395095825,
+      "learning_rate": 1.0665317491937553e-06,
+      "loss": 0.0212,
+      "step": 1306
+    },
+    {
+      "epoch": 2.8739680792515134,
+      "grad_norm": 0.20724286139011383,
+      "learning_rate": 1.0295016280365112e-06,
+      "loss": 0.0251,
+      "step": 1307
+    },
+    {
+      "epoch": 2.876169510181618,
+      "grad_norm": 0.3688121736049652,
+      "learning_rate": 9.931224350798185e-07,
+      "loss": 0.0339,
+      "step": 1308
+    },
+    {
+      "epoch": 2.8783709411117226,
+      "grad_norm": 0.43073609471321106,
+      "learning_rate": 9.573944095893294e-07,
+      "loss": 0.0331,
+      "step": 1309
+    },
+    {
+      "epoch": 2.880572372041827,
+      "grad_norm": 0.9119531512260437,
+      "learning_rate": 9.223177865480215e-07,
+      "loss": 0.0563,
+      "step": 1310
+    },
+    {
+      "epoch": 2.882773802971932,
+      "grad_norm": 0.911408007144928,
+      "learning_rate": 8.878927966545326e-07,
+      "loss": 0.1245,
+      "step": 1311
+    },
+    {
+      "epoch": 2.8849752339020363,
+      "grad_norm": 0.6210458278656006,
+      "learning_rate": 8.541196663217954e-07,
+      "loss": 0.0648,
+      "step": 1312
+    },
+    {
+      "epoch": 2.8871766648321406,
+      "grad_norm": 0.4914691746234894,
+      "learning_rate": 8.209986176753948e-07,
+      "loss": 0.1079,
+      "step": 1313
+    },
+    {
+      "epoch": 2.8893780957622455,
+      "grad_norm": 0.4978880286216736,
+      "learning_rate": 7.885298685522235e-07,
+      "loss": 0.0528,
+      "step": 1314
+    },
+    {
+      "epoch": 2.8915795266923503,
+      "grad_norm": 0.31169232726097107,
+      "learning_rate": 7.567136324989844e-07,
+      "loss": 0.0251,
+      "step": 1315
+    },
+    {
+      "epoch": 2.8937809576224547,
+      "grad_norm": 1.253857135772705,
+      "learning_rate": 7.255501187708347e-07,
+      "loss": 0.0566,
+      "step": 1316
+    },
+    {
+      "epoch": 2.895982388552559,
+      "grad_norm": 0.615176796913147,
+      "learning_rate": 6.95039532329933e-07,
+      "loss": 0.0412,
+      "step": 1317
+    },
+    {
+      "epoch": 2.8981838194826635,
+      "grad_norm": 1.291667103767395,
+      "learning_rate": 6.651820738441949e-07,
+      "loss": 0.0761,
+      "step": 1318
+    },
+    {
+      "epoch": 2.9003852504127683,
+      "grad_norm": 0.4518704116344452,
+      "learning_rate": 6.359779396858945e-07,
+      "loss": 0.0566,
+      "step": 1319
+    },
+    {
+      "epoch": 2.902586681342873,
+      "grad_norm": 0.28910595178604126,
+      "learning_rate": 6.074273219303761e-07,
+      "loss": 0.0138,
+      "step": 1320
+    },
+    {
+      "epoch": 2.9047881122729775,
+      "grad_norm": 0.581666886806488,
+      "learning_rate": 5.795304083548559e-07,
+      "loss": 0.0331,
+      "step": 1321
+    },
+    {
+      "epoch": 2.906989543203082,
+      "grad_norm": 0.07092265039682388,
+      "learning_rate": 5.522873824370889e-07,
+      "loss": 0.0059,
+      "step": 1322
+    },
+    {
+      "epoch": 2.9091909741331867,
+      "grad_norm": 0.4092436730861664,
+      "learning_rate": 5.256984233542595e-07,
+      "loss": 0.0479,
+      "step": 1323
+    },
+    {
+      "epoch": 2.911392405063291,
+      "grad_norm": 0.43878406286239624,
+      "learning_rate": 4.997637059817595e-07,
+      "loss": 0.0419,
+      "step": 1324
+    },
+    {
+      "epoch": 2.913593835993396,
+      "grad_norm": 0.28752970695495605,
+      "learning_rate": 4.744834008920007e-07,
+      "loss": 0.0106,
+      "step": 1325
+    },
+    {
+      "epoch": 2.9157952669235003,
+      "grad_norm": 0.733717679977417,
+      "learning_rate": 4.498576743533711e-07,
+      "loss": 0.0753,
+      "step": 1326
+    },
+    {
+      "epoch": 2.9179966978536047,
+      "grad_norm": 0.32417628169059753,
+      "learning_rate": 4.258866883290691e-07,
+      "loss": 0.0571,
+      "step": 1327
+    },
+    {
+      "epoch": 2.9201981287837095,
+      "grad_norm": 0.30484825372695923,
+      "learning_rate": 4.025706004760932e-07,
+      "loss": 0.0394,
+      "step": 1328
+    },
+    {
+      "epoch": 2.922399559713814,
+      "grad_norm": 0.5322850346565247,
+      "learning_rate": 3.799095641441763e-07,
+      "loss": 0.1103,
+      "step": 1329
+    },
+    {
+      "epoch": 2.9246009906439188,
+      "grad_norm": 0.3624371290206909,
+      "learning_rate": 3.5790372837477546e-07,
+      "loss": 0.035,
+      "step": 1330
+    },
+    {
+      "epoch": 2.926802421574023,
+      "grad_norm": 0.4711335599422455,
+      "learning_rate": 3.3655323790008354e-07,
+      "loss": 0.0574,
+      "step": 1331
+    },
+    {
+      "epoch": 2.9290038525041275,
+      "grad_norm": 0.6064951419830322,
+      "learning_rate": 3.1585823314214115e-07,
+      "loss": 0.0794,
+      "step": 1332
+    },
+    {
+      "epoch": 2.9312052834342324,
+      "grad_norm": 0.3930189907550812,
+      "learning_rate": 2.9581885021181533e-07,
+      "loss": 0.0472,
+      "step": 1333
+    },
+    {
+      "epoch": 2.9334067143643368,
+      "grad_norm": 0.38262468576431274,
+      "learning_rate": 2.7643522090796683e-07,
+      "loss": 0.0472,
+      "step": 1334
+    },
+    {
+      "epoch": 2.9356081452944416,
+      "grad_norm": 0.42688238620758057,
+      "learning_rate": 2.577074727165951e-07,
+      "loss": 0.017,
+      "step": 1335
+    },
+    {
+      "epoch": 2.937809576224546,
+      "grad_norm": 0.538941502571106,
+      "learning_rate": 2.3963572880995044e-07,
+      "loss": 0.0359,
+      "step": 1336
+    },
+    {
+      "epoch": 2.9400110071546504,
+      "grad_norm": 0.6182091236114502,
+      "learning_rate": 2.2222010804573423e-07,
+      "loss": 0.0486,
+      "step": 1337
+    },
+    {
+      "epoch": 2.942212438084755,
+      "grad_norm": 0.4431805908679962,
+      "learning_rate": 2.054607249663665e-07,
+      "loss": 0.0486,
+      "step": 1338
+    },
+    {
+      "epoch": 2.9444138690148596,
+      "grad_norm": 0.8684697151184082,
+      "learning_rate": 1.8935768979817524e-07,
+      "loss": 0.1061,
+      "step": 1339
+    },
+    {
+      "epoch": 2.9466152999449644,
+      "grad_norm": 0.8039331436157227,
+      "learning_rate": 1.7391110845069725e-07,
+      "loss": 0.0541,
+      "step": 1340
+    },
+    {
+      "epoch": 2.948816730875069,
+      "grad_norm": 0.8392862677574158,
+      "learning_rate": 1.5912108251597834e-07,
+      "loss": 0.1627,
+      "step": 1341
+    },
+    {
+      "epoch": 2.951018161805173,
+      "grad_norm": 0.43912941217422485,
+      "learning_rate": 1.449877092679075e-07,
+      "loss": 0.0321,
+      "step": 1342
+    },
+    {
+      "epoch": 2.953219592735278,
+      "grad_norm": 0.7228869795799255,
+      "learning_rate": 1.3151108166156168e-07,
+      "loss": 0.1113,
+      "step": 1343
+    },
+    {
+      "epoch": 2.9554210236653824,
+      "grad_norm": 0.6244905591011047,
+      "learning_rate": 1.1869128833262854e-07,
+      "loss": 0.0542,
+      "step": 1344
+    },
+    {
+      "epoch": 2.9576224545954872,
+      "grad_norm": 0.3682965040206909,
+      "learning_rate": 1.0652841359678478e-07,
+      "loss": 0.0257,
+      "step": 1345
+    },
+    {
+      "epoch": 2.9598238855255916,
+      "grad_norm": 0.32940956950187683,
+      "learning_rate": 9.502253744916311e-08,
+      "loss": 0.0308,
+      "step": 1346
+    },
+    {
+      "epoch": 2.962025316455696,
+      "grad_norm": 0.4789118766784668,
+      "learning_rate": 8.417373556381946e-08,
+      "loss": 0.0315,
+      "step": 1347
+    },
+    {
+      "epoch": 2.964226747385801,
+      "grad_norm": 1.1318228244781494,
+      "learning_rate": 7.39820792932333e-08,
+      "loss": 0.0997,
+      "step": 1348
+    },
+    {
+      "epoch": 2.966428178315905,
+      "grad_norm": 0.30917805433273315,
+      "learning_rate": 6.444763566786361e-08,
+      "loss": 0.0333,
+      "step": 1349
+    },
+    {
+      "epoch": 2.96862960924601,
+      "grad_norm": 0.6403358578681946,
+      "learning_rate": 5.55704673956714e-08,
+      "loss": 0.0581,
+      "step": 1350
+    },
+    {
+      "epoch": 2.9708310401761144,
+      "grad_norm": 0.20618505775928497,
+      "learning_rate": 4.7350632861709044e-08,
+      "loss": 0.024,
+      "step": 1351
+    },
+    {
+      "epoch": 2.973032471106219,
+      "grad_norm": 0.5780773758888245,
+      "learning_rate": 3.978818612777602e-08,
+      "loss": 0.0734,
+      "step": 1352
+    },
+    {
+      "epoch": 2.9752339020363237,
+      "grad_norm": 0.6004192233085632,
+      "learning_rate": 3.288317693201926e-08,
+      "loss": 0.0396,
+      "step": 1353
+    },
+    {
+      "epoch": 2.977435332966428,
+      "grad_norm": 0.41798269748687744,
+      "learning_rate": 2.6635650688633384e-08,
+      "loss": 0.0315,
+      "step": 1354
+    },
+    {
+      "epoch": 2.979636763896533,
+      "grad_norm": 0.2113114893436432,
+      "learning_rate": 2.1045648487549864e-08,
+      "loss": 0.0221,
+      "step": 1355
+    },
+    {
+      "epoch": 2.9818381948266373,
+      "grad_norm": 0.5978252291679382,
+      "learning_rate": 1.6113207094181626e-08,
+      "loss": 0.0544,
+      "step": 1356
+    },
+    {
+      "epoch": 2.9840396257567416,
+      "grad_norm": 0.10408246517181396,
+      "learning_rate": 1.1838358949134431e-08,
+      "loss": 0.0078,
+      "step": 1357
+    },
+    {
+      "epoch": 2.9862410566868465,
+      "grad_norm": 0.23431262373924255,
+      "learning_rate": 8.221132168073631e-09,
+      "loss": 0.0204,
+      "step": 1358
+    },
+    {
+      "epoch": 2.988442487616951,
+      "grad_norm": 0.37034013867378235,
+      "learning_rate": 5.261550541457716e-09,
+      "loss": 0.0417,
+      "step": 1359
+    },
+    {
+      "epoch": 2.9906439185470557,
+      "grad_norm": 0.6617506146430969,
+      "learning_rate": 2.9596335344383996e-09,
+      "loss": 0.0349,
+      "step": 1360
+    },
+    {
+      "epoch": 2.99284534947716,
+      "grad_norm": 0.45733800530433655,
+      "learning_rate": 1.315396286705184e-09,
+      "loss": 0.0461,
+      "step": 1361
+    },
+    {
+      "epoch": 2.9950467804072645,
+      "grad_norm": 0.48512786626815796,
+      "learning_rate": 3.2884961238544007e-10,
+      "loss": 0.0484,
+      "step": 1362
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1362,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.260686554725018e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}